winit/src/platform_impl/windows/event_loop.rs

Ignoring revisions in .git-blame-ignore-revs. Click here to bypass and see the normal blame view.

2781 lines
112 KiB
Rust
Raw Normal View History

#![allow(non_snake_case)]
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
mod runner;
use std::cell::Cell;
use std::ffi::c_void;
Improve waiting for messages on Windows Previous version used [`SetTimer`] with `GetMessageW` for waiting. The downside of UI timers like ones created by `SetTimer`, is that they may be late by up to 15-16 ms. To fix this behaviour, I added use of high resolution timers created by [`CreateWaitableTimerExW`] with the flag `CREATE_WAITABLE_TIMER_HIGH_RESOLUTION`. In my previous experience, waiting on such timers have precision of roundly 0.5 ms which is the best available on Windows at the moment. I use [`MsgWaitForMultipleObjectsEx`] to wait simultaneously for both timer and newly arriving events. Unfortunately, high resolution timers are available only since Windows 10 1803. However: 1. Win 10 is already getting to the end of support, like all previous versions, so it is OK to rely on APIs introduced in it; 2. I use `dwMilliseconds` parameter of `MsgWaitForMultipleObjectsEx` as a fallback. It should perform not worse compared to waiting for events from `SetTimer`. I also refactored code to remove event dispatching from function responsible for waiting for events. This provides more clear separations of concern and avoids unnecessary duplication of dispatching logic. After [review] from @rib, I also moved the waiting itself from `wait_for_messages` method to separate function, so it is clearly seen that `wait_for_messages` do 3 things: notify app that we about to wait, wait, notify that we have new events. I have tested behaviour using a egui app with Vulkan rendering with `VK_PRESENT_MODE_IMMEDIATE_KHR`, and older version consistently have twice less FPS than requested (e.g. 30 FPS when limit is 60 and 60 FPS when limit is 120) while newer version works more correctly (almost always 60 FPS when limit is 60, and only 5-10 frames missing when FPS is set to 120 or more). Fixes https://github.com/rust-windowing/winit/issues/1610 [`CreateWaitableTimerExW`]: https://learn.microsoft.com/en-us/windows/win32/api/synchapi/nf-synchapi-createwaitabletimerexw [`MsgWaitForMultipleObjectsEx`]: https://learn.microsoft.com/en-us/windows/win32/api/winuser/nf-winuser-msgwaitformultipleobjectsex [`SetTimer`]: https://learn.microsoft.com/en-us/windows/win32/api/winuser/nf-winuser-settimer [review]: https://github.com/rust-windowing/winit/pull/3950#discussion_r1800184479
2024-10-13 12:56:24 +02:00
use std::os::windows::io::{AsRawHandle as _, FromRawHandle as _, OwnedHandle, RawHandle};
use std::rc::Rc;
use std::sync::atomic::{AtomicU32, Ordering};
2022-08-31 18:32:19 +02:00
use std::sync::{Arc, Mutex, MutexGuard};
use std::time::{Duration, Instant};
use std::{mem, panic, ptr};
use runner::EventLoopRunner;
2022-03-07 22:58:12 +01:00
use windows_sys::Win32::Devices::HumanInterfaceDevice::MOUSE_MOVE_RELATIVE;
Improve waiting for messages on Windows Previous version used [`SetTimer`] with `GetMessageW` for waiting. The downside of UI timers like ones created by `SetTimer`, is that they may be late by up to 15-16 ms. To fix this behaviour, I added use of high resolution timers created by [`CreateWaitableTimerExW`] with the flag `CREATE_WAITABLE_TIMER_HIGH_RESOLUTION`. In my previous experience, waiting on such timers have precision of roundly 0.5 ms which is the best available on Windows at the moment. I use [`MsgWaitForMultipleObjectsEx`] to wait simultaneously for both timer and newly arriving events. Unfortunately, high resolution timers are available only since Windows 10 1803. However: 1. Win 10 is already getting to the end of support, like all previous versions, so it is OK to rely on APIs introduced in it; 2. I use `dwMilliseconds` parameter of `MsgWaitForMultipleObjectsEx` as a fallback. It should perform not worse compared to waiting for events from `SetTimer`. I also refactored code to remove event dispatching from function responsible for waiting for events. This provides more clear separations of concern and avoids unnecessary duplication of dispatching logic. After [review] from @rib, I also moved the waiting itself from `wait_for_messages` method to separate function, so it is clearly seen that `wait_for_messages` do 3 things: notify app that we about to wait, wait, notify that we have new events. I have tested behaviour using a egui app with Vulkan rendering with `VK_PRESENT_MODE_IMMEDIATE_KHR`, and older version consistently have twice less FPS than requested (e.g. 30 FPS when limit is 60 and 60 FPS when limit is 120) while newer version works more correctly (almost always 60 FPS when limit is 60, and only 5-10 frames missing when FPS is set to 120 or more). Fixes https://github.com/rust-windowing/winit/issues/1610 [`CreateWaitableTimerExW`]: https://learn.microsoft.com/en-us/windows/win32/api/synchapi/nf-synchapi-createwaitabletimerexw [`MsgWaitForMultipleObjectsEx`]: https://learn.microsoft.com/en-us/windows/win32/api/winuser/nf-winuser-msgwaitformultipleobjectsex [`SetTimer`]: https://learn.microsoft.com/en-us/windows/win32/api/winuser/nf-winuser-settimer [review]: https://github.com/rust-windowing/winit/pull/3950#discussion_r1800184479
2024-10-13 12:56:24 +02:00
use windows_sys::Win32::Foundation::{
GetLastError, FALSE, HANDLE, HWND, LPARAM, LRESULT, POINT, RECT, WAIT_FAILED, WPARAM,
};
2022-03-07 22:58:12 +01:00
use windows_sys::Win32::Graphics::Gdi::{
Windows: Implement EventLoopExtPumpEvents and EventLoopExtRunOnDemand A surprising amount of work was required to enable these extensions on Windows. I had originally assumed that pump_events was going to be very similar to run except would use PeekMessageW instead of GetMessageW to avoid blocking the external loop but I found the Windows backend broke several assumptions I had. Overall I think these changes can hopefully be considered a quite a significant simplification (I think it's a net deletion of a fair amount of code) and I think it also helps bring it into slightly closer alignment with other backends too Key changes: - I have removed the `wait_thread` that was a fairly fiddly way of handling `ControlFlow::WaitUntil` timeouts in favor of using `SetTimer` which works with the same messages picked up by `GetMessage` and `PeekMessage`. - I have removed the ordering guarantees between `MainEventsCleared`, `RedrawRequested` and `RedrawEventsCleared` events due to the complexity in maintaining this artificial ordering, which is already not supported consistently across backends anyway (in particular this ordering already isn't compatible with how MacOS / iOS work). - `RedrawRequested` events are now directly dispatched via `WM_PAINT` messages - comparable to how `RedrawRequested` is dispatched via `drawRect` in the MacOS backend. - I have re-worked how `NewEvents`, `MainEventsCleared`, and `RedrawEventsCleared` get dispatched to be more in line with the MacOS backend and also more in line with how we have recently discussed defining them for all platforms. `NewEvents` is conceptually delivered when the event loop "wakes up" and `MainEventsCleared` gets dispatched when the event loop is about to ask the OS to wait for new events. This is a more portable model, and is already how these events work in the MacOS backend. `RedrawEventsCleared` are just delivered after `MainEventsCleared` but this event no longer has a useful meaning. Probably the most controversial thing here is that this "breaks" the ordering rules for redraw event handling, but since my changes interacted with how the order is maintained I was very reluctant to figure out how to continue maintaining something that we have recently been discussing changing: https://github.com/rust-windowing/winit/issues/2640. Additionally, since the MacOS backend already doesn't strictly maintain this order it's somewhat academic to see this as a breakage if Winit applications can't really rely on it already. This updates the documentation for `request_redraw()` to reflect that we no longer guarantee that `RedrawRequested` events must be dispatched after `MainEventsCleared`.
2023-06-18 11:42:57 +01:00
GetMonitorInfoW, MonitorFromRect, MonitorFromWindow, RedrawWindow, ScreenToClient,
ValidateRect, MONITORINFO, MONITOR_DEFAULTTONULL, RDW_INTERNALPAINT, SC_SCREENSAVE,
};
2023-06-03 15:02:37 -07:00
use windows_sys::Win32::System::Ole::RevokeDragDrop;
Improve waiting for messages on Windows Previous version used [`SetTimer`] with `GetMessageW` for waiting. The downside of UI timers like ones created by `SetTimer`, is that they may be late by up to 15-16 ms. To fix this behaviour, I added use of high resolution timers created by [`CreateWaitableTimerExW`] with the flag `CREATE_WAITABLE_TIMER_HIGH_RESOLUTION`. In my previous experience, waiting on such timers have precision of roundly 0.5 ms which is the best available on Windows at the moment. I use [`MsgWaitForMultipleObjectsEx`] to wait simultaneously for both timer and newly arriving events. Unfortunately, high resolution timers are available only since Windows 10 1803. However: 1. Win 10 is already getting to the end of support, like all previous versions, so it is OK to rely on APIs introduced in it; 2. I use `dwMilliseconds` parameter of `MsgWaitForMultipleObjectsEx` as a fallback. It should perform not worse compared to waiting for events from `SetTimer`. I also refactored code to remove event dispatching from function responsible for waiting for events. This provides more clear separations of concern and avoids unnecessary duplication of dispatching logic. After [review] from @rib, I also moved the waiting itself from `wait_for_messages` method to separate function, so it is clearly seen that `wait_for_messages` do 3 things: notify app that we about to wait, wait, notify that we have new events. I have tested behaviour using a egui app with Vulkan rendering with `VK_PRESENT_MODE_IMMEDIATE_KHR`, and older version consistently have twice less FPS than requested (e.g. 30 FPS when limit is 60 and 60 FPS when limit is 120) while newer version works more correctly (almost always 60 FPS when limit is 60, and only 5-10 frames missing when FPS is set to 120 or more). Fixes https://github.com/rust-windowing/winit/issues/1610 [`CreateWaitableTimerExW`]: https://learn.microsoft.com/en-us/windows/win32/api/synchapi/nf-synchapi-createwaitabletimerexw [`MsgWaitForMultipleObjectsEx`]: https://learn.microsoft.com/en-us/windows/win32/api/winuser/nf-winuser-msgwaitformultipleobjectsex [`SetTimer`]: https://learn.microsoft.com/en-us/windows/win32/api/winuser/nf-winuser-settimer [review]: https://github.com/rust-windowing/winit/pull/3950#discussion_r1800184479
2024-10-13 12:56:24 +02:00
use windows_sys::Win32::System::Threading::{
CreateWaitableTimerExW, GetCurrentThreadId, SetWaitableTimer,
CREATE_WAITABLE_TIMER_HIGH_RESOLUTION, INFINITE, TIMER_ALL_ACCESS,
};
2022-03-07 22:58:12 +01:00
use windows_sys::Win32::UI::Controls::{HOVER_DEFAULT, WM_MOUSELEAVE};
use windows_sys::Win32::UI::Input::Ime::{GCS_COMPSTR, GCS_RESULTSTR, ISC_SHOWUICOMPOSITIONWINDOW};
2022-03-07 22:58:12 +01:00
use windows_sys::Win32::UI::Input::KeyboardAndMouse::{
ReleaseCapture, SetCapture, TrackMouseEvent, TME_LEAVE, TRACKMOUSEEVENT,
2022-03-07 22:58:12 +01:00
};
use windows_sys::Win32::UI::Input::Pointer::{
2024-08-08 00:36:36 +02:00
POINTER_FLAG_DOWN, POINTER_FLAG_PRIMARY, POINTER_FLAG_UP, POINTER_FLAG_UPDATE,
};
2022-03-07 22:58:12 +01:00
use windows_sys::Win32::UI::Input::Touch::{
2024-08-08 00:36:36 +02:00
CloseTouchInputHandle, GetTouchInputInfo, TOUCHEVENTF_DOWN, TOUCHEVENTF_MOVE,
TOUCHEVENTF_PRIMARY, TOUCHEVENTF_UP, TOUCHINPUT,
2022-03-07 22:58:12 +01:00
};
use windows_sys::Win32::UI::Input::{RAWINPUT, RIM_TYPEKEYBOARD, RIM_TYPEMOUSE};
2022-03-07 22:58:12 +01:00
use windows_sys::Win32::UI::WindowsAndMessaging::{
CreateWindowExW, DefWindowProcW, DestroyWindow, DispatchMessageW, GetClientRect, GetCursorPos,
Improve waiting for messages on Windows Previous version used [`SetTimer`] with `GetMessageW` for waiting. The downside of UI timers like ones created by `SetTimer`, is that they may be late by up to 15-16 ms. To fix this behaviour, I added use of high resolution timers created by [`CreateWaitableTimerExW`] with the flag `CREATE_WAITABLE_TIMER_HIGH_RESOLUTION`. In my previous experience, waiting on such timers have precision of roundly 0.5 ms which is the best available on Windows at the moment. I use [`MsgWaitForMultipleObjectsEx`] to wait simultaneously for both timer and newly arriving events. Unfortunately, high resolution timers are available only since Windows 10 1803. However: 1. Win 10 is already getting to the end of support, like all previous versions, so it is OK to rely on APIs introduced in it; 2. I use `dwMilliseconds` parameter of `MsgWaitForMultipleObjectsEx` as a fallback. It should perform not worse compared to waiting for events from `SetTimer`. I also refactored code to remove event dispatching from function responsible for waiting for events. This provides more clear separations of concern and avoids unnecessary duplication of dispatching logic. After [review] from @rib, I also moved the waiting itself from `wait_for_messages` method to separate function, so it is clearly seen that `wait_for_messages` do 3 things: notify app that we about to wait, wait, notify that we have new events. I have tested behaviour using a egui app with Vulkan rendering with `VK_PRESENT_MODE_IMMEDIATE_KHR`, and older version consistently have twice less FPS than requested (e.g. 30 FPS when limit is 60 and 60 FPS when limit is 120) while newer version works more correctly (almost always 60 FPS when limit is 60, and only 5-10 frames missing when FPS is set to 120 or more). Fixes https://github.com/rust-windowing/winit/issues/1610 [`CreateWaitableTimerExW`]: https://learn.microsoft.com/en-us/windows/win32/api/synchapi/nf-synchapi-createwaitabletimerexw [`MsgWaitForMultipleObjectsEx`]: https://learn.microsoft.com/en-us/windows/win32/api/winuser/nf-winuser-msgwaitformultipleobjectsex [`SetTimer`]: https://learn.microsoft.com/en-us/windows/win32/api/winuser/nf-winuser-settimer [review]: https://github.com/rust-windowing/winit/pull/3950#discussion_r1800184479
2024-10-13 12:56:24 +02:00
GetMenu, LoadCursorW, MsgWaitForMultipleObjectsEx, PeekMessageW, PostMessageW,
RegisterClassExW, RegisterWindowMessageA, SetCursor, SetWindowPos, TranslateMessage,
CREATESTRUCTW, GWL_STYLE, GWL_USERDATA, HTCAPTION, HTCLIENT, MINMAXINFO, MNC_CLOSE, MSG,
MWMO_INPUTAVAILABLE, NCCALCSIZE_PARAMS, PM_REMOVE, PT_TOUCH, QS_ALLEVENTS, RI_MOUSE_HWHEEL,
RI_MOUSE_WHEEL, SC_MINIMIZE, SC_RESTORE, SIZE_MAXIMIZED, SWP_NOACTIVATE, SWP_NOMOVE,
SWP_NOSIZE, SWP_NOZORDER, WHEEL_DELTA, WINDOWPOS, WMSZ_BOTTOM, WMSZ_BOTTOMLEFT,
WMSZ_BOTTOMRIGHT, WMSZ_LEFT, WMSZ_RIGHT, WMSZ_TOP, WMSZ_TOPLEFT, WMSZ_TOPRIGHT,
WM_CAPTURECHANGED, WM_CLOSE, WM_CREATE, WM_DESTROY, WM_DPICHANGED, WM_ENTERSIZEMOVE,
WM_EXITSIZEMOVE, WM_GETMINMAXINFO, WM_IME_COMPOSITION, WM_IME_ENDCOMPOSITION,
WM_IME_SETCONTEXT, WM_IME_STARTCOMPOSITION, WM_INPUT, WM_KEYDOWN, WM_KEYUP, WM_KILLFOCUS,
WM_LBUTTONDOWN, WM_LBUTTONUP, WM_MBUTTONDOWN, WM_MBUTTONUP, WM_MENUCHAR, WM_MOUSEHWHEEL,
WM_MOUSEMOVE, WM_MOUSEWHEEL, WM_NCACTIVATE, WM_NCCALCSIZE, WM_NCCREATE, WM_NCDESTROY,
WM_NCLBUTTONDOWN, WM_PAINT, WM_POINTERDOWN, WM_POINTERUP, WM_POINTERUPDATE, WM_RBUTTONDOWN,
WM_RBUTTONUP, WM_SETCURSOR, WM_SETFOCUS, WM_SETTINGCHANGE, WM_SIZE, WM_SIZING, WM_SYSCOMMAND,
WM_SYSKEYDOWN, WM_SYSKEYUP, WM_TOUCH, WM_WINDOWPOSCHANGED, WM_WINDOWPOSCHANGING,
WM_XBUTTONDOWN, WM_XBUTTONUP, WNDCLASSEXW, WS_EX_LAYERED, WS_EX_NOACTIVATE, WS_EX_TOOLWINDOW,
WS_EX_TRANSPARENT, WS_OVERLAPPED, WS_POPUP, WS_VISIBLE,
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
};
use super::window::set_skip_taskbar;
use super::SelectedCursor;
use crate::application::ApplicationHandler;
use crate::dpi::{PhysicalPosition, PhysicalSize};
use crate::error::{EventLoopError, RequestError};
use crate::event::{
Event, FingerId, Force, Ime, RawKeyEvent, SurfaceSizeWriter, TouchPhase, WindowEvent,
};
use crate::event_loop::{
ActiveEventLoop as RootActiveEventLoop, ControlFlow, DeviceEvents,
EventLoopProxy as RootEventLoopProxy, EventLoopProxyProvider,
OwnedDisplayHandle as CoreOwnedDisplayHandle,
};
use crate::keyboard::ModifiersState;
use crate::monitor::MonitorHandle as RootMonitorHandle;
use crate::platform::pump_events::PumpStatus;
use crate::platform_impl::platform::dark_mode::try_theme;
use crate::platform_impl::platform::dpi::{become_dpi_aware, dpi_to_scale_factor};
use crate::platform_impl::platform::drop_handler::FileDropHandler;
use crate::platform_impl::platform::icon::WinCursor;
use crate::platform_impl::platform::ime::ImeContext;
use crate::platform_impl::platform::keyboard::KeyEventBuilder;
use crate::platform_impl::platform::keyboard_layout::LAYOUT_CACHE;
use crate::platform_impl::platform::monitor::{self, MonitorHandle};
use crate::platform_impl::platform::window::InitData;
use crate::platform_impl::platform::window_state::{
CursorFlags, ImeState, WindowFlags, WindowState,
};
use crate::platform_impl::platform::{raw_input, util, wrap_device_id, Fullscreen};
use crate::platform_impl::Window;
use crate::utils::Lazy;
use crate::window::{
CustomCursor as RootCustomCursor, CustomCursorSource, Theme, Window as CoreWindow,
WindowAttributes, WindowId,
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
};
Windows: Make `EventLoopWindowTarget` independent of UserEvent type (#3061) * make `EventLoopWindowTarget` independent of UserEvent type the `EventLoopWindowTarget` is needed for window creation. conceptually, only `EventLoop` and `EventLoopProxy` need to be parameterized, and all other parts of the backend should be agnostic about the user event type, parallel to how `Event<T>` is parameterized, but `WindowEvent` is not. this change removes the dependency on the type of user events from the `EventLoopWindowTarget` for the Windows backend, but keep a phantom data to keep the API intact. to achieve this, I moved the `Receiver` end of the mpsc channel from `ThreadMsgTargetData` into `EventLoop` itself, so the `UserEvent` is only passed between `EventLoop` and `EventLoopProxy`, all other part of the backend just use unit type as a placeholder for user events. it's similar to the macos backend where an erased `EventHandler` trait object is used so all component except `EventLoop` and `EventLoopProxy` need to be parameterized. however `EventLoop` of the Windows backend already use an `Box<dyn FnMut>` to wrap the user provided event handler callback, so no need for an dedicated trait object, I just modified the wrapper to replace the placeholder user event with real value pulled from the channel. I find this is the approach which need minimum change to be made to existing code. but it does the job and could serve as a starting point to future Windows backend re-works. * fix CI clippy failure. * make UserEventPlaceholder a new type instead of alias * invariance is maintained by top-level EventLoopWindowTarget<T> this field is transitional and her to keep API compatibility only. the correct variance and such is already ensured by the top-level `EventLoopWindowTarget`, just use `PhantomData<T>` here.
2024-01-04 23:47:07 +08:00
pub(crate) struct WindowData {
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
pub window_state: Arc<Mutex<WindowState>>,
pub event_loop_runner: Rc<EventLoopRunner>,
pub key_event_builder: KeyEventBuilder,
pub _file_drop_handler: Option<FileDropHandler>,
pub userdata_removed: Cell<bool>,
pub recurse_depth: Cell<u32>,
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
}
Windows: Make `EventLoopWindowTarget` independent of UserEvent type (#3061) * make `EventLoopWindowTarget` independent of UserEvent type the `EventLoopWindowTarget` is needed for window creation. conceptually, only `EventLoop` and `EventLoopProxy` need to be parameterized, and all other parts of the backend should be agnostic about the user event type, parallel to how `Event<T>` is parameterized, but `WindowEvent` is not. this change removes the dependency on the type of user events from the `EventLoopWindowTarget` for the Windows backend, but keep a phantom data to keep the API intact. to achieve this, I moved the `Receiver` end of the mpsc channel from `ThreadMsgTargetData` into `EventLoop` itself, so the `UserEvent` is only passed between `EventLoop` and `EventLoopProxy`, all other part of the backend just use unit type as a placeholder for user events. it's similar to the macos backend where an erased `EventHandler` trait object is used so all component except `EventLoop` and `EventLoopProxy` need to be parameterized. however `EventLoop` of the Windows backend already use an `Box<dyn FnMut>` to wrap the user provided event handler callback, so no need for an dedicated trait object, I just modified the wrapper to replace the placeholder user event with real value pulled from the channel. I find this is the approach which need minimum change to be made to existing code. but it does the job and could serve as a starting point to future Windows backend re-works. * fix CI clippy failure. * make UserEventPlaceholder a new type instead of alias * invariance is maintained by top-level EventLoopWindowTarget<T> this field is transitional and her to keep API compatibility only. the correct variance and such is already ensured by the top-level `EventLoopWindowTarget`, just use `PhantomData<T>` here.
2024-01-04 23:47:07 +08:00
impl WindowData {
fn send_event(&self, event: Event) {
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
self.event_loop_runner.send_event(event);
}
2022-08-31 18:32:19 +02:00
fn window_state_lock(&self) -> MutexGuard<'_, WindowState> {
self.window_state.lock().unwrap()
}
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
}
Windows: Make `EventLoopWindowTarget` independent of UserEvent type (#3061) * make `EventLoopWindowTarget` independent of UserEvent type the `EventLoopWindowTarget` is needed for window creation. conceptually, only `EventLoop` and `EventLoopProxy` need to be parameterized, and all other parts of the backend should be agnostic about the user event type, parallel to how `Event<T>` is parameterized, but `WindowEvent` is not. this change removes the dependency on the type of user events from the `EventLoopWindowTarget` for the Windows backend, but keep a phantom data to keep the API intact. to achieve this, I moved the `Receiver` end of the mpsc channel from `ThreadMsgTargetData` into `EventLoop` itself, so the `UserEvent` is only passed between `EventLoop` and `EventLoopProxy`, all other part of the backend just use unit type as a placeholder for user events. it's similar to the macos backend where an erased `EventHandler` trait object is used so all component except `EventLoop` and `EventLoopProxy` need to be parameterized. however `EventLoop` of the Windows backend already use an `Box<dyn FnMut>` to wrap the user provided event handler callback, so no need for an dedicated trait object, I just modified the wrapper to replace the placeholder user event with real value pulled from the channel. I find this is the approach which need minimum change to be made to existing code. but it does the job and could serve as a starting point to future Windows backend re-works. * fix CI clippy failure. * make UserEventPlaceholder a new type instead of alias * invariance is maintained by top-level EventLoopWindowTarget<T> this field is transitional and her to keep API compatibility only. the correct variance and such is already ensured by the top-level `EventLoopWindowTarget`, just use `PhantomData<T>` here.
2024-01-04 23:47:07 +08:00
struct ThreadMsgTargetData {
event_loop_runner: Rc<EventLoopRunner>,
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
}
Windows: Make `EventLoopWindowTarget` independent of UserEvent type (#3061) * make `EventLoopWindowTarget` independent of UserEvent type the `EventLoopWindowTarget` is needed for window creation. conceptually, only `EventLoop` and `EventLoopProxy` need to be parameterized, and all other parts of the backend should be agnostic about the user event type, parallel to how `Event<T>` is parameterized, but `WindowEvent` is not. this change removes the dependency on the type of user events from the `EventLoopWindowTarget` for the Windows backend, but keep a phantom data to keep the API intact. to achieve this, I moved the `Receiver` end of the mpsc channel from `ThreadMsgTargetData` into `EventLoop` itself, so the `UserEvent` is only passed between `EventLoop` and `EventLoopProxy`, all other part of the backend just use unit type as a placeholder for user events. it's similar to the macos backend where an erased `EventHandler` trait object is used so all component except `EventLoop` and `EventLoopProxy` need to be parameterized. however `EventLoop` of the Windows backend already use an `Box<dyn FnMut>` to wrap the user provided event handler callback, so no need for an dedicated trait object, I just modified the wrapper to replace the placeholder user event with real value pulled from the channel. I find this is the approach which need minimum change to be made to existing code. but it does the job and could serve as a starting point to future Windows backend re-works. * fix CI clippy failure. * make UserEventPlaceholder a new type instead of alias * invariance is maintained by top-level EventLoopWindowTarget<T> this field is transitional and her to keep API compatibility only. the correct variance and such is already ensured by the top-level `EventLoopWindowTarget`, just use `PhantomData<T>` here.
2024-01-04 23:47:07 +08:00
impl ThreadMsgTargetData {
fn send_event(&self, event: Event) {
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
self.event_loop_runner.send_event(event);
}
}
/// The result of a subclass procedure (the message handling callback)
#[derive(Clone, Copy)]
pub(crate) enum ProcResult {
DefWindowProc(WPARAM),
Value(isize),
}
pub struct EventLoop {
window_target: ActiveEventLoop,
msg_hook: Option<Box<dyn FnMut(*const c_void) -> bool + 'static>>,
Improve waiting for messages on Windows Previous version used [`SetTimer`] with `GetMessageW` for waiting. The downside of UI timers like ones created by `SetTimer`, is that they may be late by up to 15-16 ms. To fix this behaviour, I added use of high resolution timers created by [`CreateWaitableTimerExW`] with the flag `CREATE_WAITABLE_TIMER_HIGH_RESOLUTION`. In my previous experience, waiting on such timers have precision of roundly 0.5 ms which is the best available on Windows at the moment. I use [`MsgWaitForMultipleObjectsEx`] to wait simultaneously for both timer and newly arriving events. Unfortunately, high resolution timers are available only since Windows 10 1803. However: 1. Win 10 is already getting to the end of support, like all previous versions, so it is OK to rely on APIs introduced in it; 2. I use `dwMilliseconds` parameter of `MsgWaitForMultipleObjectsEx` as a fallback. It should perform not worse compared to waiting for events from `SetTimer`. I also refactored code to remove event dispatching from function responsible for waiting for events. This provides more clear separations of concern and avoids unnecessary duplication of dispatching logic. After [review] from @rib, I also moved the waiting itself from `wait_for_messages` method to separate function, so it is clearly seen that `wait_for_messages` do 3 things: notify app that we about to wait, wait, notify that we have new events. I have tested behaviour using a egui app with Vulkan rendering with `VK_PRESENT_MODE_IMMEDIATE_KHR`, and older version consistently have twice less FPS than requested (e.g. 30 FPS when limit is 60 and 60 FPS when limit is 120) while newer version works more correctly (almost always 60 FPS when limit is 60, and only 5-10 frames missing when FPS is set to 120 or more). Fixes https://github.com/rust-windowing/winit/issues/1610 [`CreateWaitableTimerExW`]: https://learn.microsoft.com/en-us/windows/win32/api/synchapi/nf-synchapi-createwaitabletimerexw [`MsgWaitForMultipleObjectsEx`]: https://learn.microsoft.com/en-us/windows/win32/api/winuser/nf-winuser-msgwaitformultipleobjectsex [`SetTimer`]: https://learn.microsoft.com/en-us/windows/win32/api/winuser/nf-winuser-settimer [review]: https://github.com/rust-windowing/winit/pull/3950#discussion_r1800184479
2024-10-13 12:56:24 +02:00
// It is a timer used on timed waits.
// It is created lazily in case if we have `ControlFlow::WaitUntil`.
// Keep it as a field to avoid recreating it on every `ControlFlow::WaitUntil`.
high_resolution_timer: Option<OwnedHandle>,
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
}
pub(crate) struct PlatformSpecificEventLoopAttributes {
pub(crate) any_thread: bool,
pub(crate) dpi_aware: bool,
pub(crate) msg_hook: Option<Box<dyn FnMut(*const c_void) -> bool + 'static>>,
}
impl Default for PlatformSpecificEventLoopAttributes {
fn default() -> Self {
Self { any_thread: false, dpi_aware: true, msg_hook: None }
}
}
impl PartialEq for PlatformSpecificEventLoopAttributes {
fn eq(&self, other: &Self) -> bool {
self.any_thread.eq(&other.any_thread)
&& self.dpi_aware.eq(&other.dpi_aware)
&& match (&self.msg_hook, &other.msg_hook) {
(Some(this), Some(other)) => std::ptr::eq(&this, &other),
(None, None) => true,
_ => false,
}
}
}
impl Eq for PlatformSpecificEventLoopAttributes {}
impl std::hash::Hash for PlatformSpecificEventLoopAttributes {
fn hash<H: std::hash::Hasher>(&self, state: &mut H) {
self.any_thread.hash(state);
self.dpi_aware.hash(state);
std::ptr::hash(&self.msg_hook, state);
}
}
pub struct ActiveEventLoop {
2022-03-07 22:58:12 +01:00
thread_id: u32,
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
thread_msg_target: HWND,
pub(crate) runner_shared: Rc<EventLoopRunner>,
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
}
impl EventLoop {
pub(crate) fn new(
attributes: &mut PlatformSpecificEventLoopAttributes,
) -> Result<Self, EventLoopError> {
2022-03-07 22:58:12 +01:00
let thread_id = unsafe { GetCurrentThreadId() };
if !attributes.any_thread && thread_id != main_thread_id() {
panic!(
"Initializing the event loop outside of the main thread is a significant \
cross-platform compatibility hazard. If you absolutely need to create an \
EventLoop on a different thread, you can use the \
`EventLoopBuilderExtWindows::with_any_thread` function."
);
}
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
if attributes.dpi_aware {
become_dpi_aware();
}
Windows: Make `EventLoopWindowTarget` independent of UserEvent type (#3061) * make `EventLoopWindowTarget` independent of UserEvent type the `EventLoopWindowTarget` is needed for window creation. conceptually, only `EventLoop` and `EventLoopProxy` need to be parameterized, and all other parts of the backend should be agnostic about the user event type, parallel to how `Event<T>` is parameterized, but `WindowEvent` is not. this change removes the dependency on the type of user events from the `EventLoopWindowTarget` for the Windows backend, but keep a phantom data to keep the API intact. to achieve this, I moved the `Receiver` end of the mpsc channel from `ThreadMsgTargetData` into `EventLoop` itself, so the `UserEvent` is only passed between `EventLoop` and `EventLoopProxy`, all other part of the backend just use unit type as a placeholder for user events. it's similar to the macos backend where an erased `EventHandler` trait object is used so all component except `EventLoop` and `EventLoopProxy` need to be parameterized. however `EventLoop` of the Windows backend already use an `Box<dyn FnMut>` to wrap the user provided event handler callback, so no need for an dedicated trait object, I just modified the wrapper to replace the placeholder user event with real value pulled from the channel. I find this is the approach which need minimum change to be made to existing code. but it does the job and could serve as a starting point to future Windows backend re-works. * fix CI clippy failure. * make UserEventPlaceholder a new type instead of alias * invariance is maintained by top-level EventLoopWindowTarget<T> this field is transitional and her to keep API compatibility only. the correct variance and such is already ensured by the top-level `EventLoopWindowTarget`, just use `PhantomData<T>` here.
2024-01-04 23:47:07 +08:00
let thread_msg_target = create_event_target_window();
Windows: Implement EventLoopExtPumpEvents and EventLoopExtRunOnDemand A surprising amount of work was required to enable these extensions on Windows. I had originally assumed that pump_events was going to be very similar to run except would use PeekMessageW instead of GetMessageW to avoid blocking the external loop but I found the Windows backend broke several assumptions I had. Overall I think these changes can hopefully be considered a quite a significant simplification (I think it's a net deletion of a fair amount of code) and I think it also helps bring it into slightly closer alignment with other backends too Key changes: - I have removed the `wait_thread` that was a fairly fiddly way of handling `ControlFlow::WaitUntil` timeouts in favor of using `SetTimer` which works with the same messages picked up by `GetMessage` and `PeekMessage`. - I have removed the ordering guarantees between `MainEventsCleared`, `RedrawRequested` and `RedrawEventsCleared` events due to the complexity in maintaining this artificial ordering, which is already not supported consistently across backends anyway (in particular this ordering already isn't compatible with how MacOS / iOS work). - `RedrawRequested` events are now directly dispatched via `WM_PAINT` messages - comparable to how `RedrawRequested` is dispatched via `drawRect` in the MacOS backend. - I have re-worked how `NewEvents`, `MainEventsCleared`, and `RedrawEventsCleared` get dispatched to be more in line with the MacOS backend and also more in line with how we have recently discussed defining them for all platforms. `NewEvents` is conceptually delivered when the event loop "wakes up" and `MainEventsCleared` gets dispatched when the event loop is about to ask the OS to wait for new events. This is a more portable model, and is already how these events work in the MacOS backend. `RedrawEventsCleared` are just delivered after `MainEventsCleared` but this event no longer has a useful meaning. Probably the most controversial thing here is that this "breaks" the ordering rules for redraw event handling, but since my changes interacted with how the order is maintained I was very reluctant to figure out how to continue maintaining something that we have recently been discussing changing: https://github.com/rust-windowing/winit/issues/2640. Additionally, since the MacOS backend already doesn't strictly maintain this order it's somewhat academic to see this as a breakage if Winit applications can't really rely on it already. This updates the documentation for `request_redraw()` to reflect that we no longer guarantee that `RedrawRequested` events must be dispatched after `MainEventsCleared`.
2023-06-18 11:42:57 +01:00
let runner_shared = Rc::new(EventLoopRunner::new(thread_msg_target));
Windows: Make `EventLoopWindowTarget` independent of UserEvent type (#3061) * make `EventLoopWindowTarget` independent of UserEvent type the `EventLoopWindowTarget` is needed for window creation. conceptually, only `EventLoop` and `EventLoopProxy` need to be parameterized, and all other parts of the backend should be agnostic about the user event type, parallel to how `Event<T>` is parameterized, but `WindowEvent` is not. this change removes the dependency on the type of user events from the `EventLoopWindowTarget` for the Windows backend, but keep a phantom data to keep the API intact. to achieve this, I moved the `Receiver` end of the mpsc channel from `ThreadMsgTargetData` into `EventLoop` itself, so the `UserEvent` is only passed between `EventLoop` and `EventLoopProxy`, all other part of the backend just use unit type as a placeholder for user events. it's similar to the macos backend where an erased `EventHandler` trait object is used so all component except `EventLoop` and `EventLoopProxy` need to be parameterized. however `EventLoop` of the Windows backend already use an `Box<dyn FnMut>` to wrap the user provided event handler callback, so no need for an dedicated trait object, I just modified the wrapper to replace the placeholder user event with real value pulled from the channel. I find this is the approach which need minimum change to be made to existing code. but it does the job and could serve as a starting point to future Windows backend re-works. * fix CI clippy failure. * make UserEventPlaceholder a new type instead of alias * invariance is maintained by top-level EventLoopWindowTarget<T> this field is transitional and her to keep API compatibility only. the correct variance and such is already ensured by the top-level `EventLoopWindowTarget`, just use `PhantomData<T>` here.
2024-01-04 23:47:07 +08:00
insert_event_target_window_data(thread_msg_target, runner_shared.clone());
raw_input::register_all_mice_and_keyboards_for_raw_input(
thread_msg_target,
Default::default(),
);
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
Ok(EventLoop {
window_target: ActiveEventLoop { thread_id, thread_msg_target, runner_shared },
msg_hook: attributes.msg_hook.take(),
Improve waiting for messages on Windows Previous version used [`SetTimer`] with `GetMessageW` for waiting. The downside of UI timers like ones created by `SetTimer`, is that they may be late by up to 15-16 ms. To fix this behaviour, I added use of high resolution timers created by [`CreateWaitableTimerExW`] with the flag `CREATE_WAITABLE_TIMER_HIGH_RESOLUTION`. In my previous experience, waiting on such timers have precision of roundly 0.5 ms which is the best available on Windows at the moment. I use [`MsgWaitForMultipleObjectsEx`] to wait simultaneously for both timer and newly arriving events. Unfortunately, high resolution timers are available only since Windows 10 1803. However: 1. Win 10 is already getting to the end of support, like all previous versions, so it is OK to rely on APIs introduced in it; 2. I use `dwMilliseconds` parameter of `MsgWaitForMultipleObjectsEx` as a fallback. It should perform not worse compared to waiting for events from `SetTimer`. I also refactored code to remove event dispatching from function responsible for waiting for events. This provides more clear separations of concern and avoids unnecessary duplication of dispatching logic. After [review] from @rib, I also moved the waiting itself from `wait_for_messages` method to separate function, so it is clearly seen that `wait_for_messages` do 3 things: notify app that we about to wait, wait, notify that we have new events. I have tested behaviour using a egui app with Vulkan rendering with `VK_PRESENT_MODE_IMMEDIATE_KHR`, and older version consistently have twice less FPS than requested (e.g. 30 FPS when limit is 60 and 60 FPS when limit is 120) while newer version works more correctly (almost always 60 FPS when limit is 60, and only 5-10 frames missing when FPS is set to 120 or more). Fixes https://github.com/rust-windowing/winit/issues/1610 [`CreateWaitableTimerExW`]: https://learn.microsoft.com/en-us/windows/win32/api/synchapi/nf-synchapi-createwaitabletimerexw [`MsgWaitForMultipleObjectsEx`]: https://learn.microsoft.com/en-us/windows/win32/api/winuser/nf-winuser-msgwaitformultipleobjectsex [`SetTimer`]: https://learn.microsoft.com/en-us/windows/win32/api/winuser/nf-winuser-settimer [review]: https://github.com/rust-windowing/winit/pull/3950#discussion_r1800184479
2024-10-13 12:56:24 +02:00
high_resolution_timer: None,
})
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
}
pub fn window_target(&self) -> &dyn RootActiveEventLoop {
&self.window_target
}
pub fn run_app<A: ApplicationHandler>(mut self, app: A) -> Result<(), EventLoopError> {
self.run_app_on_demand(app)
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
}
pub fn run_app_on_demand<A: ApplicationHandler>(
&mut self,
mut app: A,
) -> Result<(), EventLoopError> {
self.window_target.clear_exit();
Windows: Implement EventLoopExtPumpEvents and EventLoopExtRunOnDemand A surprising amount of work was required to enable these extensions on Windows. I had originally assumed that pump_events was going to be very similar to run except would use PeekMessageW instead of GetMessageW to avoid blocking the external loop but I found the Windows backend broke several assumptions I had. Overall I think these changes can hopefully be considered a quite a significant simplification (I think it's a net deletion of a fair amount of code) and I think it also helps bring it into slightly closer alignment with other backends too Key changes: - I have removed the `wait_thread` that was a fairly fiddly way of handling `ControlFlow::WaitUntil` timeouts in favor of using `SetTimer` which works with the same messages picked up by `GetMessage` and `PeekMessage`. - I have removed the ordering guarantees between `MainEventsCleared`, `RedrawRequested` and `RedrawEventsCleared` events due to the complexity in maintaining this artificial ordering, which is already not supported consistently across backends anyway (in particular this ordering already isn't compatible with how MacOS / iOS work). - `RedrawRequested` events are now directly dispatched via `WM_PAINT` messages - comparable to how `RedrawRequested` is dispatched via `drawRect` in the MacOS backend. - I have re-worked how `NewEvents`, `MainEventsCleared`, and `RedrawEventsCleared` get dispatched to be more in line with the MacOS backend and also more in line with how we have recently discussed defining them for all platforms. `NewEvents` is conceptually delivered when the event loop "wakes up" and `MainEventsCleared` gets dispatched when the event loop is about to ask the OS to wait for new events. This is a more portable model, and is already how these events work in the MacOS backend. `RedrawEventsCleared` are just delivered after `MainEventsCleared` but this event no longer has a useful meaning. Probably the most controversial thing here is that this "breaks" the ordering rules for redraw event handling, but since my changes interacted with how the order is maintained I was very reluctant to figure out how to continue maintaining something that we have recently been discussing changing: https://github.com/rust-windowing/winit/issues/2640. Additionally, since the MacOS backend already doesn't strictly maintain this order it's somewhat academic to see this as a breakage if Winit applications can't really rely on it already. This updates the documentation for `request_redraw()` to reflect that we no longer guarantee that `RedrawRequested` events must be dispatched after `MainEventsCleared`.
2023-06-18 11:42:57 +01:00
{
let runner = &self.window_target.runner_shared;
Windows: Implement EventLoopExtPumpEvents and EventLoopExtRunOnDemand A surprising amount of work was required to enable these extensions on Windows. I had originally assumed that pump_events was going to be very similar to run except would use PeekMessageW instead of GetMessageW to avoid blocking the external loop but I found the Windows backend broke several assumptions I had. Overall I think these changes can hopefully be considered a quite a significant simplification (I think it's a net deletion of a fair amount of code) and I think it also helps bring it into slightly closer alignment with other backends too Key changes: - I have removed the `wait_thread` that was a fairly fiddly way of handling `ControlFlow::WaitUntil` timeouts in favor of using `SetTimer` which works with the same messages picked up by `GetMessage` and `PeekMessage`. - I have removed the ordering guarantees between `MainEventsCleared`, `RedrawRequested` and `RedrawEventsCleared` events due to the complexity in maintaining this artificial ordering, which is already not supported consistently across backends anyway (in particular this ordering already isn't compatible with how MacOS / iOS work). - `RedrawRequested` events are now directly dispatched via `WM_PAINT` messages - comparable to how `RedrawRequested` is dispatched via `drawRect` in the MacOS backend. - I have re-worked how `NewEvents`, `MainEventsCleared`, and `RedrawEventsCleared` get dispatched to be more in line with the MacOS backend and also more in line with how we have recently discussed defining them for all platforms. `NewEvents` is conceptually delivered when the event loop "wakes up" and `MainEventsCleared` gets dispatched when the event loop is about to ask the OS to wait for new events. This is a more portable model, and is already how these events work in the MacOS backend. `RedrawEventsCleared` are just delivered after `MainEventsCleared` but this event no longer has a useful meaning. Probably the most controversial thing here is that this "breaks" the ordering rules for redraw event handling, but since my changes interacted with how the order is maintained I was very reluctant to figure out how to continue maintaining something that we have recently been discussing changing: https://github.com/rust-windowing/winit/issues/2640. Additionally, since the MacOS backend already doesn't strictly maintain this order it's somewhat academic to see this as a breakage if Winit applications can't really rely on it already. This updates the documentation for `request_redraw()` to reflect that we no longer guarantee that `RedrawRequested` events must be dispatched after `MainEventsCleared`.
2023-06-18 11:42:57 +01:00
let event_loop_windows_ref = &self.window_target;
// # Safety
// We make sure to call runner.clear_event_handler() before
// returning
unsafe {
runner.set_event_handler(move |event| match event {
Event::NewEvents(cause) => app.new_events(event_loop_windows_ref, cause),
Event::WindowEvent { window_id, event } => {
app.window_event(event_loop_windows_ref, window_id, event)
},
Event::DeviceEvent { device_id, event } => {
app.device_event(event_loop_windows_ref, device_id, event)
},
Event::UserWakeUp => app.proxy_wake_up(event_loop_windows_ref),
Event::Suspended => app.suspended(event_loop_windows_ref),
Event::Resumed => app.resumed(event_loop_windows_ref),
Event::CreateSurfaces => app.can_create_surfaces(event_loop_windows_ref),
Event::AboutToWait => app.about_to_wait(event_loop_windows_ref),
Event::LoopExiting => app.exiting(event_loop_windows_ref),
Event::MemoryWarning => app.memory_warning(event_loop_windows_ref),
Windows: Make `EventLoopWindowTarget` independent of UserEvent type (#3061) * make `EventLoopWindowTarget` independent of UserEvent type the `EventLoopWindowTarget` is needed for window creation. conceptually, only `EventLoop` and `EventLoopProxy` need to be parameterized, and all other parts of the backend should be agnostic about the user event type, parallel to how `Event<T>` is parameterized, but `WindowEvent` is not. this change removes the dependency on the type of user events from the `EventLoopWindowTarget` for the Windows backend, but keep a phantom data to keep the API intact. to achieve this, I moved the `Receiver` end of the mpsc channel from `ThreadMsgTargetData` into `EventLoop` itself, so the `UserEvent` is only passed between `EventLoop` and `EventLoopProxy`, all other part of the backend just use unit type as a placeholder for user events. it's similar to the macos backend where an erased `EventHandler` trait object is used so all component except `EventLoop` and `EventLoopProxy` need to be parameterized. however `EventLoop` of the Windows backend already use an `Box<dyn FnMut>` to wrap the user provided event handler callback, so no need for an dedicated trait object, I just modified the wrapper to replace the placeholder user event with real value pulled from the channel. I find this is the approach which need minimum change to be made to existing code. but it does the job and could serve as a starting point to future Windows backend re-works. * fix CI clippy failure. * make UserEventPlaceholder a new type instead of alias * invariance is maintained by top-level EventLoopWindowTarget<T> this field is transitional and her to keep API compatibility only. the correct variance and such is already ensured by the top-level `EventLoopWindowTarget`, just use `PhantomData<T>` here.
2024-01-04 23:47:07 +08:00
});
Windows: Implement EventLoopExtPumpEvents and EventLoopExtRunOnDemand A surprising amount of work was required to enable these extensions on Windows. I had originally assumed that pump_events was going to be very similar to run except would use PeekMessageW instead of GetMessageW to avoid blocking the external loop but I found the Windows backend broke several assumptions I had. Overall I think these changes can hopefully be considered a quite a significant simplification (I think it's a net deletion of a fair amount of code) and I think it also helps bring it into slightly closer alignment with other backends too Key changes: - I have removed the `wait_thread` that was a fairly fiddly way of handling `ControlFlow::WaitUntil` timeouts in favor of using `SetTimer` which works with the same messages picked up by `GetMessage` and `PeekMessage`. - I have removed the ordering guarantees between `MainEventsCleared`, `RedrawRequested` and `RedrawEventsCleared` events due to the complexity in maintaining this artificial ordering, which is already not supported consistently across backends anyway (in particular this ordering already isn't compatible with how MacOS / iOS work). - `RedrawRequested` events are now directly dispatched via `WM_PAINT` messages - comparable to how `RedrawRequested` is dispatched via `drawRect` in the MacOS backend. - I have re-worked how `NewEvents`, `MainEventsCleared`, and `RedrawEventsCleared` get dispatched to be more in line with the MacOS backend and also more in line with how we have recently discussed defining them for all platforms. `NewEvents` is conceptually delivered when the event loop "wakes up" and `MainEventsCleared` gets dispatched when the event loop is about to ask the OS to wait for new events. This is a more portable model, and is already how these events work in the MacOS backend. `RedrawEventsCleared` are just delivered after `MainEventsCleared` but this event no longer has a useful meaning. Probably the most controversial thing here is that this "breaks" the ordering rules for redraw event handling, but since my changes interacted with how the order is maintained I was very reluctant to figure out how to continue maintaining something that we have recently been discussing changing: https://github.com/rust-windowing/winit/issues/2640. Additionally, since the MacOS backend already doesn't strictly maintain this order it's somewhat academic to see this as a breakage if Winit applications can't really rely on it already. This updates the documentation for `request_redraw()` to reflect that we no longer guarantee that `RedrawRequested` events must be dispatched after `MainEventsCleared`.
2023-06-18 11:42:57 +01:00
}
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
}
Windows: Implement EventLoopExtPumpEvents and EventLoopExtRunOnDemand A surprising amount of work was required to enable these extensions on Windows. I had originally assumed that pump_events was going to be very similar to run except would use PeekMessageW instead of GetMessageW to avoid blocking the external loop but I found the Windows backend broke several assumptions I had. Overall I think these changes can hopefully be considered a quite a significant simplification (I think it's a net deletion of a fair amount of code) and I think it also helps bring it into slightly closer alignment with other backends too Key changes: - I have removed the `wait_thread` that was a fairly fiddly way of handling `ControlFlow::WaitUntil` timeouts in favor of using `SetTimer` which works with the same messages picked up by `GetMessage` and `PeekMessage`. - I have removed the ordering guarantees between `MainEventsCleared`, `RedrawRequested` and `RedrawEventsCleared` events due to the complexity in maintaining this artificial ordering, which is already not supported consistently across backends anyway (in particular this ordering already isn't compatible with how MacOS / iOS work). - `RedrawRequested` events are now directly dispatched via `WM_PAINT` messages - comparable to how `RedrawRequested` is dispatched via `drawRect` in the MacOS backend. - I have re-worked how `NewEvents`, `MainEventsCleared`, and `RedrawEventsCleared` get dispatched to be more in line with the MacOS backend and also more in line with how we have recently discussed defining them for all platforms. `NewEvents` is conceptually delivered when the event loop "wakes up" and `MainEventsCleared` gets dispatched when the event loop is about to ask the OS to wait for new events. This is a more portable model, and is already how these events work in the MacOS backend. `RedrawEventsCleared` are just delivered after `MainEventsCleared` but this event no longer has a useful meaning. Probably the most controversial thing here is that this "breaks" the ordering rules for redraw event handling, but since my changes interacted with how the order is maintained I was very reluctant to figure out how to continue maintaining something that we have recently been discussing changing: https://github.com/rust-windowing/winit/issues/2640. Additionally, since the MacOS backend already doesn't strictly maintain this order it's somewhat academic to see this as a breakage if Winit applications can't really rely on it already. This updates the documentation for `request_redraw()` to reflect that we no longer guarantee that `RedrawRequested` events must be dispatched after `MainEventsCleared`.
2023-06-18 11:42:57 +01:00
let exit_code = loop {
Improve waiting for messages on Windows Previous version used [`SetTimer`] with `GetMessageW` for waiting. The downside of UI timers like ones created by `SetTimer`, is that they may be late by up to 15-16 ms. To fix this behaviour, I added use of high resolution timers created by [`CreateWaitableTimerExW`] with the flag `CREATE_WAITABLE_TIMER_HIGH_RESOLUTION`. In my previous experience, waiting on such timers have precision of roundly 0.5 ms which is the best available on Windows at the moment. I use [`MsgWaitForMultipleObjectsEx`] to wait simultaneously for both timer and newly arriving events. Unfortunately, high resolution timers are available only since Windows 10 1803. However: 1. Win 10 is already getting to the end of support, like all previous versions, so it is OK to rely on APIs introduced in it; 2. I use `dwMilliseconds` parameter of `MsgWaitForMultipleObjectsEx` as a fallback. It should perform not worse compared to waiting for events from `SetTimer`. I also refactored code to remove event dispatching from function responsible for waiting for events. This provides more clear separations of concern and avoids unnecessary duplication of dispatching logic. After [review] from @rib, I also moved the waiting itself from `wait_for_messages` method to separate function, so it is clearly seen that `wait_for_messages` do 3 things: notify app that we about to wait, wait, notify that we have new events. I have tested behaviour using a egui app with Vulkan rendering with `VK_PRESENT_MODE_IMMEDIATE_KHR`, and older version consistently have twice less FPS than requested (e.g. 30 FPS when limit is 60 and 60 FPS when limit is 120) while newer version works more correctly (almost always 60 FPS when limit is 60, and only 5-10 frames missing when FPS is set to 120 or more). Fixes https://github.com/rust-windowing/winit/issues/1610 [`CreateWaitableTimerExW`]: https://learn.microsoft.com/en-us/windows/win32/api/synchapi/nf-synchapi-createwaitabletimerexw [`MsgWaitForMultipleObjectsEx`]: https://learn.microsoft.com/en-us/windows/win32/api/winuser/nf-winuser-msgwaitformultipleobjectsex [`SetTimer`]: https://learn.microsoft.com/en-us/windows/win32/api/winuser/nf-winuser-settimer [review]: https://github.com/rust-windowing/winit/pull/3950#discussion_r1800184479
2024-10-13 12:56:24 +02:00
self.wait_for_messages(None);
// wait_for_messages calls user application before and after waiting
// so it may have decided to exit.
if let Some(code) = self.exit_code() {
Windows: Implement EventLoopExtPumpEvents and EventLoopExtRunOnDemand A surprising amount of work was required to enable these extensions on Windows. I had originally assumed that pump_events was going to be very similar to run except would use PeekMessageW instead of GetMessageW to avoid blocking the external loop but I found the Windows backend broke several assumptions I had. Overall I think these changes can hopefully be considered a quite a significant simplification (I think it's a net deletion of a fair amount of code) and I think it also helps bring it into slightly closer alignment with other backends too Key changes: - I have removed the `wait_thread` that was a fairly fiddly way of handling `ControlFlow::WaitUntil` timeouts in favor of using `SetTimer` which works with the same messages picked up by `GetMessage` and `PeekMessage`. - I have removed the ordering guarantees between `MainEventsCleared`, `RedrawRequested` and `RedrawEventsCleared` events due to the complexity in maintaining this artificial ordering, which is already not supported consistently across backends anyway (in particular this ordering already isn't compatible with how MacOS / iOS work). - `RedrawRequested` events are now directly dispatched via `WM_PAINT` messages - comparable to how `RedrawRequested` is dispatched via `drawRect` in the MacOS backend. - I have re-worked how `NewEvents`, `MainEventsCleared`, and `RedrawEventsCleared` get dispatched to be more in line with the MacOS backend and also more in line with how we have recently discussed defining them for all platforms. `NewEvents` is conceptually delivered when the event loop "wakes up" and `MainEventsCleared` gets dispatched when the event loop is about to ask the OS to wait for new events. This is a more portable model, and is already how these events work in the MacOS backend. `RedrawEventsCleared` are just delivered after `MainEventsCleared` but this event no longer has a useful meaning. Probably the most controversial thing here is that this "breaks" the ordering rules for redraw event handling, but since my changes interacted with how the order is maintained I was very reluctant to figure out how to continue maintaining something that we have recently been discussing changing: https://github.com/rust-windowing/winit/issues/2640. Additionally, since the MacOS backend already doesn't strictly maintain this order it's somewhat academic to see this as a breakage if Winit applications can't really rely on it already. This updates the documentation for `request_redraw()` to reflect that we no longer guarantee that `RedrawRequested` events must be dispatched after `MainEventsCleared`.
2023-06-18 11:42:57 +01:00
break code;
}
self.dispatch_peeked_messages();
if let Some(code) = self.exit_code() {
Windows: Implement EventLoopExtPumpEvents and EventLoopExtRunOnDemand A surprising amount of work was required to enable these extensions on Windows. I had originally assumed that pump_events was going to be very similar to run except would use PeekMessageW instead of GetMessageW to avoid blocking the external loop but I found the Windows backend broke several assumptions I had. Overall I think these changes can hopefully be considered a quite a significant simplification (I think it's a net deletion of a fair amount of code) and I think it also helps bring it into slightly closer alignment with other backends too Key changes: - I have removed the `wait_thread` that was a fairly fiddly way of handling `ControlFlow::WaitUntil` timeouts in favor of using `SetTimer` which works with the same messages picked up by `GetMessage` and `PeekMessage`. - I have removed the ordering guarantees between `MainEventsCleared`, `RedrawRequested` and `RedrawEventsCleared` events due to the complexity in maintaining this artificial ordering, which is already not supported consistently across backends anyway (in particular this ordering already isn't compatible with how MacOS / iOS work). - `RedrawRequested` events are now directly dispatched via `WM_PAINT` messages - comparable to how `RedrawRequested` is dispatched via `drawRect` in the MacOS backend. - I have re-worked how `NewEvents`, `MainEventsCleared`, and `RedrawEventsCleared` get dispatched to be more in line with the MacOS backend and also more in line with how we have recently discussed defining them for all platforms. `NewEvents` is conceptually delivered when the event loop "wakes up" and `MainEventsCleared` gets dispatched when the event loop is about to ask the OS to wait for new events. This is a more portable model, and is already how these events work in the MacOS backend. `RedrawEventsCleared` are just delivered after `MainEventsCleared` but this event no longer has a useful meaning. Probably the most controversial thing here is that this "breaks" the ordering rules for redraw event handling, but since my changes interacted with how the order is maintained I was very reluctant to figure out how to continue maintaining something that we have recently been discussing changing: https://github.com/rust-windowing/winit/issues/2640. Additionally, since the MacOS backend already doesn't strictly maintain this order it's somewhat academic to see this as a breakage if Winit applications can't really rely on it already. This updates the documentation for `request_redraw()` to reflect that we no longer guarantee that `RedrawRequested` events must be dispatched after `MainEventsCleared`.
2023-06-18 11:42:57 +01:00
break code;
}
};
let runner = &self.window_target.runner_shared;
Windows: Implement EventLoopExtPumpEvents and EventLoopExtRunOnDemand A surprising amount of work was required to enable these extensions on Windows. I had originally assumed that pump_events was going to be very similar to run except would use PeekMessageW instead of GetMessageW to avoid blocking the external loop but I found the Windows backend broke several assumptions I had. Overall I think these changes can hopefully be considered a quite a significant simplification (I think it's a net deletion of a fair amount of code) and I think it also helps bring it into slightly closer alignment with other backends too Key changes: - I have removed the `wait_thread` that was a fairly fiddly way of handling `ControlFlow::WaitUntil` timeouts in favor of using `SetTimer` which works with the same messages picked up by `GetMessage` and `PeekMessage`. - I have removed the ordering guarantees between `MainEventsCleared`, `RedrawRequested` and `RedrawEventsCleared` events due to the complexity in maintaining this artificial ordering, which is already not supported consistently across backends anyway (in particular this ordering already isn't compatible with how MacOS / iOS work). - `RedrawRequested` events are now directly dispatched via `WM_PAINT` messages - comparable to how `RedrawRequested` is dispatched via `drawRect` in the MacOS backend. - I have re-worked how `NewEvents`, `MainEventsCleared`, and `RedrawEventsCleared` get dispatched to be more in line with the MacOS backend and also more in line with how we have recently discussed defining them for all platforms. `NewEvents` is conceptually delivered when the event loop "wakes up" and `MainEventsCleared` gets dispatched when the event loop is about to ask the OS to wait for new events. This is a more portable model, and is already how these events work in the MacOS backend. `RedrawEventsCleared` are just delivered after `MainEventsCleared` but this event no longer has a useful meaning. Probably the most controversial thing here is that this "breaks" the ordering rules for redraw event handling, but since my changes interacted with how the order is maintained I was very reluctant to figure out how to continue maintaining something that we have recently been discussing changing: https://github.com/rust-windowing/winit/issues/2640. Additionally, since the MacOS backend already doesn't strictly maintain this order it's somewhat academic to see this as a breakage if Winit applications can't really rely on it already. This updates the documentation for `request_redraw()` to reflect that we no longer guarantee that `RedrawRequested` events must be dispatched after `MainEventsCleared`.
2023-06-18 11:42:57 +01:00
runner.loop_destroyed();
// # Safety
// We assume that this will effectively call `runner.clear_event_handler()`
// to meet the safety requirements for calling `runner.set_event_handler()` above.
runner.reset_runner();
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
Windows: Implement EventLoopExtPumpEvents and EventLoopExtRunOnDemand A surprising amount of work was required to enable these extensions on Windows. I had originally assumed that pump_events was going to be very similar to run except would use PeekMessageW instead of GetMessageW to avoid blocking the external loop but I found the Windows backend broke several assumptions I had. Overall I think these changes can hopefully be considered a quite a significant simplification (I think it's a net deletion of a fair amount of code) and I think it also helps bring it into slightly closer alignment with other backends too Key changes: - I have removed the `wait_thread` that was a fairly fiddly way of handling `ControlFlow::WaitUntil` timeouts in favor of using `SetTimer` which works with the same messages picked up by `GetMessage` and `PeekMessage`. - I have removed the ordering guarantees between `MainEventsCleared`, `RedrawRequested` and `RedrawEventsCleared` events due to the complexity in maintaining this artificial ordering, which is already not supported consistently across backends anyway (in particular this ordering already isn't compatible with how MacOS / iOS work). - `RedrawRequested` events are now directly dispatched via `WM_PAINT` messages - comparable to how `RedrawRequested` is dispatched via `drawRect` in the MacOS backend. - I have re-worked how `NewEvents`, `MainEventsCleared`, and `RedrawEventsCleared` get dispatched to be more in line with the MacOS backend and also more in line with how we have recently discussed defining them for all platforms. `NewEvents` is conceptually delivered when the event loop "wakes up" and `MainEventsCleared` gets dispatched when the event loop is about to ask the OS to wait for new events. This is a more portable model, and is already how these events work in the MacOS backend. `RedrawEventsCleared` are just delivered after `MainEventsCleared` but this event no longer has a useful meaning. Probably the most controversial thing here is that this "breaks" the ordering rules for redraw event handling, but since my changes interacted with how the order is maintained I was very reluctant to figure out how to continue maintaining something that we have recently been discussing changing: https://github.com/rust-windowing/winit/issues/2640. Additionally, since the MacOS backend already doesn't strictly maintain this order it's somewhat academic to see this as a breakage if Winit applications can't really rely on it already. This updates the documentation for `request_redraw()` to reflect that we no longer guarantee that `RedrawRequested` events must be dispatched after `MainEventsCleared`.
2023-06-18 11:42:57 +01:00
if exit_code == 0 {
Ok(())
} else {
Err(EventLoopError::ExitFailure(exit_code))
Windows: Implement EventLoopExtPumpEvents and EventLoopExtRunOnDemand A surprising amount of work was required to enable these extensions on Windows. I had originally assumed that pump_events was going to be very similar to run except would use PeekMessageW instead of GetMessageW to avoid blocking the external loop but I found the Windows backend broke several assumptions I had. Overall I think these changes can hopefully be considered a quite a significant simplification (I think it's a net deletion of a fair amount of code) and I think it also helps bring it into slightly closer alignment with other backends too Key changes: - I have removed the `wait_thread` that was a fairly fiddly way of handling `ControlFlow::WaitUntil` timeouts in favor of using `SetTimer` which works with the same messages picked up by `GetMessage` and `PeekMessage`. - I have removed the ordering guarantees between `MainEventsCleared`, `RedrawRequested` and `RedrawEventsCleared` events due to the complexity in maintaining this artificial ordering, which is already not supported consistently across backends anyway (in particular this ordering already isn't compatible with how MacOS / iOS work). - `RedrawRequested` events are now directly dispatched via `WM_PAINT` messages - comparable to how `RedrawRequested` is dispatched via `drawRect` in the MacOS backend. - I have re-worked how `NewEvents`, `MainEventsCleared`, and `RedrawEventsCleared` get dispatched to be more in line with the MacOS backend and also more in line with how we have recently discussed defining them for all platforms. `NewEvents` is conceptually delivered when the event loop "wakes up" and `MainEventsCleared` gets dispatched when the event loop is about to ask the OS to wait for new events. This is a more portable model, and is already how these events work in the MacOS backend. `RedrawEventsCleared` are just delivered after `MainEventsCleared` but this event no longer has a useful meaning. Probably the most controversial thing here is that this "breaks" the ordering rules for redraw event handling, but since my changes interacted with how the order is maintained I was very reluctant to figure out how to continue maintaining something that we have recently been discussing changing: https://github.com/rust-windowing/winit/issues/2640. Additionally, since the MacOS backend already doesn't strictly maintain this order it's somewhat academic to see this as a breakage if Winit applications can't really rely on it already. This updates the documentation for `request_redraw()` to reflect that we no longer guarantee that `RedrawRequested` events must be dispatched after `MainEventsCleared`.
2023-06-18 11:42:57 +01:00
}
}
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
pub fn pump_app_events<A: ApplicationHandler>(
&mut self,
timeout: Option<Duration>,
mut app: A,
) -> PumpStatus {
Windows: Implement EventLoopExtPumpEvents and EventLoopExtRunOnDemand A surprising amount of work was required to enable these extensions on Windows. I had originally assumed that pump_events was going to be very similar to run except would use PeekMessageW instead of GetMessageW to avoid blocking the external loop but I found the Windows backend broke several assumptions I had. Overall I think these changes can hopefully be considered a quite a significant simplification (I think it's a net deletion of a fair amount of code) and I think it also helps bring it into slightly closer alignment with other backends too Key changes: - I have removed the `wait_thread` that was a fairly fiddly way of handling `ControlFlow::WaitUntil` timeouts in favor of using `SetTimer` which works with the same messages picked up by `GetMessage` and `PeekMessage`. - I have removed the ordering guarantees between `MainEventsCleared`, `RedrawRequested` and `RedrawEventsCleared` events due to the complexity in maintaining this artificial ordering, which is already not supported consistently across backends anyway (in particular this ordering already isn't compatible with how MacOS / iOS work). - `RedrawRequested` events are now directly dispatched via `WM_PAINT` messages - comparable to how `RedrawRequested` is dispatched via `drawRect` in the MacOS backend. - I have re-worked how `NewEvents`, `MainEventsCleared`, and `RedrawEventsCleared` get dispatched to be more in line with the MacOS backend and also more in line with how we have recently discussed defining them for all platforms. `NewEvents` is conceptually delivered when the event loop "wakes up" and `MainEventsCleared` gets dispatched when the event loop is about to ask the OS to wait for new events. This is a more portable model, and is already how these events work in the MacOS backend. `RedrawEventsCleared` are just delivered after `MainEventsCleared` but this event no longer has a useful meaning. Probably the most controversial thing here is that this "breaks" the ordering rules for redraw event handling, but since my changes interacted with how the order is maintained I was very reluctant to figure out how to continue maintaining something that we have recently been discussing changing: https://github.com/rust-windowing/winit/issues/2640. Additionally, since the MacOS backend already doesn't strictly maintain this order it's somewhat academic to see this as a breakage if Winit applications can't really rely on it already. This updates the documentation for `request_redraw()` to reflect that we no longer guarantee that `RedrawRequested` events must be dispatched after `MainEventsCleared`.
2023-06-18 11:42:57 +01:00
{
let runner = &self.window_target.runner_shared;
Windows: Implement EventLoopExtPumpEvents and EventLoopExtRunOnDemand A surprising amount of work was required to enable these extensions on Windows. I had originally assumed that pump_events was going to be very similar to run except would use PeekMessageW instead of GetMessageW to avoid blocking the external loop but I found the Windows backend broke several assumptions I had. Overall I think these changes can hopefully be considered a quite a significant simplification (I think it's a net deletion of a fair amount of code) and I think it also helps bring it into slightly closer alignment with other backends too Key changes: - I have removed the `wait_thread` that was a fairly fiddly way of handling `ControlFlow::WaitUntil` timeouts in favor of using `SetTimer` which works with the same messages picked up by `GetMessage` and `PeekMessage`. - I have removed the ordering guarantees between `MainEventsCleared`, `RedrawRequested` and `RedrawEventsCleared` events due to the complexity in maintaining this artificial ordering, which is already not supported consistently across backends anyway (in particular this ordering already isn't compatible with how MacOS / iOS work). - `RedrawRequested` events are now directly dispatched via `WM_PAINT` messages - comparable to how `RedrawRequested` is dispatched via `drawRect` in the MacOS backend. - I have re-worked how `NewEvents`, `MainEventsCleared`, and `RedrawEventsCleared` get dispatched to be more in line with the MacOS backend and also more in line with how we have recently discussed defining them for all platforms. `NewEvents` is conceptually delivered when the event loop "wakes up" and `MainEventsCleared` gets dispatched when the event loop is about to ask the OS to wait for new events. This is a more portable model, and is already how these events work in the MacOS backend. `RedrawEventsCleared` are just delivered after `MainEventsCleared` but this event no longer has a useful meaning. Probably the most controversial thing here is that this "breaks" the ordering rules for redraw event handling, but since my changes interacted with how the order is maintained I was very reluctant to figure out how to continue maintaining something that we have recently been discussing changing: https://github.com/rust-windowing/winit/issues/2640. Additionally, since the MacOS backend already doesn't strictly maintain this order it's somewhat academic to see this as a breakage if Winit applications can't really rely on it already. This updates the documentation for `request_redraw()` to reflect that we no longer guarantee that `RedrawRequested` events must be dispatched after `MainEventsCleared`.
2023-06-18 11:42:57 +01:00
let event_loop_windows_ref = &self.window_target;
// let user_event_receiver = &self.user_event_receiver;
Windows: Implement EventLoopExtPumpEvents and EventLoopExtRunOnDemand A surprising amount of work was required to enable these extensions on Windows. I had originally assumed that pump_events was going to be very similar to run except would use PeekMessageW instead of GetMessageW to avoid blocking the external loop but I found the Windows backend broke several assumptions I had. Overall I think these changes can hopefully be considered a quite a significant simplification (I think it's a net deletion of a fair amount of code) and I think it also helps bring it into slightly closer alignment with other backends too Key changes: - I have removed the `wait_thread` that was a fairly fiddly way of handling `ControlFlow::WaitUntil` timeouts in favor of using `SetTimer` which works with the same messages picked up by `GetMessage` and `PeekMessage`. - I have removed the ordering guarantees between `MainEventsCleared`, `RedrawRequested` and `RedrawEventsCleared` events due to the complexity in maintaining this artificial ordering, which is already not supported consistently across backends anyway (in particular this ordering already isn't compatible with how MacOS / iOS work). - `RedrawRequested` events are now directly dispatched via `WM_PAINT` messages - comparable to how `RedrawRequested` is dispatched via `drawRect` in the MacOS backend. - I have re-worked how `NewEvents`, `MainEventsCleared`, and `RedrawEventsCleared` get dispatched to be more in line with the MacOS backend and also more in line with how we have recently discussed defining them for all platforms. `NewEvents` is conceptually delivered when the event loop "wakes up" and `MainEventsCleared` gets dispatched when the event loop is about to ask the OS to wait for new events. This is a more portable model, and is already how these events work in the MacOS backend. `RedrawEventsCleared` are just delivered after `MainEventsCleared` but this event no longer has a useful meaning. Probably the most controversial thing here is that this "breaks" the ordering rules for redraw event handling, but since my changes interacted with how the order is maintained I was very reluctant to figure out how to continue maintaining something that we have recently been discussing changing: https://github.com/rust-windowing/winit/issues/2640. Additionally, since the MacOS backend already doesn't strictly maintain this order it's somewhat academic to see this as a breakage if Winit applications can't really rely on it already. This updates the documentation for `request_redraw()` to reflect that we no longer guarantee that `RedrawRequested` events must be dispatched after `MainEventsCleared`.
2023-06-18 11:42:57 +01:00
// # Safety
// We make sure to call runner.clear_event_handler() before
// returning
//
// Note: we're currently assuming nothing can panic and unwind
// to leave the runner in an unsound state with an associated
// event handler.
unsafe {
runner.set_event_handler(move |event| match event {
Event::NewEvents(cause) => app.new_events(event_loop_windows_ref, cause),
Event::WindowEvent { window_id, event } => {
app.window_event(event_loop_windows_ref, window_id, event)
},
Event::DeviceEvent { device_id, event } => {
app.device_event(event_loop_windows_ref, device_id, event)
},
Event::UserWakeUp => app.proxy_wake_up(event_loop_windows_ref),
Event::Suspended => app.suspended(event_loop_windows_ref),
Event::Resumed => app.resumed(event_loop_windows_ref),
Event::CreateSurfaces => app.can_create_surfaces(event_loop_windows_ref),
Event::AboutToWait => app.about_to_wait(event_loop_windows_ref),
Event::LoopExiting => app.exiting(event_loop_windows_ref),
Event::MemoryWarning => app.memory_warning(event_loop_windows_ref),
Windows: Make `EventLoopWindowTarget` independent of UserEvent type (#3061) * make `EventLoopWindowTarget` independent of UserEvent type the `EventLoopWindowTarget` is needed for window creation. conceptually, only `EventLoop` and `EventLoopProxy` need to be parameterized, and all other parts of the backend should be agnostic about the user event type, parallel to how `Event<T>` is parameterized, but `WindowEvent` is not. this change removes the dependency on the type of user events from the `EventLoopWindowTarget` for the Windows backend, but keep a phantom data to keep the API intact. to achieve this, I moved the `Receiver` end of the mpsc channel from `ThreadMsgTargetData` into `EventLoop` itself, so the `UserEvent` is only passed between `EventLoop` and `EventLoopProxy`, all other part of the backend just use unit type as a placeholder for user events. it's similar to the macos backend where an erased `EventHandler` trait object is used so all component except `EventLoop` and `EventLoopProxy` need to be parameterized. however `EventLoop` of the Windows backend already use an `Box<dyn FnMut>` to wrap the user provided event handler callback, so no need for an dedicated trait object, I just modified the wrapper to replace the placeholder user event with real value pulled from the channel. I find this is the approach which need minimum change to be made to existing code. but it does the job and could serve as a starting point to future Windows backend re-works. * fix CI clippy failure. * make UserEventPlaceholder a new type instead of alias * invariance is maintained by top-level EventLoopWindowTarget<T> this field is transitional and her to keep API compatibility only. the correct variance and such is already ensured by the top-level `EventLoopWindowTarget`, just use `PhantomData<T>` here.
2024-01-04 23:47:07 +08:00
});
Windows: Implement EventLoopExtPumpEvents and EventLoopExtRunOnDemand A surprising amount of work was required to enable these extensions on Windows. I had originally assumed that pump_events was going to be very similar to run except would use PeekMessageW instead of GetMessageW to avoid blocking the external loop but I found the Windows backend broke several assumptions I had. Overall I think these changes can hopefully be considered a quite a significant simplification (I think it's a net deletion of a fair amount of code) and I think it also helps bring it into slightly closer alignment with other backends too Key changes: - I have removed the `wait_thread` that was a fairly fiddly way of handling `ControlFlow::WaitUntil` timeouts in favor of using `SetTimer` which works with the same messages picked up by `GetMessage` and `PeekMessage`. - I have removed the ordering guarantees between `MainEventsCleared`, `RedrawRequested` and `RedrawEventsCleared` events due to the complexity in maintaining this artificial ordering, which is already not supported consistently across backends anyway (in particular this ordering already isn't compatible with how MacOS / iOS work). - `RedrawRequested` events are now directly dispatched via `WM_PAINT` messages - comparable to how `RedrawRequested` is dispatched via `drawRect` in the MacOS backend. - I have re-worked how `NewEvents`, `MainEventsCleared`, and `RedrawEventsCleared` get dispatched to be more in line with the MacOS backend and also more in line with how we have recently discussed defining them for all platforms. `NewEvents` is conceptually delivered when the event loop "wakes up" and `MainEventsCleared` gets dispatched when the event loop is about to ask the OS to wait for new events. This is a more portable model, and is already how these events work in the MacOS backend. `RedrawEventsCleared` are just delivered after `MainEventsCleared` but this event no longer has a useful meaning. Probably the most controversial thing here is that this "breaks" the ordering rules for redraw event handling, but since my changes interacted with how the order is maintained I was very reluctant to figure out how to continue maintaining something that we have recently been discussing changing: https://github.com/rust-windowing/winit/issues/2640. Additionally, since the MacOS backend already doesn't strictly maintain this order it's somewhat academic to see this as a breakage if Winit applications can't really rely on it already. This updates the documentation for `request_redraw()` to reflect that we no longer guarantee that `RedrawRequested` events must be dispatched after `MainEventsCleared`.
2023-06-18 11:42:57 +01:00
runner.wakeup();
}
}
Improve waiting for messages on Windows Previous version used [`SetTimer`] with `GetMessageW` for waiting. The downside of UI timers like ones created by `SetTimer`, is that they may be late by up to 15-16 ms. To fix this behaviour, I added use of high resolution timers created by [`CreateWaitableTimerExW`] with the flag `CREATE_WAITABLE_TIMER_HIGH_RESOLUTION`. In my previous experience, waiting on such timers have precision of roundly 0.5 ms which is the best available on Windows at the moment. I use [`MsgWaitForMultipleObjectsEx`] to wait simultaneously for both timer and newly arriving events. Unfortunately, high resolution timers are available only since Windows 10 1803. However: 1. Win 10 is already getting to the end of support, like all previous versions, so it is OK to rely on APIs introduced in it; 2. I use `dwMilliseconds` parameter of `MsgWaitForMultipleObjectsEx` as a fallback. It should perform not worse compared to waiting for events from `SetTimer`. I also refactored code to remove event dispatching from function responsible for waiting for events. This provides more clear separations of concern and avoids unnecessary duplication of dispatching logic. After [review] from @rib, I also moved the waiting itself from `wait_for_messages` method to separate function, so it is clearly seen that `wait_for_messages` do 3 things: notify app that we about to wait, wait, notify that we have new events. I have tested behaviour using a egui app with Vulkan rendering with `VK_PRESENT_MODE_IMMEDIATE_KHR`, and older version consistently have twice less FPS than requested (e.g. 30 FPS when limit is 60 and 60 FPS when limit is 120) while newer version works more correctly (almost always 60 FPS when limit is 60, and only 5-10 frames missing when FPS is set to 120 or more). Fixes https://github.com/rust-windowing/winit/issues/1610 [`CreateWaitableTimerExW`]: https://learn.microsoft.com/en-us/windows/win32/api/synchapi/nf-synchapi-createwaitabletimerexw [`MsgWaitForMultipleObjectsEx`]: https://learn.microsoft.com/en-us/windows/win32/api/winuser/nf-winuser-msgwaitformultipleobjectsex [`SetTimer`]: https://learn.microsoft.com/en-us/windows/win32/api/winuser/nf-winuser-settimer [review]: https://github.com/rust-windowing/winit/pull/3950#discussion_r1800184479
2024-10-13 12:56:24 +02:00
if self.exit_code().is_none() {
self.wait_for_messages(timeout);
}
// wait_for_messages calls user application before and after waiting
// so it may have decided to exit.
if self.exit_code().is_none() {
self.dispatch_peeked_messages();
}
Windows: Implement EventLoopExtPumpEvents and EventLoopExtRunOnDemand A surprising amount of work was required to enable these extensions on Windows. I had originally assumed that pump_events was going to be very similar to run except would use PeekMessageW instead of GetMessageW to avoid blocking the external loop but I found the Windows backend broke several assumptions I had. Overall I think these changes can hopefully be considered a quite a significant simplification (I think it's a net deletion of a fair amount of code) and I think it also helps bring it into slightly closer alignment with other backends too Key changes: - I have removed the `wait_thread` that was a fairly fiddly way of handling `ControlFlow::WaitUntil` timeouts in favor of using `SetTimer` which works with the same messages picked up by `GetMessage` and `PeekMessage`. - I have removed the ordering guarantees between `MainEventsCleared`, `RedrawRequested` and `RedrawEventsCleared` events due to the complexity in maintaining this artificial ordering, which is already not supported consistently across backends anyway (in particular this ordering already isn't compatible with how MacOS / iOS work). - `RedrawRequested` events are now directly dispatched via `WM_PAINT` messages - comparable to how `RedrawRequested` is dispatched via `drawRect` in the MacOS backend. - I have re-worked how `NewEvents`, `MainEventsCleared`, and `RedrawEventsCleared` get dispatched to be more in line with the MacOS backend and also more in line with how we have recently discussed defining them for all platforms. `NewEvents` is conceptually delivered when the event loop "wakes up" and `MainEventsCleared` gets dispatched when the event loop is about to ask the OS to wait for new events. This is a more portable model, and is already how these events work in the MacOS backend. `RedrawEventsCleared` are just delivered after `MainEventsCleared` but this event no longer has a useful meaning. Probably the most controversial thing here is that this "breaks" the ordering rules for redraw event handling, but since my changes interacted with how the order is maintained I was very reluctant to figure out how to continue maintaining something that we have recently been discussing changing: https://github.com/rust-windowing/winit/issues/2640. Additionally, since the MacOS backend already doesn't strictly maintain this order it's somewhat academic to see this as a breakage if Winit applications can't really rely on it already. This updates the documentation for `request_redraw()` to reflect that we no longer guarantee that `RedrawRequested` events must be dispatched after `MainEventsCleared`.
2023-06-18 11:42:57 +01:00
let runner = &self.window_target.runner_shared;
Windows: Implement EventLoopExtPumpEvents and EventLoopExtRunOnDemand A surprising amount of work was required to enable these extensions on Windows. I had originally assumed that pump_events was going to be very similar to run except would use PeekMessageW instead of GetMessageW to avoid blocking the external loop but I found the Windows backend broke several assumptions I had. Overall I think these changes can hopefully be considered a quite a significant simplification (I think it's a net deletion of a fair amount of code) and I think it also helps bring it into slightly closer alignment with other backends too Key changes: - I have removed the `wait_thread` that was a fairly fiddly way of handling `ControlFlow::WaitUntil` timeouts in favor of using `SetTimer` which works with the same messages picked up by `GetMessage` and `PeekMessage`. - I have removed the ordering guarantees between `MainEventsCleared`, `RedrawRequested` and `RedrawEventsCleared` events due to the complexity in maintaining this artificial ordering, which is already not supported consistently across backends anyway (in particular this ordering already isn't compatible with how MacOS / iOS work). - `RedrawRequested` events are now directly dispatched via `WM_PAINT` messages - comparable to how `RedrawRequested` is dispatched via `drawRect` in the MacOS backend. - I have re-worked how `NewEvents`, `MainEventsCleared`, and `RedrawEventsCleared` get dispatched to be more in line with the MacOS backend and also more in line with how we have recently discussed defining them for all platforms. `NewEvents` is conceptually delivered when the event loop "wakes up" and `MainEventsCleared` gets dispatched when the event loop is about to ask the OS to wait for new events. This is a more portable model, and is already how these events work in the MacOS backend. `RedrawEventsCleared` are just delivered after `MainEventsCleared` but this event no longer has a useful meaning. Probably the most controversial thing here is that this "breaks" the ordering rules for redraw event handling, but since my changes interacted with how the order is maintained I was very reluctant to figure out how to continue maintaining something that we have recently been discussing changing: https://github.com/rust-windowing/winit/issues/2640. Additionally, since the MacOS backend already doesn't strictly maintain this order it's somewhat academic to see this as a breakage if Winit applications can't really rely on it already. This updates the documentation for `request_redraw()` to reflect that we no longer guarantee that `RedrawRequested` events must be dispatched after `MainEventsCleared`.
2023-06-18 11:42:57 +01:00
let status = if let Some(code) = runner.exit_code() {
Windows: Implement EventLoopExtPumpEvents and EventLoopExtRunOnDemand A surprising amount of work was required to enable these extensions on Windows. I had originally assumed that pump_events was going to be very similar to run except would use PeekMessageW instead of GetMessageW to avoid blocking the external loop but I found the Windows backend broke several assumptions I had. Overall I think these changes can hopefully be considered a quite a significant simplification (I think it's a net deletion of a fair amount of code) and I think it also helps bring it into slightly closer alignment with other backends too Key changes: - I have removed the `wait_thread` that was a fairly fiddly way of handling `ControlFlow::WaitUntil` timeouts in favor of using `SetTimer` which works with the same messages picked up by `GetMessage` and `PeekMessage`. - I have removed the ordering guarantees between `MainEventsCleared`, `RedrawRequested` and `RedrawEventsCleared` events due to the complexity in maintaining this artificial ordering, which is already not supported consistently across backends anyway (in particular this ordering already isn't compatible with how MacOS / iOS work). - `RedrawRequested` events are now directly dispatched via `WM_PAINT` messages - comparable to how `RedrawRequested` is dispatched via `drawRect` in the MacOS backend. - I have re-worked how `NewEvents`, `MainEventsCleared`, and `RedrawEventsCleared` get dispatched to be more in line with the MacOS backend and also more in line with how we have recently discussed defining them for all platforms. `NewEvents` is conceptually delivered when the event loop "wakes up" and `MainEventsCleared` gets dispatched when the event loop is about to ask the OS to wait for new events. This is a more portable model, and is already how these events work in the MacOS backend. `RedrawEventsCleared` are just delivered after `MainEventsCleared` but this event no longer has a useful meaning. Probably the most controversial thing here is that this "breaks" the ordering rules for redraw event handling, but since my changes interacted with how the order is maintained I was very reluctant to figure out how to continue maintaining something that we have recently been discussing changing: https://github.com/rust-windowing/winit/issues/2640. Additionally, since the MacOS backend already doesn't strictly maintain this order it's somewhat academic to see this as a breakage if Winit applications can't really rely on it already. This updates the documentation for `request_redraw()` to reflect that we no longer guarantee that `RedrawRequested` events must be dispatched after `MainEventsCleared`.
2023-06-18 11:42:57 +01:00
runner.loop_destroyed();
// Immediately reset the internal state for the loop to allow
// the loop to be run more than once.
runner.reset_runner();
PumpStatus::Exit(code)
} else {
runner.prepare_wait();
PumpStatus::Continue
};
// We wait until we've checked for an exit status before clearing the
// application callback, in case we need to dispatch a LoopExiting event
Windows: Implement EventLoopExtPumpEvents and EventLoopExtRunOnDemand A surprising amount of work was required to enable these extensions on Windows. I had originally assumed that pump_events was going to be very similar to run except would use PeekMessageW instead of GetMessageW to avoid blocking the external loop but I found the Windows backend broke several assumptions I had. Overall I think these changes can hopefully be considered a quite a significant simplification (I think it's a net deletion of a fair amount of code) and I think it also helps bring it into slightly closer alignment with other backends too Key changes: - I have removed the `wait_thread` that was a fairly fiddly way of handling `ControlFlow::WaitUntil` timeouts in favor of using `SetTimer` which works with the same messages picked up by `GetMessage` and `PeekMessage`. - I have removed the ordering guarantees between `MainEventsCleared`, `RedrawRequested` and `RedrawEventsCleared` events due to the complexity in maintaining this artificial ordering, which is already not supported consistently across backends anyway (in particular this ordering already isn't compatible with how MacOS / iOS work). - `RedrawRequested` events are now directly dispatched via `WM_PAINT` messages - comparable to how `RedrawRequested` is dispatched via `drawRect` in the MacOS backend. - I have re-worked how `NewEvents`, `MainEventsCleared`, and `RedrawEventsCleared` get dispatched to be more in line with the MacOS backend and also more in line with how we have recently discussed defining them for all platforms. `NewEvents` is conceptually delivered when the event loop "wakes up" and `MainEventsCleared` gets dispatched when the event loop is about to ask the OS to wait for new events. This is a more portable model, and is already how these events work in the MacOS backend. `RedrawEventsCleared` are just delivered after `MainEventsCleared` but this event no longer has a useful meaning. Probably the most controversial thing here is that this "breaks" the ordering rules for redraw event handling, but since my changes interacted with how the order is maintained I was very reluctant to figure out how to continue maintaining something that we have recently been discussing changing: https://github.com/rust-windowing/winit/issues/2640. Additionally, since the MacOS backend already doesn't strictly maintain this order it's somewhat academic to see this as a breakage if Winit applications can't really rely on it already. This updates the documentation for `request_redraw()` to reflect that we no longer guarantee that `RedrawRequested` events must be dispatched after `MainEventsCleared`.
2023-06-18 11:42:57 +01:00
//
// # Safety
// This pairs up with our call to `runner.set_event_handler` and ensures
// the application's callback can't be held beyond its lifetime.
runner.clear_event_handler();
status
}
Improve waiting for messages on Windows Previous version used [`SetTimer`] with `GetMessageW` for waiting. The downside of UI timers like ones created by `SetTimer`, is that they may be late by up to 15-16 ms. To fix this behaviour, I added use of high resolution timers created by [`CreateWaitableTimerExW`] with the flag `CREATE_WAITABLE_TIMER_HIGH_RESOLUTION`. In my previous experience, waiting on such timers have precision of roundly 0.5 ms which is the best available on Windows at the moment. I use [`MsgWaitForMultipleObjectsEx`] to wait simultaneously for both timer and newly arriving events. Unfortunately, high resolution timers are available only since Windows 10 1803. However: 1. Win 10 is already getting to the end of support, like all previous versions, so it is OK to rely on APIs introduced in it; 2. I use `dwMilliseconds` parameter of `MsgWaitForMultipleObjectsEx` as a fallback. It should perform not worse compared to waiting for events from `SetTimer`. I also refactored code to remove event dispatching from function responsible for waiting for events. This provides more clear separations of concern and avoids unnecessary duplication of dispatching logic. After [review] from @rib, I also moved the waiting itself from `wait_for_messages` method to separate function, so it is clearly seen that `wait_for_messages` do 3 things: notify app that we about to wait, wait, notify that we have new events. I have tested behaviour using a egui app with Vulkan rendering with `VK_PRESENT_MODE_IMMEDIATE_KHR`, and older version consistently have twice less FPS than requested (e.g. 30 FPS when limit is 60 and 60 FPS when limit is 120) while newer version works more correctly (almost always 60 FPS when limit is 60, and only 5-10 frames missing when FPS is set to 120 or more). Fixes https://github.com/rust-windowing/winit/issues/1610 [`CreateWaitableTimerExW`]: https://learn.microsoft.com/en-us/windows/win32/api/synchapi/nf-synchapi-createwaitabletimerexw [`MsgWaitForMultipleObjectsEx`]: https://learn.microsoft.com/en-us/windows/win32/api/winuser/nf-winuser-msgwaitformultipleobjectsex [`SetTimer`]: https://learn.microsoft.com/en-us/windows/win32/api/winuser/nf-winuser-settimer [review]: https://github.com/rust-windowing/winit/pull/3950#discussion_r1800184479
2024-10-13 12:56:24 +02:00
/// Waits until new event messages arrive to be peeked.
/// Doesn't peek messages itself.
///
/// Parameter timeout is optional. This method would wait for the smaller timeout
/// between the argument and a timeout from control flow.
fn wait_for_messages(&mut self, timeout: Option<Duration>) {
let runner = &self.window_target.runner_shared;
Windows: Implement EventLoopExtPumpEvents and EventLoopExtRunOnDemand A surprising amount of work was required to enable these extensions on Windows. I had originally assumed that pump_events was going to be very similar to run except would use PeekMessageW instead of GetMessageW to avoid blocking the external loop but I found the Windows backend broke several assumptions I had. Overall I think these changes can hopefully be considered a quite a significant simplification (I think it's a net deletion of a fair amount of code) and I think it also helps bring it into slightly closer alignment with other backends too Key changes: - I have removed the `wait_thread` that was a fairly fiddly way of handling `ControlFlow::WaitUntil` timeouts in favor of using `SetTimer` which works with the same messages picked up by `GetMessage` and `PeekMessage`. - I have removed the ordering guarantees between `MainEventsCleared`, `RedrawRequested` and `RedrawEventsCleared` events due to the complexity in maintaining this artificial ordering, which is already not supported consistently across backends anyway (in particular this ordering already isn't compatible with how MacOS / iOS work). - `RedrawRequested` events are now directly dispatched via `WM_PAINT` messages - comparable to how `RedrawRequested` is dispatched via `drawRect` in the MacOS backend. - I have re-worked how `NewEvents`, `MainEventsCleared`, and `RedrawEventsCleared` get dispatched to be more in line with the MacOS backend and also more in line with how we have recently discussed defining them for all platforms. `NewEvents` is conceptually delivered when the event loop "wakes up" and `MainEventsCleared` gets dispatched when the event loop is about to ask the OS to wait for new events. This is a more portable model, and is already how these events work in the MacOS backend. `RedrawEventsCleared` are just delivered after `MainEventsCleared` but this event no longer has a useful meaning. Probably the most controversial thing here is that this "breaks" the ordering rules for redraw event handling, but since my changes interacted with how the order is maintained I was very reluctant to figure out how to continue maintaining something that we have recently been discussing changing: https://github.com/rust-windowing/winit/issues/2640. Additionally, since the MacOS backend already doesn't strictly maintain this order it's somewhat academic to see this as a breakage if Winit applications can't really rely on it already. This updates the documentation for `request_redraw()` to reflect that we no longer guarantee that `RedrawRequested` events must be dispatched after `MainEventsCleared`.
2023-06-18 11:42:57 +01:00
// We aim to be consistent with the MacOS backend which has a RunLoop
Remove RedrawEventsCleared + MainEventsCleared, and added AboutToWait The idea that redraw events are dispatched with a specific ordering that makes it possible to specifically report when we have finished dispatching redraw events isn't portable and the way in which we dispatched RedrawEventsCleared was inconsistent across backends. More generally speaking, there is no inherent relationship between redrawing and event loop iterations. An event loop may wake up at any frequency depending on what sources of input events are being listened to but redrawing is generally throttled and in some way synchronized with the display frequency. Similarly there's no inherent relationship between a single event loop iteration and the dispatching of any specific kind of "main" event. An event loop wakes up when there are events to read (e.g. input events or responses from a display server / compositor) and goes back to waiting when there's nothing else to read. There isn't really a special kind of "main" event that is dispatched in order with respect to other events. What we can do more portably is emit an event when the event loop is about to block and wait for new events. In practice this is very similar to how MainEventsCleared was implemented except it wasn't the very last event previously since redraw events could be dispatched afterwards. The main backend where we don't strictly know when we're going to wait for events is Web (since the real event loop is internal to the browser). For now we emulate AboutToWait on Web similar to how MainEventsCleared was dispatched. In practice most applications almost certainly shouldn't care about AboutToWait because the frequency of event loop iterations is essentially arbitrary and usually irrelevant.
2023-07-28 17:37:56 +01:00
// observer that will dispatch AboutToWait when about to wait for
Windows: Implement EventLoopExtPumpEvents and EventLoopExtRunOnDemand A surprising amount of work was required to enable these extensions on Windows. I had originally assumed that pump_events was going to be very similar to run except would use PeekMessageW instead of GetMessageW to avoid blocking the external loop but I found the Windows backend broke several assumptions I had. Overall I think these changes can hopefully be considered a quite a significant simplification (I think it's a net deletion of a fair amount of code) and I think it also helps bring it into slightly closer alignment with other backends too Key changes: - I have removed the `wait_thread` that was a fairly fiddly way of handling `ControlFlow::WaitUntil` timeouts in favor of using `SetTimer` which works with the same messages picked up by `GetMessage` and `PeekMessage`. - I have removed the ordering guarantees between `MainEventsCleared`, `RedrawRequested` and `RedrawEventsCleared` events due to the complexity in maintaining this artificial ordering, which is already not supported consistently across backends anyway (in particular this ordering already isn't compatible with how MacOS / iOS work). - `RedrawRequested` events are now directly dispatched via `WM_PAINT` messages - comparable to how `RedrawRequested` is dispatched via `drawRect` in the MacOS backend. - I have re-worked how `NewEvents`, `MainEventsCleared`, and `RedrawEventsCleared` get dispatched to be more in line with the MacOS backend and also more in line with how we have recently discussed defining them for all platforms. `NewEvents` is conceptually delivered when the event loop "wakes up" and `MainEventsCleared` gets dispatched when the event loop is about to ask the OS to wait for new events. This is a more portable model, and is already how these events work in the MacOS backend. `RedrawEventsCleared` are just delivered after `MainEventsCleared` but this event no longer has a useful meaning. Probably the most controversial thing here is that this "breaks" the ordering rules for redraw event handling, but since my changes interacted with how the order is maintained I was very reluctant to figure out how to continue maintaining something that we have recently been discussing changing: https://github.com/rust-windowing/winit/issues/2640. Additionally, since the MacOS backend already doesn't strictly maintain this order it's somewhat academic to see this as a breakage if Winit applications can't really rely on it already. This updates the documentation for `request_redraw()` to reflect that we no longer guarantee that `RedrawRequested` events must be dispatched after `MainEventsCleared`.
2023-06-18 11:42:57 +01:00
// events, and NewEvents after the RunLoop wakes up.
//
Improve waiting for messages on Windows Previous version used [`SetTimer`] with `GetMessageW` for waiting. The downside of UI timers like ones created by `SetTimer`, is that they may be late by up to 15-16 ms. To fix this behaviour, I added use of high resolution timers created by [`CreateWaitableTimerExW`] with the flag `CREATE_WAITABLE_TIMER_HIGH_RESOLUTION`. In my previous experience, waiting on such timers have precision of roundly 0.5 ms which is the best available on Windows at the moment. I use [`MsgWaitForMultipleObjectsEx`] to wait simultaneously for both timer and newly arriving events. Unfortunately, high resolution timers are available only since Windows 10 1803. However: 1. Win 10 is already getting to the end of support, like all previous versions, so it is OK to rely on APIs introduced in it; 2. I use `dwMilliseconds` parameter of `MsgWaitForMultipleObjectsEx` as a fallback. It should perform not worse compared to waiting for events from `SetTimer`. I also refactored code to remove event dispatching from function responsible for waiting for events. This provides more clear separations of concern and avoids unnecessary duplication of dispatching logic. After [review] from @rib, I also moved the waiting itself from `wait_for_messages` method to separate function, so it is clearly seen that `wait_for_messages` do 3 things: notify app that we about to wait, wait, notify that we have new events. I have tested behaviour using a egui app with Vulkan rendering with `VK_PRESENT_MODE_IMMEDIATE_KHR`, and older version consistently have twice less FPS than requested (e.g. 30 FPS when limit is 60 and 60 FPS when limit is 120) while newer version works more correctly (almost always 60 FPS when limit is 60, and only 5-10 frames missing when FPS is set to 120 or more). Fixes https://github.com/rust-windowing/winit/issues/1610 [`CreateWaitableTimerExW`]: https://learn.microsoft.com/en-us/windows/win32/api/synchapi/nf-synchapi-createwaitabletimerexw [`MsgWaitForMultipleObjectsEx`]: https://learn.microsoft.com/en-us/windows/win32/api/winuser/nf-winuser-msgwaitformultipleobjectsex [`SetTimer`]: https://learn.microsoft.com/en-us/windows/win32/api/winuser/nf-winuser-settimer [review]: https://github.com/rust-windowing/winit/pull/3950#discussion_r1800184479
2024-10-13 12:56:24 +02:00
// We emulate similar behaviour by treating `MsgWaitForMultipleObjectsEx` as our wait
Windows: Implement EventLoopExtPumpEvents and EventLoopExtRunOnDemand A surprising amount of work was required to enable these extensions on Windows. I had originally assumed that pump_events was going to be very similar to run except would use PeekMessageW instead of GetMessageW to avoid blocking the external loop but I found the Windows backend broke several assumptions I had. Overall I think these changes can hopefully be considered a quite a significant simplification (I think it's a net deletion of a fair amount of code) and I think it also helps bring it into slightly closer alignment with other backends too Key changes: - I have removed the `wait_thread` that was a fairly fiddly way of handling `ControlFlow::WaitUntil` timeouts in favor of using `SetTimer` which works with the same messages picked up by `GetMessage` and `PeekMessage`. - I have removed the ordering guarantees between `MainEventsCleared`, `RedrawRequested` and `RedrawEventsCleared` events due to the complexity in maintaining this artificial ordering, which is already not supported consistently across backends anyway (in particular this ordering already isn't compatible with how MacOS / iOS work). - `RedrawRequested` events are now directly dispatched via `WM_PAINT` messages - comparable to how `RedrawRequested` is dispatched via `drawRect` in the MacOS backend. - I have re-worked how `NewEvents`, `MainEventsCleared`, and `RedrawEventsCleared` get dispatched to be more in line with the MacOS backend and also more in line with how we have recently discussed defining them for all platforms. `NewEvents` is conceptually delivered when the event loop "wakes up" and `MainEventsCleared` gets dispatched when the event loop is about to ask the OS to wait for new events. This is a more portable model, and is already how these events work in the MacOS backend. `RedrawEventsCleared` are just delivered after `MainEventsCleared` but this event no longer has a useful meaning. Probably the most controversial thing here is that this "breaks" the ordering rules for redraw event handling, but since my changes interacted with how the order is maintained I was very reluctant to figure out how to continue maintaining something that we have recently been discussing changing: https://github.com/rust-windowing/winit/issues/2640. Additionally, since the MacOS backend already doesn't strictly maintain this order it's somewhat academic to see this as a breakage if Winit applications can't really rely on it already. This updates the documentation for `request_redraw()` to reflect that we no longer guarantee that `RedrawRequested` events must be dispatched after `MainEventsCleared`.
2023-06-18 11:42:57 +01:00
// point and wake up point (when it returns) and we drain all other
// pending messages via `PeekMessage` until we come back to "wait" via
Improve waiting for messages on Windows Previous version used [`SetTimer`] with `GetMessageW` for waiting. The downside of UI timers like ones created by `SetTimer`, is that they may be late by up to 15-16 ms. To fix this behaviour, I added use of high resolution timers created by [`CreateWaitableTimerExW`] with the flag `CREATE_WAITABLE_TIMER_HIGH_RESOLUTION`. In my previous experience, waiting on such timers have precision of roundly 0.5 ms which is the best available on Windows at the moment. I use [`MsgWaitForMultipleObjectsEx`] to wait simultaneously for both timer and newly arriving events. Unfortunately, high resolution timers are available only since Windows 10 1803. However: 1. Win 10 is already getting to the end of support, like all previous versions, so it is OK to rely on APIs introduced in it; 2. I use `dwMilliseconds` parameter of `MsgWaitForMultipleObjectsEx` as a fallback. It should perform not worse compared to waiting for events from `SetTimer`. I also refactored code to remove event dispatching from function responsible for waiting for events. This provides more clear separations of concern and avoids unnecessary duplication of dispatching logic. After [review] from @rib, I also moved the waiting itself from `wait_for_messages` method to separate function, so it is clearly seen that `wait_for_messages` do 3 things: notify app that we about to wait, wait, notify that we have new events. I have tested behaviour using a egui app with Vulkan rendering with `VK_PRESENT_MODE_IMMEDIATE_KHR`, and older version consistently have twice less FPS than requested (e.g. 30 FPS when limit is 60 and 60 FPS when limit is 120) while newer version works more correctly (almost always 60 FPS when limit is 60, and only 5-10 frames missing when FPS is set to 120 or more). Fixes https://github.com/rust-windowing/winit/issues/1610 [`CreateWaitableTimerExW`]: https://learn.microsoft.com/en-us/windows/win32/api/synchapi/nf-synchapi-createwaitabletimerexw [`MsgWaitForMultipleObjectsEx`]: https://learn.microsoft.com/en-us/windows/win32/api/winuser/nf-winuser-msgwaitformultipleobjectsex [`SetTimer`]: https://learn.microsoft.com/en-us/windows/win32/api/winuser/nf-winuser-settimer [review]: https://github.com/rust-windowing/winit/pull/3950#discussion_r1800184479
2024-10-13 12:56:24 +02:00
// `MsgWaitForMultipleObjectsEx`.
Windows: Implement EventLoopExtPumpEvents and EventLoopExtRunOnDemand A surprising amount of work was required to enable these extensions on Windows. I had originally assumed that pump_events was going to be very similar to run except would use PeekMessageW instead of GetMessageW to avoid blocking the external loop but I found the Windows backend broke several assumptions I had. Overall I think these changes can hopefully be considered a quite a significant simplification (I think it's a net deletion of a fair amount of code) and I think it also helps bring it into slightly closer alignment with other backends too Key changes: - I have removed the `wait_thread` that was a fairly fiddly way of handling `ControlFlow::WaitUntil` timeouts in favor of using `SetTimer` which works with the same messages picked up by `GetMessage` and `PeekMessage`. - I have removed the ordering guarantees between `MainEventsCleared`, `RedrawRequested` and `RedrawEventsCleared` events due to the complexity in maintaining this artificial ordering, which is already not supported consistently across backends anyway (in particular this ordering already isn't compatible with how MacOS / iOS work). - `RedrawRequested` events are now directly dispatched via `WM_PAINT` messages - comparable to how `RedrawRequested` is dispatched via `drawRect` in the MacOS backend. - I have re-worked how `NewEvents`, `MainEventsCleared`, and `RedrawEventsCleared` get dispatched to be more in line with the MacOS backend and also more in line with how we have recently discussed defining them for all platforms. `NewEvents` is conceptually delivered when the event loop "wakes up" and `MainEventsCleared` gets dispatched when the event loop is about to ask the OS to wait for new events. This is a more portable model, and is already how these events work in the MacOS backend. `RedrawEventsCleared` are just delivered after `MainEventsCleared` but this event no longer has a useful meaning. Probably the most controversial thing here is that this "breaks" the ordering rules for redraw event handling, but since my changes interacted with how the order is maintained I was very reluctant to figure out how to continue maintaining something that we have recently been discussing changing: https://github.com/rust-windowing/winit/issues/2640. Additionally, since the MacOS backend already doesn't strictly maintain this order it's somewhat academic to see this as a breakage if Winit applications can't really rely on it already. This updates the documentation for `request_redraw()` to reflect that we no longer guarantee that `RedrawRequested` events must be dispatched after `MainEventsCleared`.
2023-06-18 11:42:57 +01:00
//
runner.prepare_wait();
Improve waiting for messages on Windows Previous version used [`SetTimer`] with `GetMessageW` for waiting. The downside of UI timers like ones created by `SetTimer`, is that they may be late by up to 15-16 ms. To fix this behaviour, I added use of high resolution timers created by [`CreateWaitableTimerExW`] with the flag `CREATE_WAITABLE_TIMER_HIGH_RESOLUTION`. In my previous experience, waiting on such timers have precision of roundly 0.5 ms which is the best available on Windows at the moment. I use [`MsgWaitForMultipleObjectsEx`] to wait simultaneously for both timer and newly arriving events. Unfortunately, high resolution timers are available only since Windows 10 1803. However: 1. Win 10 is already getting to the end of support, like all previous versions, so it is OK to rely on APIs introduced in it; 2. I use `dwMilliseconds` parameter of `MsgWaitForMultipleObjectsEx` as a fallback. It should perform not worse compared to waiting for events from `SetTimer`. I also refactored code to remove event dispatching from function responsible for waiting for events. This provides more clear separations of concern and avoids unnecessary duplication of dispatching logic. After [review] from @rib, I also moved the waiting itself from `wait_for_messages` method to separate function, so it is clearly seen that `wait_for_messages` do 3 things: notify app that we about to wait, wait, notify that we have new events. I have tested behaviour using a egui app with Vulkan rendering with `VK_PRESENT_MODE_IMMEDIATE_KHR`, and older version consistently have twice less FPS than requested (e.g. 30 FPS when limit is 60 and 60 FPS when limit is 120) while newer version works more correctly (almost always 60 FPS when limit is 60, and only 5-10 frames missing when FPS is set to 120 or more). Fixes https://github.com/rust-windowing/winit/issues/1610 [`CreateWaitableTimerExW`]: https://learn.microsoft.com/en-us/windows/win32/api/synchapi/nf-synchapi-createwaitabletimerexw [`MsgWaitForMultipleObjectsEx`]: https://learn.microsoft.com/en-us/windows/win32/api/winuser/nf-winuser-msgwaitformultipleobjectsex [`SetTimer`]: https://learn.microsoft.com/en-us/windows/win32/api/winuser/nf-winuser-settimer [review]: https://github.com/rust-windowing/winit/pull/3950#discussion_r1800184479
2024-10-13 12:56:24 +02:00
wait_for_messages_impl(&mut self.high_resolution_timer, runner.control_flow(), timeout);
Windows: Implement EventLoopExtPumpEvents and EventLoopExtRunOnDemand A surprising amount of work was required to enable these extensions on Windows. I had originally assumed that pump_events was going to be very similar to run except would use PeekMessageW instead of GetMessageW to avoid blocking the external loop but I found the Windows backend broke several assumptions I had. Overall I think these changes can hopefully be considered a quite a significant simplification (I think it's a net deletion of a fair amount of code) and I think it also helps bring it into slightly closer alignment with other backends too Key changes: - I have removed the `wait_thread` that was a fairly fiddly way of handling `ControlFlow::WaitUntil` timeouts in favor of using `SetTimer` which works with the same messages picked up by `GetMessage` and `PeekMessage`. - I have removed the ordering guarantees between `MainEventsCleared`, `RedrawRequested` and `RedrawEventsCleared` events due to the complexity in maintaining this artificial ordering, which is already not supported consistently across backends anyway (in particular this ordering already isn't compatible with how MacOS / iOS work). - `RedrawRequested` events are now directly dispatched via `WM_PAINT` messages - comparable to how `RedrawRequested` is dispatched via `drawRect` in the MacOS backend. - I have re-worked how `NewEvents`, `MainEventsCleared`, and `RedrawEventsCleared` get dispatched to be more in line with the MacOS backend and also more in line with how we have recently discussed defining them for all platforms. `NewEvents` is conceptually delivered when the event loop "wakes up" and `MainEventsCleared` gets dispatched when the event loop is about to ask the OS to wait for new events. This is a more portable model, and is already how these events work in the MacOS backend. `RedrawEventsCleared` are just delivered after `MainEventsCleared` but this event no longer has a useful meaning. Probably the most controversial thing here is that this "breaks" the ordering rules for redraw event handling, but since my changes interacted with how the order is maintained I was very reluctant to figure out how to continue maintaining something that we have recently been discussing changing: https://github.com/rust-windowing/winit/issues/2640. Additionally, since the MacOS backend already doesn't strictly maintain this order it's somewhat academic to see this as a breakage if Winit applications can't really rely on it already. This updates the documentation for `request_redraw()` to reflect that we no longer guarantee that `RedrawRequested` events must be dispatched after `MainEventsCleared`.
2023-06-18 11:42:57 +01:00
// Before we potentially exit, make sure to consistently emit an event for the wake up
runner.wakeup();
}
/// Dispatch all queued messages via `PeekMessageW`
fn dispatch_peeked_messages(&mut self) {
let runner = &self.window_target.runner_shared;
Windows: Implement EventLoopExtPumpEvents and EventLoopExtRunOnDemand A surprising amount of work was required to enable these extensions on Windows. I had originally assumed that pump_events was going to be very similar to run except would use PeekMessageW instead of GetMessageW to avoid blocking the external loop but I found the Windows backend broke several assumptions I had. Overall I think these changes can hopefully be considered a quite a significant simplification (I think it's a net deletion of a fair amount of code) and I think it also helps bring it into slightly closer alignment with other backends too Key changes: - I have removed the `wait_thread` that was a fairly fiddly way of handling `ControlFlow::WaitUntil` timeouts in favor of using `SetTimer` which works with the same messages picked up by `GetMessage` and `PeekMessage`. - I have removed the ordering guarantees between `MainEventsCleared`, `RedrawRequested` and `RedrawEventsCleared` events due to the complexity in maintaining this artificial ordering, which is already not supported consistently across backends anyway (in particular this ordering already isn't compatible with how MacOS / iOS work). - `RedrawRequested` events are now directly dispatched via `WM_PAINT` messages - comparable to how `RedrawRequested` is dispatched via `drawRect` in the MacOS backend. - I have re-worked how `NewEvents`, `MainEventsCleared`, and `RedrawEventsCleared` get dispatched to be more in line with the MacOS backend and also more in line with how we have recently discussed defining them for all platforms. `NewEvents` is conceptually delivered when the event loop "wakes up" and `MainEventsCleared` gets dispatched when the event loop is about to ask the OS to wait for new events. This is a more portable model, and is already how these events work in the MacOS backend. `RedrawEventsCleared` are just delivered after `MainEventsCleared` but this event no longer has a useful meaning. Probably the most controversial thing here is that this "breaks" the ordering rules for redraw event handling, but since my changes interacted with how the order is maintained I was very reluctant to figure out how to continue maintaining something that we have recently been discussing changing: https://github.com/rust-windowing/winit/issues/2640. Additionally, since the MacOS backend already doesn't strictly maintain this order it's somewhat academic to see this as a breakage if Winit applications can't really rely on it already. This updates the documentation for `request_redraw()` to reflect that we no longer guarantee that `RedrawRequested` events must be dispatched after `MainEventsCleared`.
2023-06-18 11:42:57 +01:00
// We generally want to continue dispatching all pending messages
// but we also allow dispatching to be interrupted as a means to
// ensure the `pump_events` won't indefinitely block an external
// event loop if there are too many pending events. This interrupt
// flag will be set after dispatching `RedrawRequested` events.
runner.interrupt_msg_dispatch.set(false);
// # Safety
// The Windows API has no documented requirement for bitwise
// initializing a `MSG` struct (it can be uninitialized memory for the C
// API) and there's no API to construct or initialize a `MSG`. This
// is the simplest way avoid uninitialized memory in Rust
Improve waiting for messages on Windows Previous version used [`SetTimer`] with `GetMessageW` for waiting. The downside of UI timers like ones created by `SetTimer`, is that they may be late by up to 15-16 ms. To fix this behaviour, I added use of high resolution timers created by [`CreateWaitableTimerExW`] with the flag `CREATE_WAITABLE_TIMER_HIGH_RESOLUTION`. In my previous experience, waiting on such timers have precision of roundly 0.5 ms which is the best available on Windows at the moment. I use [`MsgWaitForMultipleObjectsEx`] to wait simultaneously for both timer and newly arriving events. Unfortunately, high resolution timers are available only since Windows 10 1803. However: 1. Win 10 is already getting to the end of support, like all previous versions, so it is OK to rely on APIs introduced in it; 2. I use `dwMilliseconds` parameter of `MsgWaitForMultipleObjectsEx` as a fallback. It should perform not worse compared to waiting for events from `SetTimer`. I also refactored code to remove event dispatching from function responsible for waiting for events. This provides more clear separations of concern and avoids unnecessary duplication of dispatching logic. After [review] from @rib, I also moved the waiting itself from `wait_for_messages` method to separate function, so it is clearly seen that `wait_for_messages` do 3 things: notify app that we about to wait, wait, notify that we have new events. I have tested behaviour using a egui app with Vulkan rendering with `VK_PRESENT_MODE_IMMEDIATE_KHR`, and older version consistently have twice less FPS than requested (e.g. 30 FPS when limit is 60 and 60 FPS when limit is 120) while newer version works more correctly (almost always 60 FPS when limit is 60, and only 5-10 frames missing when FPS is set to 120 or more). Fixes https://github.com/rust-windowing/winit/issues/1610 [`CreateWaitableTimerExW`]: https://learn.microsoft.com/en-us/windows/win32/api/synchapi/nf-synchapi-createwaitabletimerexw [`MsgWaitForMultipleObjectsEx`]: https://learn.microsoft.com/en-us/windows/win32/api/winuser/nf-winuser-msgwaitformultipleobjectsex [`SetTimer`]: https://learn.microsoft.com/en-us/windows/win32/api/winuser/nf-winuser-settimer [review]: https://github.com/rust-windowing/winit/pull/3950#discussion_r1800184479
2024-10-13 12:56:24 +02:00
let mut msg: MSG = unsafe { mem::zeroed() };
Windows: Implement EventLoopExtPumpEvents and EventLoopExtRunOnDemand A surprising amount of work was required to enable these extensions on Windows. I had originally assumed that pump_events was going to be very similar to run except would use PeekMessageW instead of GetMessageW to avoid blocking the external loop but I found the Windows backend broke several assumptions I had. Overall I think these changes can hopefully be considered a quite a significant simplification (I think it's a net deletion of a fair amount of code) and I think it also helps bring it into slightly closer alignment with other backends too Key changes: - I have removed the `wait_thread` that was a fairly fiddly way of handling `ControlFlow::WaitUntil` timeouts in favor of using `SetTimer` which works with the same messages picked up by `GetMessage` and `PeekMessage`. - I have removed the ordering guarantees between `MainEventsCleared`, `RedrawRequested` and `RedrawEventsCleared` events due to the complexity in maintaining this artificial ordering, which is already not supported consistently across backends anyway (in particular this ordering already isn't compatible with how MacOS / iOS work). - `RedrawRequested` events are now directly dispatched via `WM_PAINT` messages - comparable to how `RedrawRequested` is dispatched via `drawRect` in the MacOS backend. - I have re-worked how `NewEvents`, `MainEventsCleared`, and `RedrawEventsCleared` get dispatched to be more in line with the MacOS backend and also more in line with how we have recently discussed defining them for all platforms. `NewEvents` is conceptually delivered when the event loop "wakes up" and `MainEventsCleared` gets dispatched when the event loop is about to ask the OS to wait for new events. This is a more portable model, and is already how these events work in the MacOS backend. `RedrawEventsCleared` are just delivered after `MainEventsCleared` but this event no longer has a useful meaning. Probably the most controversial thing here is that this "breaks" the ordering rules for redraw event handling, but since my changes interacted with how the order is maintained I was very reluctant to figure out how to continue maintaining something that we have recently been discussing changing: https://github.com/rust-windowing/winit/issues/2640. Additionally, since the MacOS backend already doesn't strictly maintain this order it's somewhat academic to see this as a breakage if Winit applications can't really rely on it already. This updates the documentation for `request_redraw()` to reflect that we no longer guarantee that `RedrawRequested` events must be dispatched after `MainEventsCleared`.
2023-06-18 11:42:57 +01:00
loop {
unsafe {
if PeekMessageW(&mut msg, 0, 0, 0, PM_REMOVE) == false.into() {
break;
}
let handled = if let Some(callback) = self.msg_hook.as_deref_mut() {
callback(&mut msg as *mut _ as *mut _)
} else {
false
};
if !handled {
TranslateMessage(&msg);
DispatchMessageW(&msg);
}
Windows: Implement EventLoopExtPumpEvents and EventLoopExtRunOnDemand A surprising amount of work was required to enable these extensions on Windows. I had originally assumed that pump_events was going to be very similar to run except would use PeekMessageW instead of GetMessageW to avoid blocking the external loop but I found the Windows backend broke several assumptions I had. Overall I think these changes can hopefully be considered a quite a significant simplification (I think it's a net deletion of a fair amount of code) and I think it also helps bring it into slightly closer alignment with other backends too Key changes: - I have removed the `wait_thread` that was a fairly fiddly way of handling `ControlFlow::WaitUntil` timeouts in favor of using `SetTimer` which works with the same messages picked up by `GetMessage` and `PeekMessage`. - I have removed the ordering guarantees between `MainEventsCleared`, `RedrawRequested` and `RedrawEventsCleared` events due to the complexity in maintaining this artificial ordering, which is already not supported consistently across backends anyway (in particular this ordering already isn't compatible with how MacOS / iOS work). - `RedrawRequested` events are now directly dispatched via `WM_PAINT` messages - comparable to how `RedrawRequested` is dispatched via `drawRect` in the MacOS backend. - I have re-worked how `NewEvents`, `MainEventsCleared`, and `RedrawEventsCleared` get dispatched to be more in line with the MacOS backend and also more in line with how we have recently discussed defining them for all platforms. `NewEvents` is conceptually delivered when the event loop "wakes up" and `MainEventsCleared` gets dispatched when the event loop is about to ask the OS to wait for new events. This is a more portable model, and is already how these events work in the MacOS backend. `RedrawEventsCleared` are just delivered after `MainEventsCleared` but this event no longer has a useful meaning. Probably the most controversial thing here is that this "breaks" the ordering rules for redraw event handling, but since my changes interacted with how the order is maintained I was very reluctant to figure out how to continue maintaining something that we have recently been discussing changing: https://github.com/rust-windowing/winit/issues/2640. Additionally, since the MacOS backend already doesn't strictly maintain this order it's somewhat academic to see this as a breakage if Winit applications can't really rely on it already. This updates the documentation for `request_redraw()` to reflect that we no longer guarantee that `RedrawRequested` events must be dispatched after `MainEventsCleared`.
2023-06-18 11:42:57 +01:00
}
Windows: Implement EventLoopExtPumpEvents and EventLoopExtRunOnDemand A surprising amount of work was required to enable these extensions on Windows. I had originally assumed that pump_events was going to be very similar to run except would use PeekMessageW instead of GetMessageW to avoid blocking the external loop but I found the Windows backend broke several assumptions I had. Overall I think these changes can hopefully be considered a quite a significant simplification (I think it's a net deletion of a fair amount of code) and I think it also helps bring it into slightly closer alignment with other backends too Key changes: - I have removed the `wait_thread` that was a fairly fiddly way of handling `ControlFlow::WaitUntil` timeouts in favor of using `SetTimer` which works with the same messages picked up by `GetMessage` and `PeekMessage`. - I have removed the ordering guarantees between `MainEventsCleared`, `RedrawRequested` and `RedrawEventsCleared` events due to the complexity in maintaining this artificial ordering, which is already not supported consistently across backends anyway (in particular this ordering already isn't compatible with how MacOS / iOS work). - `RedrawRequested` events are now directly dispatched via `WM_PAINT` messages - comparable to how `RedrawRequested` is dispatched via `drawRect` in the MacOS backend. - I have re-worked how `NewEvents`, `MainEventsCleared`, and `RedrawEventsCleared` get dispatched to be more in line with the MacOS backend and also more in line with how we have recently discussed defining them for all platforms. `NewEvents` is conceptually delivered when the event loop "wakes up" and `MainEventsCleared` gets dispatched when the event loop is about to ask the OS to wait for new events. This is a more portable model, and is already how these events work in the MacOS backend. `RedrawEventsCleared` are just delivered after `MainEventsCleared` but this event no longer has a useful meaning. Probably the most controversial thing here is that this "breaks" the ordering rules for redraw event handling, but since my changes interacted with how the order is maintained I was very reluctant to figure out how to continue maintaining something that we have recently been discussing changing: https://github.com/rust-windowing/winit/issues/2640. Additionally, since the MacOS backend already doesn't strictly maintain this order it's somewhat academic to see this as a breakage if Winit applications can't really rely on it already. This updates the documentation for `request_redraw()` to reflect that we no longer guarantee that `RedrawRequested` events must be dispatched after `MainEventsCleared`.
2023-06-18 11:42:57 +01:00
if let Err(payload) = runner.take_panic_error() {
runner.reset_runner();
panic::resume_unwind(payload);
}
if let Some(_code) = runner.exit_code() {
Windows: Implement EventLoopExtPumpEvents and EventLoopExtRunOnDemand A surprising amount of work was required to enable these extensions on Windows. I had originally assumed that pump_events was going to be very similar to run except would use PeekMessageW instead of GetMessageW to avoid blocking the external loop but I found the Windows backend broke several assumptions I had. Overall I think these changes can hopefully be considered a quite a significant simplification (I think it's a net deletion of a fair amount of code) and I think it also helps bring it into slightly closer alignment with other backends too Key changes: - I have removed the `wait_thread` that was a fairly fiddly way of handling `ControlFlow::WaitUntil` timeouts in favor of using `SetTimer` which works with the same messages picked up by `GetMessage` and `PeekMessage`. - I have removed the ordering guarantees between `MainEventsCleared`, `RedrawRequested` and `RedrawEventsCleared` events due to the complexity in maintaining this artificial ordering, which is already not supported consistently across backends anyway (in particular this ordering already isn't compatible with how MacOS / iOS work). - `RedrawRequested` events are now directly dispatched via `WM_PAINT` messages - comparable to how `RedrawRequested` is dispatched via `drawRect` in the MacOS backend. - I have re-worked how `NewEvents`, `MainEventsCleared`, and `RedrawEventsCleared` get dispatched to be more in line with the MacOS backend and also more in line with how we have recently discussed defining them for all platforms. `NewEvents` is conceptually delivered when the event loop "wakes up" and `MainEventsCleared` gets dispatched when the event loop is about to ask the OS to wait for new events. This is a more portable model, and is already how these events work in the MacOS backend. `RedrawEventsCleared` are just delivered after `MainEventsCleared` but this event no longer has a useful meaning. Probably the most controversial thing here is that this "breaks" the ordering rules for redraw event handling, but since my changes interacted with how the order is maintained I was very reluctant to figure out how to continue maintaining something that we have recently been discussing changing: https://github.com/rust-windowing/winit/issues/2640. Additionally, since the MacOS backend already doesn't strictly maintain this order it's somewhat academic to see this as a breakage if Winit applications can't really rely on it already. This updates the documentation for `request_redraw()` to reflect that we no longer guarantee that `RedrawRequested` events must be dispatched after `MainEventsCleared`.
2023-06-18 11:42:57 +01:00
break;
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
}
Windows: Implement EventLoopExtPumpEvents and EventLoopExtRunOnDemand A surprising amount of work was required to enable these extensions on Windows. I had originally assumed that pump_events was going to be very similar to run except would use PeekMessageW instead of GetMessageW to avoid blocking the external loop but I found the Windows backend broke several assumptions I had. Overall I think these changes can hopefully be considered a quite a significant simplification (I think it's a net deletion of a fair amount of code) and I think it also helps bring it into slightly closer alignment with other backends too Key changes: - I have removed the `wait_thread` that was a fairly fiddly way of handling `ControlFlow::WaitUntil` timeouts in favor of using `SetTimer` which works with the same messages picked up by `GetMessage` and `PeekMessage`. - I have removed the ordering guarantees between `MainEventsCleared`, `RedrawRequested` and `RedrawEventsCleared` events due to the complexity in maintaining this artificial ordering, which is already not supported consistently across backends anyway (in particular this ordering already isn't compatible with how MacOS / iOS work). - `RedrawRequested` events are now directly dispatched via `WM_PAINT` messages - comparable to how `RedrawRequested` is dispatched via `drawRect` in the MacOS backend. - I have re-worked how `NewEvents`, `MainEventsCleared`, and `RedrawEventsCleared` get dispatched to be more in line with the MacOS backend and also more in line with how we have recently discussed defining them for all platforms. `NewEvents` is conceptually delivered when the event loop "wakes up" and `MainEventsCleared` gets dispatched when the event loop is about to ask the OS to wait for new events. This is a more portable model, and is already how these events work in the MacOS backend. `RedrawEventsCleared` are just delivered after `MainEventsCleared` but this event no longer has a useful meaning. Probably the most controversial thing here is that this "breaks" the ordering rules for redraw event handling, but since my changes interacted with how the order is maintained I was very reluctant to figure out how to continue maintaining something that we have recently been discussing changing: https://github.com/rust-windowing/winit/issues/2640. Additionally, since the MacOS backend already doesn't strictly maintain this order it's somewhat academic to see this as a breakage if Winit applications can't really rely on it already. This updates the documentation for `request_redraw()` to reflect that we no longer guarantee that `RedrawRequested` events must be dispatched after `MainEventsCleared`.
2023-06-18 11:42:57 +01:00
if runner.interrupt_msg_dispatch.get() {
break;
}
}
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
}
fn exit_code(&self) -> Option<i32> {
self.window_target.exit_code()
}
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
}
Improve waiting for messages on Windows Previous version used [`SetTimer`] with `GetMessageW` for waiting. The downside of UI timers like ones created by `SetTimer`, is that they may be late by up to 15-16 ms. To fix this behaviour, I added use of high resolution timers created by [`CreateWaitableTimerExW`] with the flag `CREATE_WAITABLE_TIMER_HIGH_RESOLUTION`. In my previous experience, waiting on such timers have precision of roundly 0.5 ms which is the best available on Windows at the moment. I use [`MsgWaitForMultipleObjectsEx`] to wait simultaneously for both timer and newly arriving events. Unfortunately, high resolution timers are available only since Windows 10 1803. However: 1. Win 10 is already getting to the end of support, like all previous versions, so it is OK to rely on APIs introduced in it; 2. I use `dwMilliseconds` parameter of `MsgWaitForMultipleObjectsEx` as a fallback. It should perform not worse compared to waiting for events from `SetTimer`. I also refactored code to remove event dispatching from function responsible for waiting for events. This provides more clear separations of concern and avoids unnecessary duplication of dispatching logic. After [review] from @rib, I also moved the waiting itself from `wait_for_messages` method to separate function, so it is clearly seen that `wait_for_messages` do 3 things: notify app that we about to wait, wait, notify that we have new events. I have tested behaviour using a egui app with Vulkan rendering with `VK_PRESENT_MODE_IMMEDIATE_KHR`, and older version consistently have twice less FPS than requested (e.g. 30 FPS when limit is 60 and 60 FPS when limit is 120) while newer version works more correctly (almost always 60 FPS when limit is 60, and only 5-10 frames missing when FPS is set to 120 or more). Fixes https://github.com/rust-windowing/winit/issues/1610 [`CreateWaitableTimerExW`]: https://learn.microsoft.com/en-us/windows/win32/api/synchapi/nf-synchapi-createwaitabletimerexw [`MsgWaitForMultipleObjectsEx`]: https://learn.microsoft.com/en-us/windows/win32/api/winuser/nf-winuser-msgwaitformultipleobjectsex [`SetTimer`]: https://learn.microsoft.com/en-us/windows/win32/api/winuser/nf-winuser-settimer [review]: https://github.com/rust-windowing/winit/pull/3950#discussion_r1800184479
2024-10-13 12:56:24 +02:00
impl Drop for EventLoop {
fn drop(&mut self) {
unsafe {
DestroyWindow(self.window_target.thread_msg_target);
}
}
}
impl ActiveEventLoop {
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
#[inline(always)]
pub(crate) fn create_thread_executor(&self) -> EventLoopThreadExecutor {
EventLoopThreadExecutor { thread_id: self.thread_id, target_window: self.thread_msg_target }
}
pub(crate) fn clear_exit(&self) {
self.runner_shared.clear_exit();
}
fn exit_code(&self) -> Option<i32> {
self.runner_shared.exit_code()
}
}
impl RootActiveEventLoop for ActiveEventLoop {
fn create_proxy(&self) -> RootEventLoopProxy {
let event_loop_proxy = EventLoopProxy { target_window: self.thread_msg_target };
RootEventLoopProxy::new(Arc::new(event_loop_proxy))
}
fn create_window(
&self,
window_attributes: WindowAttributes,
) -> Result<Box<dyn CoreWindow>, RequestError> {
Ok(Box::new(Window::new(self, window_attributes)?))
}
fn create_custom_cursor(
&self,
source: CustomCursorSource,
) -> Result<RootCustomCursor, RequestError> {
Ok(RootCustomCursor { inner: WinCursor::new(&source.inner.0)? })
}
fn available_monitors(&self) -> Box<dyn Iterator<Item = crate::monitor::MonitorHandle>> {
Box::new(
monitor::available_monitors()
.into_iter()
.map(|inner| crate::monitor::MonitorHandle { inner }),
)
}
fn primary_monitor(&self) -> Option<crate::monitor::MonitorHandle> {
Some(RootMonitorHandle { inner: monitor::primary_monitor() })
}
fn exiting(&self) -> bool {
self.runner_shared.exit_code().is_some()
}
fn system_theme(&self) -> Option<Theme> {
Some(if super::dark_mode::should_use_dark_mode() { Theme::Dark } else { Theme::Light })
}
fn listen_device_events(&self, allowed: DeviceEvents) {
raw_input::register_all_mice_and_keyboards_for_raw_input(self.thread_msg_target, allowed);
}
fn set_control_flow(&self, control_flow: ControlFlow) {
self.runner_shared.set_control_flow(control_flow)
}
fn control_flow(&self) -> ControlFlow {
self.runner_shared.control_flow()
}
fn exit(&self) {
self.runner_shared.set_exit_code(0)
}
fn owned_display_handle(&self) -> CoreOwnedDisplayHandle {
CoreOwnedDisplayHandle::new(Arc::new(OwnedDisplayHandle))
}
fn rwh_06_handle(&self) -> &dyn rwh_06::HasDisplayHandle {
self
}
}
impl rwh_06::HasDisplayHandle for ActiveEventLoop {
fn display_handle(&self) -> Result<rwh_06::DisplayHandle<'_>, rwh_06::HandleError> {
let raw = rwh_06::RawDisplayHandle::Windows(rwh_06::WindowsDisplayHandle::new());
unsafe { Ok(rwh_06::DisplayHandle::borrow_raw(raw)) }
}
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
}
#[derive(Clone)]
pub(crate) struct OwnedDisplayHandle;
impl rwh_06::HasDisplayHandle for OwnedDisplayHandle {
fn display_handle(&self) -> Result<rwh_06::DisplayHandle<'_>, rwh_06::HandleError> {
let raw = rwh_06::RawDisplayHandle::Windows(rwh_06::WindowsDisplayHandle::new());
unsafe { Ok(rwh_06::DisplayHandle::borrow_raw(raw)) }
}
}
/// Returns the id of the main thread.
///
/// Windows has no real API to check if the current executing thread is the "main thread", unlike
/// macOS.
///
/// Windows will let us look up the current thread's id, but there's no API that lets us check what
/// the id of the main thread is. We would somehow need to get the main thread's id before a
/// developer could spin off any other threads inside of the main entrypoint in order to emulate the
/// capabilities of other platforms.
///
/// We can get the id of the main thread by using CRT initialization. CRT initialization can be used
/// to setup global state within a program. The OS will call a list of function pointers which
/// assign values to a static variable. To have get a hold of the main thread id, we need to place
/// our function pointer inside of the `.CRT$XCU` section so it is called before the main
/// entrypoint.
///
/// Full details of CRT initialization can be found here:
/// <https://docs.microsoft.com/en-us/cpp/c-runtime-library/crt-initialization?view=msvc-160>
2022-03-07 22:58:12 +01:00
fn main_thread_id() -> u32 {
static mut MAIN_THREAD_ID: u32 = 0;
2024-12-02 12:51:26 +01:00
// Function pointer used in CRT initialization section to set the above static field's value.
// Mark as used so this is not removable.
#[used]
#[allow(non_upper_case_globals)]
// Place the function pointer inside of CRT initialization section so it is loaded before
// main entrypoint.
//
// See: https://doc.rust-lang.org/stable/reference/abi.html#the-link_section-attribute
#[link_section = ".CRT$XCU"]
static INIT_MAIN_THREAD_ID: unsafe fn() = {
unsafe fn initer() {
unsafe { MAIN_THREAD_ID = GetCurrentThreadId() };
}
initer
};
unsafe { MAIN_THREAD_ID }
}
/// Returns the minimum `Option<Duration>`, taking into account that `None`
/// equates to an infinite timeout, not a zero timeout (so can't just use
/// `Option::min`)
fn min_timeout(a: Option<Duration>, b: Option<Duration>) -> Option<Duration> {
a.map_or(b, |a_timeout| b.map_or(Some(a_timeout), |b_timeout| Some(a_timeout.min(b_timeout))))
}
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
// Implementation taken from https://github.com/rust-lang/rust/blob/db5476571d9b27c862b95c1e64764b0ac8980e23/src/libstd/sys/windows/mod.rs
2022-03-07 22:58:12 +01:00
fn dur2timeout(dur: Duration) -> u32 {
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
// Note that a duration is a (u64, u32) (seconds, nanoseconds) pair, and the
// timeouts in windows APIs are typically u32 milliseconds. To translate, we
// have two pieces to take care of:
//
// * Nanosecond precision is rounded up
// * Greater than u32::MAX milliseconds (50 days) is rounded up to INFINITE (never time out).
dur.as_secs()
.checked_mul(1000)
.and_then(|ms| ms.checked_add((dur.subsec_nanos() as u64) / 1_000_000))
.and_then(
|ms| {
if dur.subsec_nanos() % 1_000_000 > 0 {
ms.checked_add(1)
} else {
Some(ms)
}
},
)
2022-03-07 22:58:12 +01:00
.map(|ms| if ms > u32::MAX as u64 { INFINITE } else { ms as u32 })
.unwrap_or(INFINITE)
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
}
Improve waiting for messages on Windows Previous version used [`SetTimer`] with `GetMessageW` for waiting. The downside of UI timers like ones created by `SetTimer`, is that they may be late by up to 15-16 ms. To fix this behaviour, I added use of high resolution timers created by [`CreateWaitableTimerExW`] with the flag `CREATE_WAITABLE_TIMER_HIGH_RESOLUTION`. In my previous experience, waiting on such timers have precision of roundly 0.5 ms which is the best available on Windows at the moment. I use [`MsgWaitForMultipleObjectsEx`] to wait simultaneously for both timer and newly arriving events. Unfortunately, high resolution timers are available only since Windows 10 1803. However: 1. Win 10 is already getting to the end of support, like all previous versions, so it is OK to rely on APIs introduced in it; 2. I use `dwMilliseconds` parameter of `MsgWaitForMultipleObjectsEx` as a fallback. It should perform not worse compared to waiting for events from `SetTimer`. I also refactored code to remove event dispatching from function responsible for waiting for events. This provides more clear separations of concern and avoids unnecessary duplication of dispatching logic. After [review] from @rib, I also moved the waiting itself from `wait_for_messages` method to separate function, so it is clearly seen that `wait_for_messages` do 3 things: notify app that we about to wait, wait, notify that we have new events. I have tested behaviour using a egui app with Vulkan rendering with `VK_PRESENT_MODE_IMMEDIATE_KHR`, and older version consistently have twice less FPS than requested (e.g. 30 FPS when limit is 60 and 60 FPS when limit is 120) while newer version works more correctly (almost always 60 FPS when limit is 60, and only 5-10 frames missing when FPS is set to 120 or more). Fixes https://github.com/rust-windowing/winit/issues/1610 [`CreateWaitableTimerExW`]: https://learn.microsoft.com/en-us/windows/win32/api/synchapi/nf-synchapi-createwaitabletimerexw [`MsgWaitForMultipleObjectsEx`]: https://learn.microsoft.com/en-us/windows/win32/api/winuser/nf-winuser-msgwaitformultipleobjectsex [`SetTimer`]: https://learn.microsoft.com/en-us/windows/win32/api/winuser/nf-winuser-settimer [review]: https://github.com/rust-windowing/winit/pull/3950#discussion_r1800184479
2024-10-13 12:56:24 +02:00
/// Set upper limit for waiting time to avoid overflows.
/// I chose 50 days as a limit because it is used in dur2timeout.
const FIFTY_DAYS: Duration = Duration::from_secs(50_u64 * 24 * 60 * 60);
/// Waitable timers use 100 ns intervals to indicate due time.
/// <https://learn.microsoft.com/en-us/windows/win32/api/synchapi/nf-synchapi-setwaitabletimer#parameters>
/// And there is no point waiting using other ways for such small timings
/// because they are even less precise (can overshoot by few ms).
const MIN_WAIT: Duration = Duration::from_nanos(100);
fn create_high_resolution_timer() -> Option<OwnedHandle> {
unsafe {
let handle: HANDLE = CreateWaitableTimerExW(
ptr::null(),
ptr::null(),
CREATE_WAITABLE_TIMER_HIGH_RESOLUTION,
TIMER_ALL_ACCESS,
);
// CREATE_WAITABLE_TIMER_HIGH_RESOLUTION is supported only after
// Win10 1803 but it is already default option for rustc
// (std uses it to implement `std::thread::sleep`).
if handle == 0 {
None
} else {
Some(OwnedHandle::from_raw_handle(handle as *mut c_void))
}
}
}
/// This function should not return error if parameters are valid
/// but there is no guarantee about that at MSDN docs
/// so we return result of GetLastError if fail.
///
/// ## Safety
///
/// timer must be a valid timer handle created by [create_high_resolution_timer].
/// timeout divided by 100 nanoseconds must be more than 0 and less than i64::MAX.
unsafe fn set_high_resolution_timer(timer: RawHandle, timeout: Duration) -> Result<(), u32> {
const INTERVAL_NS: u32 = MIN_WAIT.subsec_nanos();
const INTERVALS_IN_SEC: u64 = (Duration::from_secs(1).as_nanos() / INTERVAL_NS as u128) as u64;
let intervals_to_wait: u64 =
timeout.as_secs() * INTERVALS_IN_SEC + u64::from(timeout.subsec_nanos() / INTERVAL_NS);
debug_assert!(intervals_to_wait < i64::MAX as u64, "Must be called with smaller duration",);
// Use negative time to indicate relative time.
let due_time: i64 = -(intervals_to_wait as i64);
unsafe {
let set_result = SetWaitableTimer(timer as HANDLE, &due_time, 0, None, ptr::null(), FALSE);
if set_result != FALSE {
Ok(())
} else {
Err(GetLastError())
}
}
}
/// Implementation detail of [EventLoop::wait_for_messages].
///
/// Does actual system-level waiting and doesn't process any messages itself,
/// including winits internal notifications about waiting and new messages arrival.
fn wait_for_messages_impl(
high_resolution_timer: &mut Option<OwnedHandle>,
control_flow: ControlFlow,
timeout: Option<Duration>,
) {
let timeout = {
let control_flow_timeout = match control_flow {
ControlFlow::Wait => None,
ControlFlow::Poll => Some(Duration::ZERO),
ControlFlow::WaitUntil(wait_deadline) => {
let start = Instant::now();
Some(wait_deadline.saturating_duration_since(start))
},
};
let timeout = min_timeout(timeout, control_flow_timeout);
if timeout == Some(Duration::ZERO) {
// Do not wait if we don't have time.
return;
}
// Now we decided to wait so need to do some clamping
// to avoid problems with overflow and calling WinAPI with invalid parameters.
timeout
.map(|t| t.min(FIFTY_DAYS))
// If timeout is less than minimally supported by Windows,
// increase it to that minimum. Who want less than microsecond delays anyway?
.map(|t| t.max(MIN_WAIT))
};
if timeout.is_some() && high_resolution_timer.is_none() {
*high_resolution_timer = create_high_resolution_timer();
}
let high_resolution_timer: Option<RawHandle> =
high_resolution_timer.as_ref().map(OwnedHandle::as_raw_handle);
let use_timer: bool;
if let (Some(handle), Some(timeout)) = (high_resolution_timer, timeout) {
let res = unsafe {
// Safety: handle can be Some only if we succeeded in creating high resolution
// timer. We properly clamped timeout so it can be used as argument
// to timer.
set_high_resolution_timer(handle, timeout)
};
if let Err(error_code) = res {
// We successfully got timer but failed to set it?
// Should be some bug in our code.
tracing::trace!("Failed to set high resolution timer: last error {}", error_code);
use_timer = false;
} else {
use_timer = true;
}
} else {
use_timer = false;
}
unsafe {
// Either:
// 1. User wants to wait indefinely if timeout is not set.
// 2. We failed to get and set high resolution timer and we need something instead of it.
let wait_duration_ms = timeout.map(dur2timeout).unwrap_or(INFINITE);
let (num_handles, raw_handles) =
if use_timer { (1, [high_resolution_timer.unwrap()]) } else { (0, [ptr::null_mut()]) };
let result = MsgWaitForMultipleObjectsEx(
num_handles,
raw_handles.as_ptr() as *const _,
wait_duration_ms,
QS_ALLEVENTS,
MWMO_INPUTAVAILABLE,
);
if result == WAIT_FAILED {
// Well, nothing smart to do in such case.
// Treat it as spurious wake up.
tracing::warn!("Failed to MsgWaitForMultipleObjectsEx: error code {}", GetLastError(),);
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
}
}
}
pub(crate) struct EventLoopThreadExecutor {
2022-03-07 22:58:12 +01:00
thread_id: u32,
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
target_window: HWND,
}
unsafe impl Send for EventLoopThreadExecutor {}
unsafe impl Sync for EventLoopThreadExecutor {}
impl EventLoopThreadExecutor {
/// Check to see if we're in the parent event loop's thread.
pub(super) fn in_event_loop_thread(&self) -> bool {
2022-03-07 22:58:12 +01:00
let cur_thread_id = unsafe { GetCurrentThreadId() };
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
self.thread_id == cur_thread_id
}
/// Executes a function in the event loop thread. If we're already in the event loop thread,
/// we just call the function directly.
///
/// The `Inserted` can be used to inject a `WindowState` for the callback to use. The state is
/// removed automatically if the callback receives a `WM_CLOSE` message for the window.
///
/// Note that if you are using this to change some property of a window and updating
/// `WindowState` then you should call this within the lock of `WindowState`. Otherwise the
/// events may be sent to the other thread in different order to the one in which you set
/// `WindowState`, leaving them out of sync.
///
/// Note that we use a FnMut instead of a FnOnce because we're too lazy to create an equivalent
/// to the unstable FnBox.
pub(super) fn execute_in_thread<F>(&self, mut function: F)
where
F: FnMut() + Send + 'static,
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
{
unsafe {
if self.in_event_loop_thread() {
function();
} else {
// We double-box because the first box is a fat pointer.
2022-03-07 22:58:12 +01:00
let boxed2: ThreadExecFn = Box::new(Box::new(function));
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
let raw = Box::into_raw(boxed2);
let res = PostMessageW(self.target_window, EXEC_MSG_ID.get(), raw as usize, 0);
2022-03-07 22:58:12 +01:00
assert!(res != false.into(), "PostMessage failed; is the messages queue full?");
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
}
}
}
}
type ThreadExecFn = Box<Box<dyn FnMut()>>;
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
pub struct EventLoopProxy {
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
target_window: HWND,
}
unsafe impl Send for EventLoopProxy {}
impl EventLoopProxyProvider for EventLoopProxy {
fn wake_up(&self) {
unsafe { PostMessageW(self.target_window, USER_EVENT_MSG_ID.get(), 0, 0) };
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
}
}
/// A lazily-initialized window message ID.
pub struct LazyMessageId {
/// The ID.
id: AtomicU32,
/// The name of the message.
name: &'static str,
}
/// An invalid custom window ID.
const INVALID_ID: u32 = 0x0;
impl LazyMessageId {
/// Create a new `LazyId`.
const fn new(name: &'static str) -> Self {
Self { id: AtomicU32::new(INVALID_ID), name }
}
/// Get the message ID.
pub fn get(&self) -> u32 {
// Load the ID.
let id = self.id.load(Ordering::Relaxed);
if id != INVALID_ID {
return id;
}
// Register the message.
// SAFETY: We are sure that the pointer is a valid C string ending with '\0'.
assert!(self.name.ends_with('\0'));
let new_id = unsafe { RegisterWindowMessageA(self.name.as_ptr()) };
assert_ne!(
new_id,
0,
"RegisterWindowMessageA returned zero for '{}': {}",
self.name,
std::io::Error::last_os_error()
);
// Store the new ID. Since `RegisterWindowMessageA` returns the same value for any given
// string, the target value will always either be a). `INVALID_ID` or b). the
// correct ID. Therefore a compare-and-swap operation here (or really any
// consideration) is never necessary.
self.id.store(new_id, Ordering::Relaxed);
new_id
}
}
2022-06-08 11:50:26 -07:00
// Message sent by the `EventLoopProxy` when we want to wake up the thread.
// WPARAM and LPARAM are unused.
static USER_EVENT_MSG_ID: LazyMessageId = LazyMessageId::new("Winit::WakeupMsg\0");
2022-06-08 11:50:26 -07:00
// Message sent when we want to execute a closure in the thread.
// WPARAM contains a Box<Box<dyn FnMut()>> that must be retrieved with `Box::from_raw`,
// and LPARAM is unused.
static EXEC_MSG_ID: LazyMessageId = LazyMessageId::new("Winit::ExecMsg\0");
2022-06-08 11:50:26 -07:00
// Message sent by a `Window` when it wants to be destroyed by the main thread.
// WPARAM and LPARAM are unused.
pub(crate) static DESTROY_MSG_ID: LazyMessageId = LazyMessageId::new("Winit::DestroyMsg\0");
2022-06-08 11:50:26 -07:00
// WPARAM is a bool specifying the `WindowFlags::MARKER_RETAIN_STATE_ON_SIZE` flag. See the
// documentation in the `window_state` module for more information.
pub(crate) static SET_RETAIN_STATE_ON_SIZE_MSG_ID: LazyMessageId =
LazyMessageId::new("Winit::SetRetainMaximized\0");
2022-06-08 11:50:26 -07:00
static THREAD_EVENT_TARGET_WINDOW_CLASS: Lazy<Vec<u16>> =
Lazy::new(|| util::encode_wide("Winit Thread Event Target"));
/// When the taskbar is created, it registers a message with the "TaskbarCreated" string and then
/// broadcasts this message to all top-level windows <https://docs.microsoft.com/en-us/windows/win32/shell/taskbar#taskbar-creation-notification>
pub(crate) static TASKBAR_CREATED: LazyMessageId = LazyMessageId::new("TaskbarCreated\0");
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
Windows: Make `EventLoopWindowTarget` independent of UserEvent type (#3061) * make `EventLoopWindowTarget` independent of UserEvent type the `EventLoopWindowTarget` is needed for window creation. conceptually, only `EventLoop` and `EventLoopProxy` need to be parameterized, and all other parts of the backend should be agnostic about the user event type, parallel to how `Event<T>` is parameterized, but `WindowEvent` is not. this change removes the dependency on the type of user events from the `EventLoopWindowTarget` for the Windows backend, but keep a phantom data to keep the API intact. to achieve this, I moved the `Receiver` end of the mpsc channel from `ThreadMsgTargetData` into `EventLoop` itself, so the `UserEvent` is only passed between `EventLoop` and `EventLoopProxy`, all other part of the backend just use unit type as a placeholder for user events. it's similar to the macos backend where an erased `EventHandler` trait object is used so all component except `EventLoop` and `EventLoopProxy` need to be parameterized. however `EventLoop` of the Windows backend already use an `Box<dyn FnMut>` to wrap the user provided event handler callback, so no need for an dedicated trait object, I just modified the wrapper to replace the placeholder user event with real value pulled from the channel. I find this is the approach which need minimum change to be made to existing code. but it does the job and could serve as a starting point to future Windows backend re-works. * fix CI clippy failure. * make UserEventPlaceholder a new type instead of alias * invariance is maintained by top-level EventLoopWindowTarget<T> this field is transitional and her to keep API compatibility only. the correct variance and such is already ensured by the top-level `EventLoopWindowTarget`, just use `PhantomData<T>` here.
2024-01-04 23:47:07 +08:00
fn create_event_target_window() -> HWND {
use windows_sys::Win32::UI::WindowsAndMessaging::{CS_HREDRAW, CS_VREDRAW};
unsafe {
2022-03-07 22:58:12 +01:00
let class = WNDCLASSEXW {
cbSize: mem::size_of::<WNDCLASSEXW>() as u32,
style: CS_HREDRAW | CS_VREDRAW,
Windows: Make `EventLoopWindowTarget` independent of UserEvent type (#3061) * make `EventLoopWindowTarget` independent of UserEvent type the `EventLoopWindowTarget` is needed for window creation. conceptually, only `EventLoop` and `EventLoopProxy` need to be parameterized, and all other parts of the backend should be agnostic about the user event type, parallel to how `Event<T>` is parameterized, but `WindowEvent` is not. this change removes the dependency on the type of user events from the `EventLoopWindowTarget` for the Windows backend, but keep a phantom data to keep the API intact. to achieve this, I moved the `Receiver` end of the mpsc channel from `ThreadMsgTargetData` into `EventLoop` itself, so the `UserEvent` is only passed between `EventLoop` and `EventLoopProxy`, all other part of the backend just use unit type as a placeholder for user events. it's similar to the macos backend where an erased `EventHandler` trait object is used so all component except `EventLoop` and `EventLoopProxy` need to be parameterized. however `EventLoop` of the Windows backend already use an `Box<dyn FnMut>` to wrap the user provided event handler callback, so no need for an dedicated trait object, I just modified the wrapper to replace the placeholder user event with real value pulled from the channel. I find this is the approach which need minimum change to be made to existing code. but it does the job and could serve as a starting point to future Windows backend re-works. * fix CI clippy failure. * make UserEventPlaceholder a new type instead of alias * invariance is maintained by top-level EventLoopWindowTarget<T> this field is transitional and her to keep API compatibility only. the correct variance and such is already ensured by the top-level `EventLoopWindowTarget`, just use `PhantomData<T>` here.
2024-01-04 23:47:07 +08:00
lpfnWndProc: Some(thread_event_target_callback),
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
cbClsExtra: 0,
cbWndExtra: 0,
hInstance: util::get_instance_handle(),
2022-03-07 22:58:12 +01:00
hIcon: 0,
hCursor: 0, // must be null in order for cursor state to work properly
hbrBackground: 0,
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
lpszMenuName: ptr::null(),
lpszClassName: THREAD_EVENT_TARGET_WINDOW_CLASS.as_ptr(),
2022-03-07 22:58:12 +01:00
hIconSm: 0,
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
};
2022-03-07 22:58:12 +01:00
RegisterClassExW(&class);
}
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
unsafe {
// WS_EX_TOOLWINDOW prevents this window from ever showing up in the taskbar, which
// we want to avoid. If you remove this style, this window won't show up in the
// taskbar *initially*, but it can show up at some later point. This can sometimes
// happen on its own after several hours have passed, although this has proven
// difficult to reproduce. Alternatively, it can be manually triggered by killing
// `explorer.exe` and then starting the process back up.
// It is unclear why the bug is triggered by waiting for several hours.
2022-03-07 22:58:12 +01:00
let window = CreateWindowExW(
WS_EX_NOACTIVATE | WS_EX_TRANSPARENT | WS_EX_LAYERED | WS_EX_TOOLWINDOW,
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
THREAD_EVENT_TARGET_WINDOW_CLASS.as_ptr(),
2022-03-07 22:58:12 +01:00
ptr::null(),
WS_OVERLAPPED,
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
0,
0,
0,
0,
0,
2022-03-07 22:58:12 +01:00
0,
util::get_instance_handle(),
2022-03-07 22:58:12 +01:00
ptr::null(),
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
);
2022-03-07 22:58:12 +01:00
super::set_window_long(
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
window,
2022-03-07 22:58:12 +01:00
GWL_STYLE,
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
// The window technically has to be visible to receive WM_PAINT messages (which are
// used for delivering events during resizes), but it isn't displayed to
// the user because of the LAYERED style.
2022-03-07 22:58:12 +01:00
(WS_VISIBLE | WS_POPUP) as isize,
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
);
window
}
}
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
Windows: Make `EventLoopWindowTarget` independent of UserEvent type (#3061) * make `EventLoopWindowTarget` independent of UserEvent type the `EventLoopWindowTarget` is needed for window creation. conceptually, only `EventLoop` and `EventLoopProxy` need to be parameterized, and all other parts of the backend should be agnostic about the user event type, parallel to how `Event<T>` is parameterized, but `WindowEvent` is not. this change removes the dependency on the type of user events from the `EventLoopWindowTarget` for the Windows backend, but keep a phantom data to keep the API intact. to achieve this, I moved the `Receiver` end of the mpsc channel from `ThreadMsgTargetData` into `EventLoop` itself, so the `UserEvent` is only passed between `EventLoop` and `EventLoopProxy`, all other part of the backend just use unit type as a placeholder for user events. it's similar to the macos backend where an erased `EventHandler` trait object is used so all component except `EventLoop` and `EventLoopProxy` need to be parameterized. however `EventLoop` of the Windows backend already use an `Box<dyn FnMut>` to wrap the user provided event handler callback, so no need for an dedicated trait object, I just modified the wrapper to replace the placeholder user event with real value pulled from the channel. I find this is the approach which need minimum change to be made to existing code. but it does the job and could serve as a starting point to future Windows backend re-works. * fix CI clippy failure. * make UserEventPlaceholder a new type instead of alias * invariance is maintained by top-level EventLoopWindowTarget<T> this field is transitional and her to keep API compatibility only. the correct variance and such is already ensured by the top-level `EventLoopWindowTarget`, just use `PhantomData<T>` here.
2024-01-04 23:47:07 +08:00
fn insert_event_target_window_data(
thread_msg_target: HWND,
event_loop_runner: Rc<EventLoopRunner>,
Windows: Make `EventLoopWindowTarget` independent of UserEvent type (#3061) * make `EventLoopWindowTarget` independent of UserEvent type the `EventLoopWindowTarget` is needed for window creation. conceptually, only `EventLoop` and `EventLoopProxy` need to be parameterized, and all other parts of the backend should be agnostic about the user event type, parallel to how `Event<T>` is parameterized, but `WindowEvent` is not. this change removes the dependency on the type of user events from the `EventLoopWindowTarget` for the Windows backend, but keep a phantom data to keep the API intact. to achieve this, I moved the `Receiver` end of the mpsc channel from `ThreadMsgTargetData` into `EventLoop` itself, so the `UserEvent` is only passed between `EventLoop` and `EventLoopProxy`, all other part of the backend just use unit type as a placeholder for user events. it's similar to the macos backend where an erased `EventHandler` trait object is used so all component except `EventLoop` and `EventLoopProxy` need to be parameterized. however `EventLoop` of the Windows backend already use an `Box<dyn FnMut>` to wrap the user provided event handler callback, so no need for an dedicated trait object, I just modified the wrapper to replace the placeholder user event with real value pulled from the channel. I find this is the approach which need minimum change to be made to existing code. but it does the job and could serve as a starting point to future Windows backend re-works. * fix CI clippy failure. * make UserEventPlaceholder a new type instead of alias * invariance is maintained by top-level EventLoopWindowTarget<T> this field is transitional and her to keep API compatibility only. the correct variance and such is already ensured by the top-level `EventLoopWindowTarget`, just use `PhantomData<T>` here.
2024-01-04 23:47:07 +08:00
) {
let userdata = ThreadMsgTargetData { event_loop_runner };
let input_ptr = Box::into_raw(Box::new(userdata));
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
2022-03-07 22:58:12 +01:00
unsafe { super::set_window_long(thread_msg_target, GWL_USERDATA, input_ptr as isize) };
}
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
/// Capture mouse input, allowing `window` to receive mouse events when the cursor is outside of
/// the window.
unsafe fn capture_mouse(window: HWND, window_state: &mut WindowState) {
window_state.mouse.capture_count += 1;
unsafe { SetCapture(window) };
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
}
/// Release mouse input, stopping windows on this thread from receiving mouse input when the cursor
/// is outside the window.
2022-08-31 18:32:19 +02:00
unsafe fn release_mouse(mut window_state: MutexGuard<'_, WindowState>) {
window_state.mouse.capture_count = window_state.mouse.capture_count.saturating_sub(1);
if window_state.mouse.capture_count == 0 {
// ReleaseCapture() causes a WM_CAPTURECHANGED where we lock the window_state.
drop(window_state);
unsafe { ReleaseCapture() };
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
}
}
fn normalize_pointer_pressure(pressure: u32) -> Option<Force> {
match pressure {
1..=1024 => Some(Force::Normalized(pressure as f64 / 1024.0)),
_ => None,
}
}
Move `ModifiersChanged` variant to `WindowEvent` (#1381) * Move `ModifiersChanged` variant to `WindowEvent` * macos: Fix flags_changed for ModifiersChanged variant move I haven't look too deep at what this does internally, but at least cargo-check is fully happy now. :) Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * macos: Fire a ModifiersChanged event on window_did_resign_key From debugging, I determined that macOS' emission of a flagsChanged around window switching is inconsistent. It is fair to assume, I think, that when the user switches windows, they do not expect their former modifiers state to remain effective; so I think it's best to clear that state by sending a ModifiersChanged(ModifiersState::empty()). Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * windows: Fix build I don't know enough about the code to implement the fix as it is done on this branch, but this commit at least fixes the build. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * windows: Send ModifiersChanged(ModifiersState::empty) on KILLFOCUS Very similar to the changes made in [1], as focus is lost, send an event to the window indicating that the modifiers have been released. It's unclear to me (without a Windows device to test this on) whether this is necessary, but it certainly ensures that unfocused windows will have at least received this event, which is an improvement. [1]: f79f21641a31da3e4039d41be89047cdcc6028f7 Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * macos: Add a hook to update stale modifiers Sometimes, `ViewState` and `event` might have different values for their stored `modifiers` flags. These are internally stored as a bitmask in the latter and an enum in the former. We can check to see if they differ, and if they do, automatically dispatch an event to update consumers of modifier state as well as the stored `state.modifiers`. That's what the hook does. This hook is then called in the key_down, mouse_entered, mouse_exited, mouse_click, scroll_wheel, and pressure_change_with_event callbacks, which each will contain updated modifiers. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * Only call event_mods once when determining whether to update state Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * flags_changed: Memoize window_id collection Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * window_did_resign_key: Remove synthetic ModifiersChanged event We no longer need to emit this event, since we are checking the state of our modifiers before emitting most other events. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * mouse_motion: Add a call to update_potentially_stale_modifiers Now, cover all events (that I can think of, at least) where stale modifiers might affect how user programs behave. Effectively, every human-interface event (keypress, mouse click, keydown, etc.) will cause a ModifiersChanged event to be fired if something has changed. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * key_up: Add a call to update_potentially_stale_modifiers We also want to make sure modifiers state is synchronized here, too. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * mouse_motion: Remove update_potentially_stale_modifiers invocation Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * Retry CI * ViewState: Promote visibility of modifiers to the macos impl This is so that we can interact with the ViewState directly from the WindowDelegate. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * window_delegate: Synthetically set modifiers state to empty on resignKey This logic is implemented similarly on other platforms, so we wish to regain parity here. Originally this behavior was implemented to always fire an event with ModifiersState::empty(), but that was not the best as it was not necessarily correct and could be a duplicate event. This solution is perhaps the most elegant possible to implement the desired behavior of sending a synthetic empty modifiers event when a window loses focus, trading some safety for interoperation between the NSWindowDelegate and the NSView (as the objc runtime must now be consulted in order to acquire access to the ViewState which is "owned" by the NSView). Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * Check for modifiers change in window events * Fix modifier changed on macOS Since the `mouse_entered` function was generating a mouse motion, which updates the modifier state, a modifiers changed event was incorrectly generated. The updating of the modifier state has also been changed to make sure it consistently happens before events that have a modifier state attached to it, without happening on any other event. This of course means that no `CursorMoved` event is generated anymore when the user enters the window without it being focused, however I'd say that is consistent with how winit should behave. * Fix unused variable warning * Move changelog entry into `Unreleased` section Co-authored-by: Freya Gentz <zegentzy@protonmail.com> Co-authored-by: Kristofer Rye <kristofer.rye@gmail.com> Co-authored-by: Christian Duerr <contact@christianduerr.com>
2020-03-06 15:43:55 -07:00
/// Emit a `ModifiersChanged` event whenever modifiers have changed.
/// Returns the current modifier state
Windows: Make `EventLoopWindowTarget` independent of UserEvent type (#3061) * make `EventLoopWindowTarget` independent of UserEvent type the `EventLoopWindowTarget` is needed for window creation. conceptually, only `EventLoop` and `EventLoopProxy` need to be parameterized, and all other parts of the backend should be agnostic about the user event type, parallel to how `Event<T>` is parameterized, but `WindowEvent` is not. this change removes the dependency on the type of user events from the `EventLoopWindowTarget` for the Windows backend, but keep a phantom data to keep the API intact. to achieve this, I moved the `Receiver` end of the mpsc channel from `ThreadMsgTargetData` into `EventLoop` itself, so the `UserEvent` is only passed between `EventLoop` and `EventLoopProxy`, all other part of the backend just use unit type as a placeholder for user events. it's similar to the macos backend where an erased `EventHandler` trait object is used so all component except `EventLoop` and `EventLoopProxy` need to be parameterized. however `EventLoop` of the Windows backend already use an `Box<dyn FnMut>` to wrap the user provided event handler callback, so no need for an dedicated trait object, I just modified the wrapper to replace the placeholder user event with real value pulled from the channel. I find this is the approach which need minimum change to be made to existing code. but it does the job and could serve as a starting point to future Windows backend re-works. * fix CI clippy failure. * make UserEventPlaceholder a new type instead of alias * invariance is maintained by top-level EventLoopWindowTarget<T> this field is transitional and her to keep API compatibility only. the correct variance and such is already ensured by the top-level `EventLoopWindowTarget`, just use `PhantomData<T>` here.
2024-01-04 23:47:07 +08:00
fn update_modifiers(window: HWND, userdata: &WindowData) {
Move `ModifiersChanged` variant to `WindowEvent` (#1381) * Move `ModifiersChanged` variant to `WindowEvent` * macos: Fix flags_changed for ModifiersChanged variant move I haven't look too deep at what this does internally, but at least cargo-check is fully happy now. :) Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * macos: Fire a ModifiersChanged event on window_did_resign_key From debugging, I determined that macOS' emission of a flagsChanged around window switching is inconsistent. It is fair to assume, I think, that when the user switches windows, they do not expect their former modifiers state to remain effective; so I think it's best to clear that state by sending a ModifiersChanged(ModifiersState::empty()). Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * windows: Fix build I don't know enough about the code to implement the fix as it is done on this branch, but this commit at least fixes the build. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * windows: Send ModifiersChanged(ModifiersState::empty) on KILLFOCUS Very similar to the changes made in [1], as focus is lost, send an event to the window indicating that the modifiers have been released. It's unclear to me (without a Windows device to test this on) whether this is necessary, but it certainly ensures that unfocused windows will have at least received this event, which is an improvement. [1]: f79f21641a31da3e4039d41be89047cdcc6028f7 Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * macos: Add a hook to update stale modifiers Sometimes, `ViewState` and `event` might have different values for their stored `modifiers` flags. These are internally stored as a bitmask in the latter and an enum in the former. We can check to see if they differ, and if they do, automatically dispatch an event to update consumers of modifier state as well as the stored `state.modifiers`. That's what the hook does. This hook is then called in the key_down, mouse_entered, mouse_exited, mouse_click, scroll_wheel, and pressure_change_with_event callbacks, which each will contain updated modifiers. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * Only call event_mods once when determining whether to update state Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * flags_changed: Memoize window_id collection Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * window_did_resign_key: Remove synthetic ModifiersChanged event We no longer need to emit this event, since we are checking the state of our modifiers before emitting most other events. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * mouse_motion: Add a call to update_potentially_stale_modifiers Now, cover all events (that I can think of, at least) where stale modifiers might affect how user programs behave. Effectively, every human-interface event (keypress, mouse click, keydown, etc.) will cause a ModifiersChanged event to be fired if something has changed. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * key_up: Add a call to update_potentially_stale_modifiers We also want to make sure modifiers state is synchronized here, too. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * mouse_motion: Remove update_potentially_stale_modifiers invocation Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * Retry CI * ViewState: Promote visibility of modifiers to the macos impl This is so that we can interact with the ViewState directly from the WindowDelegate. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * window_delegate: Synthetically set modifiers state to empty on resignKey This logic is implemented similarly on other platforms, so we wish to regain parity here. Originally this behavior was implemented to always fire an event with ModifiersState::empty(), but that was not the best as it was not necessarily correct and could be a duplicate event. This solution is perhaps the most elegant possible to implement the desired behavior of sending a synthetic empty modifiers event when a window loses focus, trading some safety for interoperation between the NSWindowDelegate and the NSView (as the objc runtime must now be consulted in order to acquire access to the ViewState which is "owned" by the NSView). Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * Check for modifiers change in window events * Fix modifier changed on macOS Since the `mouse_entered` function was generating a mouse motion, which updates the modifier state, a modifiers changed event was incorrectly generated. The updating of the modifier state has also been changed to make sure it consistently happens before events that have a modifier state attached to it, without happening on any other event. This of course means that no `CursorMoved` event is generated anymore when the user enters the window without it being focused, however I'd say that is consistent with how winit should behave. * Fix unused variable warning * Move changelog entry into `Unreleased` section Co-authored-by: Freya Gentz <zegentzy@protonmail.com> Co-authored-by: Kristofer Rye <kristofer.rye@gmail.com> Co-authored-by: Christian Duerr <contact@christianduerr.com>
2020-03-06 15:43:55 -07:00
use crate::event::WindowEvent::ModifiersChanged;
let modifiers = {
let mut layouts = LAYOUT_CACHE.lock().unwrap();
layouts.get_agnostic_mods()
};
let mut window_state = userdata.window_state.lock().unwrap();
Move `ModifiersChanged` variant to `WindowEvent` (#1381) * Move `ModifiersChanged` variant to `WindowEvent` * macos: Fix flags_changed for ModifiersChanged variant move I haven't look too deep at what this does internally, but at least cargo-check is fully happy now. :) Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * macos: Fire a ModifiersChanged event on window_did_resign_key From debugging, I determined that macOS' emission of a flagsChanged around window switching is inconsistent. It is fair to assume, I think, that when the user switches windows, they do not expect their former modifiers state to remain effective; so I think it's best to clear that state by sending a ModifiersChanged(ModifiersState::empty()). Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * windows: Fix build I don't know enough about the code to implement the fix as it is done on this branch, but this commit at least fixes the build. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * windows: Send ModifiersChanged(ModifiersState::empty) on KILLFOCUS Very similar to the changes made in [1], as focus is lost, send an event to the window indicating that the modifiers have been released. It's unclear to me (without a Windows device to test this on) whether this is necessary, but it certainly ensures that unfocused windows will have at least received this event, which is an improvement. [1]: f79f21641a31da3e4039d41be89047cdcc6028f7 Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * macos: Add a hook to update stale modifiers Sometimes, `ViewState` and `event` might have different values for their stored `modifiers` flags. These are internally stored as a bitmask in the latter and an enum in the former. We can check to see if they differ, and if they do, automatically dispatch an event to update consumers of modifier state as well as the stored `state.modifiers`. That's what the hook does. This hook is then called in the key_down, mouse_entered, mouse_exited, mouse_click, scroll_wheel, and pressure_change_with_event callbacks, which each will contain updated modifiers. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * Only call event_mods once when determining whether to update state Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * flags_changed: Memoize window_id collection Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * window_did_resign_key: Remove synthetic ModifiersChanged event We no longer need to emit this event, since we are checking the state of our modifiers before emitting most other events. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * mouse_motion: Add a call to update_potentially_stale_modifiers Now, cover all events (that I can think of, at least) where stale modifiers might affect how user programs behave. Effectively, every human-interface event (keypress, mouse click, keydown, etc.) will cause a ModifiersChanged event to be fired if something has changed. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * key_up: Add a call to update_potentially_stale_modifiers We also want to make sure modifiers state is synchronized here, too. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * mouse_motion: Remove update_potentially_stale_modifiers invocation Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * Retry CI * ViewState: Promote visibility of modifiers to the macos impl This is so that we can interact with the ViewState directly from the WindowDelegate. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * window_delegate: Synthetically set modifiers state to empty on resignKey This logic is implemented similarly on other platforms, so we wish to regain parity here. Originally this behavior was implemented to always fire an event with ModifiersState::empty(), but that was not the best as it was not necessarily correct and could be a duplicate event. This solution is perhaps the most elegant possible to implement the desired behavior of sending a synthetic empty modifiers event when a window loses focus, trading some safety for interoperation between the NSWindowDelegate and the NSView (as the objc runtime must now be consulted in order to acquire access to the ViewState which is "owned" by the NSView). Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * Check for modifiers change in window events * Fix modifier changed on macOS Since the `mouse_entered` function was generating a mouse motion, which updates the modifier state, a modifiers changed event was incorrectly generated. The updating of the modifier state has also been changed to make sure it consistently happens before events that have a modifier state attached to it, without happening on any other event. This of course means that no `CursorMoved` event is generated anymore when the user enters the window without it being focused, however I'd say that is consistent with how winit should behave. * Fix unused variable warning * Move changelog entry into `Unreleased` section Co-authored-by: Freya Gentz <zegentzy@protonmail.com> Co-authored-by: Kristofer Rye <kristofer.rye@gmail.com> Co-authored-by: Christian Duerr <contact@christianduerr.com>
2020-03-06 15:43:55 -07:00
if window_state.modifiers_state != modifiers {
window_state.modifiers_state = modifiers;
// Drop lock
drop(window_state);
userdata.send_event(Event::WindowEvent {
window_id: WindowId::from_raw(window as usize),
event: ModifiersChanged(modifiers.into()),
});
Move `ModifiersChanged` variant to `WindowEvent` (#1381) * Move `ModifiersChanged` variant to `WindowEvent` * macos: Fix flags_changed for ModifiersChanged variant move I haven't look too deep at what this does internally, but at least cargo-check is fully happy now. :) Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * macos: Fire a ModifiersChanged event on window_did_resign_key From debugging, I determined that macOS' emission of a flagsChanged around window switching is inconsistent. It is fair to assume, I think, that when the user switches windows, they do not expect their former modifiers state to remain effective; so I think it's best to clear that state by sending a ModifiersChanged(ModifiersState::empty()). Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * windows: Fix build I don't know enough about the code to implement the fix as it is done on this branch, but this commit at least fixes the build. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * windows: Send ModifiersChanged(ModifiersState::empty) on KILLFOCUS Very similar to the changes made in [1], as focus is lost, send an event to the window indicating that the modifiers have been released. It's unclear to me (without a Windows device to test this on) whether this is necessary, but it certainly ensures that unfocused windows will have at least received this event, which is an improvement. [1]: f79f21641a31da3e4039d41be89047cdcc6028f7 Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * macos: Add a hook to update stale modifiers Sometimes, `ViewState` and `event` might have different values for their stored `modifiers` flags. These are internally stored as a bitmask in the latter and an enum in the former. We can check to see if they differ, and if they do, automatically dispatch an event to update consumers of modifier state as well as the stored `state.modifiers`. That's what the hook does. This hook is then called in the key_down, mouse_entered, mouse_exited, mouse_click, scroll_wheel, and pressure_change_with_event callbacks, which each will contain updated modifiers. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * Only call event_mods once when determining whether to update state Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * flags_changed: Memoize window_id collection Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * window_did_resign_key: Remove synthetic ModifiersChanged event We no longer need to emit this event, since we are checking the state of our modifiers before emitting most other events. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * mouse_motion: Add a call to update_potentially_stale_modifiers Now, cover all events (that I can think of, at least) where stale modifiers might affect how user programs behave. Effectively, every human-interface event (keypress, mouse click, keydown, etc.) will cause a ModifiersChanged event to be fired if something has changed. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * key_up: Add a call to update_potentially_stale_modifiers We also want to make sure modifiers state is synchronized here, too. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * mouse_motion: Remove update_potentially_stale_modifiers invocation Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * Retry CI * ViewState: Promote visibility of modifiers to the macos impl This is so that we can interact with the ViewState directly from the WindowDelegate. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * window_delegate: Synthetically set modifiers state to empty on resignKey This logic is implemented similarly on other platforms, so we wish to regain parity here. Originally this behavior was implemented to always fire an event with ModifiersState::empty(), but that was not the best as it was not necessarily correct and could be a duplicate event. This solution is perhaps the most elegant possible to implement the desired behavior of sending a synthetic empty modifiers event when a window loses focus, trading some safety for interoperation between the NSWindowDelegate and the NSView (as the objc runtime must now be consulted in order to acquire access to the ViewState which is "owned" by the NSView). Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * Check for modifiers change in window events * Fix modifier changed on macOS Since the `mouse_entered` function was generating a mouse motion, which updates the modifier state, a modifiers changed event was incorrectly generated. The updating of the modifier state has also been changed to make sure it consistently happens before events that have a modifier state attached to it, without happening on any other event. This of course means that no `CursorMoved` event is generated anymore when the user enters the window without it being focused, however I'd say that is consistent with how winit should behave. * Fix unused variable warning * Move changelog entry into `Unreleased` section Co-authored-by: Freya Gentz <zegentzy@protonmail.com> Co-authored-by: Kristofer Rye <kristofer.rye@gmail.com> Co-authored-by: Christian Duerr <contact@christianduerr.com>
2020-03-06 15:43:55 -07:00
}
}
Windows: Make `EventLoopWindowTarget` independent of UserEvent type (#3061) * make `EventLoopWindowTarget` independent of UserEvent type the `EventLoopWindowTarget` is needed for window creation. conceptually, only `EventLoop` and `EventLoopProxy` need to be parameterized, and all other parts of the backend should be agnostic about the user event type, parallel to how `Event<T>` is parameterized, but `WindowEvent` is not. this change removes the dependency on the type of user events from the `EventLoopWindowTarget` for the Windows backend, but keep a phantom data to keep the API intact. to achieve this, I moved the `Receiver` end of the mpsc channel from `ThreadMsgTargetData` into `EventLoop` itself, so the `UserEvent` is only passed between `EventLoop` and `EventLoopProxy`, all other part of the backend just use unit type as a placeholder for user events. it's similar to the macos backend where an erased `EventHandler` trait object is used so all component except `EventLoop` and `EventLoopProxy` need to be parameterized. however `EventLoop` of the Windows backend already use an `Box<dyn FnMut>` to wrap the user provided event handler callback, so no need for an dedicated trait object, I just modified the wrapper to replace the placeholder user event with real value pulled from the channel. I find this is the approach which need minimum change to be made to existing code. but it does the job and could serve as a starting point to future Windows backend re-works. * fix CI clippy failure. * make UserEventPlaceholder a new type instead of alias * invariance is maintained by top-level EventLoopWindowTarget<T> this field is transitional and her to keep API compatibility only. the correct variance and such is already ensured by the top-level `EventLoopWindowTarget`, just use `PhantomData<T>` here.
2024-01-04 23:47:07 +08:00
unsafe fn gain_active_focus(window: HWND, userdata: &WindowData) {
use crate::event::WindowEvent::Focused;
update_modifiers(window, userdata);
userdata.send_event(Event::WindowEvent {
window_id: WindowId::from_raw(window as usize),
event: Focused(true),
});
}
Windows: Make `EventLoopWindowTarget` independent of UserEvent type (#3061) * make `EventLoopWindowTarget` independent of UserEvent type the `EventLoopWindowTarget` is needed for window creation. conceptually, only `EventLoop` and `EventLoopProxy` need to be parameterized, and all other parts of the backend should be agnostic about the user event type, parallel to how `Event<T>` is parameterized, but `WindowEvent` is not. this change removes the dependency on the type of user events from the `EventLoopWindowTarget` for the Windows backend, but keep a phantom data to keep the API intact. to achieve this, I moved the `Receiver` end of the mpsc channel from `ThreadMsgTargetData` into `EventLoop` itself, so the `UserEvent` is only passed between `EventLoop` and `EventLoopProxy`, all other part of the backend just use unit type as a placeholder for user events. it's similar to the macos backend where an erased `EventHandler` trait object is used so all component except `EventLoop` and `EventLoopProxy` need to be parameterized. however `EventLoop` of the Windows backend already use an `Box<dyn FnMut>` to wrap the user provided event handler callback, so no need for an dedicated trait object, I just modified the wrapper to replace the placeholder user event with real value pulled from the channel. I find this is the approach which need minimum change to be made to existing code. but it does the job and could serve as a starting point to future Windows backend re-works. * fix CI clippy failure. * make UserEventPlaceholder a new type instead of alias * invariance is maintained by top-level EventLoopWindowTarget<T> this field is transitional and her to keep API compatibility only. the correct variance and such is already ensured by the top-level `EventLoopWindowTarget`, just use `PhantomData<T>` here.
2024-01-04 23:47:07 +08:00
unsafe fn lose_active_focus(window: HWND, userdata: &WindowData) {
use crate::event::WindowEvent::{Focused, ModifiersChanged};
2022-08-31 18:32:19 +02:00
userdata.window_state_lock().modifiers_state = ModifiersState::empty();
userdata.send_event(Event::WindowEvent {
window_id: WindowId::from_raw(window as usize),
event: ModifiersChanged(ModifiersState::empty().into()),
});
userdata.send_event(Event::WindowEvent {
window_id: WindowId::from_raw(window as usize),
event: Focused(false),
});
}
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
/// Any window whose callback is configured to this function will have its events propagated
/// through the events loop of the thread the window was created in.
// This is the callback that is called by `DispatchMessage` in the events loop.
//
// Returning 0 tells the Win32 API that the message has been processed.
// FIXME: detect WM_DWMCOMPOSITIONCHANGED and call DwmEnableBlurBehindWindow if necessary
pub(super) unsafe extern "system" fn public_window_callback(
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
window: HWND,
2022-03-07 22:58:12 +01:00
msg: u32,
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
wparam: WPARAM,
lparam: LPARAM,
) -> LRESULT {
let userdata = unsafe { super::get_window_long(window, GWL_USERDATA) };
2022-03-07 22:58:12 +01:00
let userdata_ptr = match (userdata, msg) {
2022-03-07 22:58:12 +01:00
(0, WM_NCCREATE) => {
let createstruct = unsafe { &mut *(lparam as *mut CREATESTRUCTW) };
let initdata = unsafe { &mut *(createstruct.lpCreateParams as *mut InitData<'_>) };
let result = match unsafe { initdata.on_nccreate(window) } {
Some(userdata) => unsafe {
2022-03-07 22:58:12 +01:00
super::set_window_long(window, GWL_USERDATA, userdata as _);
DefWindowProcW(window, msg, wparam, lparam)
},
None => -1, // failed to create the window
};
return result;
},
// Getting here should quite frankly be impossible,
// but we'll make window creation fail here just in case.
2022-03-07 22:58:12 +01:00
(0, WM_CREATE) => return -1,
(_, WM_CREATE) => unsafe {
2022-03-07 22:58:12 +01:00
let createstruct = &mut *(lparam as *mut CREATESTRUCTW);
let initdata = createstruct.lpCreateParams;
let initdata = &mut *(initdata as *mut InitData<'_>);
initdata.on_create();
2022-03-07 22:58:12 +01:00
return DefWindowProcW(window, msg, wparam, lparam);
},
(0, _) => return unsafe { DefWindowProcW(window, msg, wparam, lparam) },
Windows: Make `EventLoopWindowTarget` independent of UserEvent type (#3061) * make `EventLoopWindowTarget` independent of UserEvent type the `EventLoopWindowTarget` is needed for window creation. conceptually, only `EventLoop` and `EventLoopProxy` need to be parameterized, and all other parts of the backend should be agnostic about the user event type, parallel to how `Event<T>` is parameterized, but `WindowEvent` is not. this change removes the dependency on the type of user events from the `EventLoopWindowTarget` for the Windows backend, but keep a phantom data to keep the API intact. to achieve this, I moved the `Receiver` end of the mpsc channel from `ThreadMsgTargetData` into `EventLoop` itself, so the `UserEvent` is only passed between `EventLoop` and `EventLoopProxy`, all other part of the backend just use unit type as a placeholder for user events. it's similar to the macos backend where an erased `EventHandler` trait object is used so all component except `EventLoop` and `EventLoopProxy` need to be parameterized. however `EventLoop` of the Windows backend already use an `Box<dyn FnMut>` to wrap the user provided event handler callback, so no need for an dedicated trait object, I just modified the wrapper to replace the placeholder user event with real value pulled from the channel. I find this is the approach which need minimum change to be made to existing code. but it does the job and could serve as a starting point to future Windows backend re-works. * fix CI clippy failure. * make UserEventPlaceholder a new type instead of alias * invariance is maintained by top-level EventLoopWindowTarget<T> this field is transitional and her to keep API compatibility only. the correct variance and such is already ensured by the top-level `EventLoopWindowTarget`, just use `PhantomData<T>` here.
2024-01-04 23:47:07 +08:00
_ => userdata as *mut WindowData,
};
let (result, userdata_removed, recurse_depth) = {
let userdata = unsafe { &*(userdata_ptr) };
userdata.recurse_depth.set(userdata.recurse_depth.get() + 1);
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
let result = unsafe { public_window_callback_inner(window, msg, wparam, lparam, userdata) };
let userdata_removed = userdata.userdata_removed.get();
let recurse_depth = userdata.recurse_depth.get() - 1;
userdata.recurse_depth.set(recurse_depth);
(result, userdata_removed, recurse_depth)
};
if userdata_removed && recurse_depth == 0 {
drop(unsafe { Box::from_raw(userdata_ptr) });
}
result
}
Windows: Make `EventLoopWindowTarget` independent of UserEvent type (#3061) * make `EventLoopWindowTarget` independent of UserEvent type the `EventLoopWindowTarget` is needed for window creation. conceptually, only `EventLoop` and `EventLoopProxy` need to be parameterized, and all other parts of the backend should be agnostic about the user event type, parallel to how `Event<T>` is parameterized, but `WindowEvent` is not. this change removes the dependency on the type of user events from the `EventLoopWindowTarget` for the Windows backend, but keep a phantom data to keep the API intact. to achieve this, I moved the `Receiver` end of the mpsc channel from `ThreadMsgTargetData` into `EventLoop` itself, so the `UserEvent` is only passed between `EventLoop` and `EventLoopProxy`, all other part of the backend just use unit type as a placeholder for user events. it's similar to the macos backend where an erased `EventHandler` trait object is used so all component except `EventLoop` and `EventLoopProxy` need to be parameterized. however `EventLoop` of the Windows backend already use an `Box<dyn FnMut>` to wrap the user provided event handler callback, so no need for an dedicated trait object, I just modified the wrapper to replace the placeholder user event with real value pulled from the channel. I find this is the approach which need minimum change to be made to existing code. but it does the job and could serve as a starting point to future Windows backend re-works. * fix CI clippy failure. * make UserEventPlaceholder a new type instead of alias * invariance is maintained by top-level EventLoopWindowTarget<T> this field is transitional and her to keep API compatibility only. the correct variance and such is already ensured by the top-level `EventLoopWindowTarget`, just use `PhantomData<T>` here.
2024-01-04 23:47:07 +08:00
unsafe fn public_window_callback_inner(
window: HWND,
2022-03-07 22:58:12 +01:00
msg: u32,
wparam: WPARAM,
lparam: LPARAM,
Windows: Make `EventLoopWindowTarget` independent of UserEvent type (#3061) * make `EventLoopWindowTarget` independent of UserEvent type the `EventLoopWindowTarget` is needed for window creation. conceptually, only `EventLoop` and `EventLoopProxy` need to be parameterized, and all other parts of the backend should be agnostic about the user event type, parallel to how `Event<T>` is parameterized, but `WindowEvent` is not. this change removes the dependency on the type of user events from the `EventLoopWindowTarget` for the Windows backend, but keep a phantom data to keep the API intact. to achieve this, I moved the `Receiver` end of the mpsc channel from `ThreadMsgTargetData` into `EventLoop` itself, so the `UserEvent` is only passed between `EventLoop` and `EventLoopProxy`, all other part of the backend just use unit type as a placeholder for user events. it's similar to the macos backend where an erased `EventHandler` trait object is used so all component except `EventLoop` and `EventLoopProxy` need to be parameterized. however `EventLoop` of the Windows backend already use an `Box<dyn FnMut>` to wrap the user provided event handler callback, so no need for an dedicated trait object, I just modified the wrapper to replace the placeholder user event with real value pulled from the channel. I find this is the approach which need minimum change to be made to existing code. but it does the job and could serve as a starting point to future Windows backend re-works. * fix CI clippy failure. * make UserEventPlaceholder a new type instead of alias * invariance is maintained by top-level EventLoopWindowTarget<T> this field is transitional and her to keep API compatibility only. the correct variance and such is already ensured by the top-level `EventLoopWindowTarget`, just use `PhantomData<T>` here.
2024-01-04 23:47:07 +08:00
userdata: &WindowData,
) -> LRESULT {
let mut result = ProcResult::DefWindowProc(wparam);
// Send new modifiers before sending key events.
let mods_changed_callback = || match msg {
WM_KEYDOWN | WM_SYSKEYDOWN | WM_KEYUP | WM_SYSKEYUP => {
update_modifiers(window, userdata);
result = ProcResult::Value(0);
},
_ => (),
};
userdata
.event_loop_runner
.catch_unwind(mods_changed_callback)
.unwrap_or_else(|| result = ProcResult::Value(-1));
let keyboard_callback = || {
use crate::event::WindowEvent::KeyboardInput;
let events =
userdata.key_event_builder.process_message(window, msg, wparam, lparam, &mut result);
for event in events {
userdata.send_event(Event::WindowEvent {
window_id: WindowId::from_raw(window as usize),
event: KeyboardInput {
device_id: None,
event: event.event,
is_synthetic: event.is_synthetic,
},
});
}
};
userdata
.event_loop_runner
.catch_unwind(keyboard_callback)
.unwrap_or_else(|| result = ProcResult::Value(-1));
// I decided to bind the closure to `callback` and pass it to catch_unwind rather than passing
// the closure to catch_unwind directly so that the match body indentation wouldn't change and
// the git blame and history would be preserved.
let callback = || match msg {
WM_NCCALCSIZE => {
2022-08-31 18:32:19 +02:00
let window_flags = userdata.window_state_lock().window_flags;
if wparam == 0 || window_flags.contains(WindowFlags::MARKER_DECORATIONS) {
result = ProcResult::DefWindowProc(wparam);
return;
}
let params = unsafe { &mut *(lparam as *mut NCCALCSIZE_PARAMS) };
if util::is_maximized(window) {
// Limit the window size when maximized to the current monitor.
// Otherwise it would include the non-existent decorations.
//
// Use `MonitorFromRect` instead of `MonitorFromWindow` to select
// the correct monitor here.
// See https://github.com/MicrosoftEdge/WebView2Feedback/issues/2549
let monitor = unsafe { MonitorFromRect(&params.rgrc[0], MONITOR_DEFAULTTONULL) };
if let Ok(monitor_info) = monitor::get_monitor_info(monitor) {
params.rgrc[0] = monitor_info.monitorInfo.rcWork;
}
} else if window_flags.contains(WindowFlags::MARKER_UNDECORATED_SHADOW) {
// Extend the client area to cover the whole non-client area.
// https://docs.microsoft.com/en-us/windows/win32/winmsg/wm-nccalcsize#remarks
//
// HACK(msiglreith): To add the drop shadow we slightly tweak the non-client area.
// This leads to a small black 1px border on the top. Adding a margin manually
// on all 4 borders would result in the caption getting drawn by the DWM.
//
// Another option would be to allow the DWM to paint inside the client area.
// Unfortunately this results in janky resize behavior, where the compositor is
// ahead of the window surface. Currently, there seems no option to achieve this
// with the Windows API.
params.rgrc[0].top += 1;
params.rgrc[0].bottom += 1;
}
result = ProcResult::Value(0);
},
2022-03-07 22:58:12 +01:00
WM_ENTERSIZEMOVE => {
userdata
2022-08-31 18:32:19 +02:00
.window_state_lock()
.set_window_flags_in_place(|f| f.insert(WindowFlags::MARKER_IN_SIZE_MOVE));
result = ProcResult::Value(0);
2019-06-24 12:14:55 -04:00
},
2022-03-07 22:58:12 +01:00
WM_EXITSIZEMOVE => {
let mut state = userdata.window_state_lock();
if state.dragging {
state.dragging = false;
unsafe { PostMessageW(window, WM_LBUTTONUP, 0, lparam) };
}
state.set_window_flags_in_place(|f| f.remove(WindowFlags::MARKER_IN_SIZE_MOVE));
result = ProcResult::Value(0);
2019-06-24 12:14:55 -04:00
},
2022-03-07 22:58:12 +01:00
WM_NCLBUTTONDOWN => {
if wparam == HTCAPTION as _ {
unsafe { PostMessageW(window, WM_MOUSEMOVE, 0, lparam) };
}
result = ProcResult::DefWindowProc(wparam);
2019-06-24 12:14:55 -04:00
},
2022-03-07 22:58:12 +01:00
WM_CLOSE => {
use crate::event::WindowEvent::CloseRequested;
userdata.send_event(Event::WindowEvent {
window_id: WindowId::from_raw(window as usize),
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
event: CloseRequested,
});
result = ProcResult::Value(0);
2019-06-24 12:14:55 -04:00
},
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
2022-03-07 22:58:12 +01:00
WM_DESTROY => {
use crate::event::WindowEvent::Destroyed;
unsafe { RevokeDragDrop(window) };
userdata.send_event(Event::WindowEvent {
window_id: WindowId::from_raw(window as usize),
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
event: Destroyed,
});
result = ProcResult::Value(0);
},
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
2022-03-07 22:58:12 +01:00
WM_NCDESTROY => {
unsafe { super::set_window_long(window, GWL_USERDATA, 0) };
userdata.userdata_removed.set(true);
result = ProcResult::Value(0);
2019-06-24 12:14:55 -04:00
},
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
2022-03-07 22:58:12 +01:00
WM_PAINT => {
userdata.window_state_lock().redraw_requested =
userdata.event_loop_runner.should_buffer();
// We'll buffer only in response to `UpdateWindow`, if win32 decides to redraw the
// window outside the normal flow of the event loop. This way mark event as handled
// and request a normal redraw with `RedrawWindow`.
if !userdata.event_loop_runner.should_buffer() {
userdata.send_event(Event::WindowEvent {
window_id: WindowId::from_raw(window as usize),
event: WindowEvent::RedrawRequested,
});
}
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
// NOTE: calling `RedrawWindow` during `WM_PAINT` does nothing, since to mark
// `WM_PAINT` as handled we should call the `DefWindowProcW`. Call it and check whether
// user asked for redraw during `RedrawRequested` event handling and request it again
// after marking `WM_PAINT` as handled.
result = ProcResult::Value(unsafe { DefWindowProcW(window, msg, wparam, lparam) });
if std::mem::take(&mut userdata.window_state_lock().redraw_requested) {
unsafe { RedrawWindow(window, ptr::null(), 0, RDW_INTERNALPAINT) };
}
},
2022-03-07 22:58:12 +01:00
WM_WINDOWPOSCHANGING => {
2022-08-31 18:32:19 +02:00
let mut window_state = userdata.window_state_lock();
if let Some(ref mut fullscreen) = window_state.fullscreen {
let window_pos = unsafe { &mut *(lparam as *mut WINDOWPOS) };
let new_rect = RECT {
left: window_pos.x,
top: window_pos.y,
right: window_pos.x + window_pos.cx,
bottom: window_pos.y + window_pos.cy,
};
const NOMOVE_OR_NOSIZE: u32 = SWP_NOMOVE | SWP_NOSIZE;
let new_rect = if window_pos.flags & NOMOVE_OR_NOSIZE != 0 {
let cur_rect = util::WindowArea::Outer.get_rect(window).expect(
"Unexpected GetWindowRect failure; please report this error to \
rust-windowing/winit",
);
match window_pos.flags & NOMOVE_OR_NOSIZE {
NOMOVE_OR_NOSIZE => None,
SWP_NOMOVE => Some(RECT {
left: cur_rect.left,
top: cur_rect.top,
right: cur_rect.left + window_pos.cx,
bottom: cur_rect.top + window_pos.cy,
}),
SWP_NOSIZE => Some(RECT {
left: window_pos.x,
top: window_pos.y,
right: window_pos.x - cur_rect.left + cur_rect.right,
bottom: window_pos.y - cur_rect.top + cur_rect.bottom,
}),
_ => unreachable!(),
}
} else {
Some(new_rect)
};
if let Some(new_rect) = new_rect {
let new_monitor = unsafe { MonitorFromRect(&new_rect, MONITOR_DEFAULTTONULL) };
match fullscreen {
Fullscreen::Borderless(ref mut fullscreen_monitor) => {
if new_monitor != 0
&& fullscreen_monitor
.as_ref()
2022-09-21 10:04:28 +02:00
.map(|monitor| new_monitor != monitor.hmonitor())
.unwrap_or(true)
{
if let Ok(new_monitor_info) = monitor::get_monitor_info(new_monitor)
{
let new_monitor_rect = new_monitor_info.monitorInfo.rcMonitor;
window_pos.x = new_monitor_rect.left;
window_pos.y = new_monitor_rect.top;
window_pos.cx = new_monitor_rect.right - new_monitor_rect.left;
window_pos.cy = new_monitor_rect.bottom - new_monitor_rect.top;
}
2022-09-21 10:04:28 +02:00
*fullscreen_monitor = Some(MonitorHandle::new(new_monitor));
}
},
Fullscreen::Exclusive(ref video_mode) => {
2022-09-21 10:04:28 +02:00
let old_monitor = video_mode.monitor.hmonitor();
if let Ok(old_monitor_info) = monitor::get_monitor_info(old_monitor) {
let old_monitor_rect = old_monitor_info.monitorInfo.rcMonitor;
window_pos.x = old_monitor_rect.left;
window_pos.y = old_monitor_rect.top;
window_pos.cx = old_monitor_rect.right - old_monitor_rect.left;
window_pos.cy = old_monitor_rect.bottom - old_monitor_rect.top;
}
},
}
}
}
result = ProcResult::Value(0);
},
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
// WM_MOVE supplies client area positions, so we send Moved here instead.
2022-03-07 22:58:12 +01:00
WM_WINDOWPOSCHANGED => {
use crate::event::WindowEvent::Moved;
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
2022-03-07 22:58:12 +01:00
let windowpos = lparam as *const WINDOWPOS;
if unsafe { (*windowpos).flags & SWP_NOMOVE != SWP_NOMOVE } {
let physical_position =
unsafe { PhysicalPosition::new((*windowpos).x, (*windowpos).y) };
userdata.send_event(Event::WindowEvent {
window_id: WindowId::from_raw(window as usize),
event: Moved(physical_position),
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
});
}
// This is necessary for us to still get sent WM_SIZE.
result = ProcResult::DefWindowProc(wparam);
2019-06-24 12:14:55 -04:00
},
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
2022-03-07 22:58:12 +01:00
WM_SIZE => {
use crate::event::WindowEvent::SurfaceResized;
2022-03-07 22:58:12 +01:00
let w = super::loword(lparam as u32) as u32;
let h = super::hiword(lparam as u32) as u32;
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
let physical_size = PhysicalSize::new(w, h);
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
let event = Event::WindowEvent {
window_id: WindowId::from_raw(window as usize),
event: SurfaceResized(physical_size),
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
};
{
2022-08-31 18:32:19 +02:00
let mut w = userdata.window_state_lock();
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
// See WindowFlags::MARKER_RETAIN_STATE_ON_SIZE docs for info on why this `if` check
// exists.
if !w.window_flags().contains(WindowFlags::MARKER_RETAIN_STATE_ON_SIZE) {
2022-03-07 22:58:12 +01:00
let maximized = wparam == SIZE_MAXIMIZED as usize;
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
w.set_window_flags_in_place(|f| f.set(WindowFlags::MAXIMIZED, maximized));
}
}
userdata.send_event(event);
result = ProcResult::Value(0);
2019-06-24 12:14:55 -04:00
},
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
WM_SIZING => {
/// Calculate the amount to add to round `value` to the nearest multiple of `increment`.
fn snap_to_nearest_increment_delta(value: i32, increment: i32) -> i32 {
let half_one = increment / 2;
let half_two = increment - half_one;
half_one - (value - half_two) % increment
}
let scale_factor = userdata.window_state_lock().scale_factor;
let Some(inc) = userdata
.window_state_lock()
.surface_resize_increments
.map(|inc| inc.to_physical(scale_factor))
.filter(|inc| inc.width > 0 && inc.height > 0)
else {
result = ProcResult::Value(0);
return;
};
let side = wparam as u32;
// The desired new size of the window, decorations included.
let rect = unsafe { &mut *(lparam as *mut RECT) };
// We need to calculate the dimensions of the window decorations to get the true
// size of the window's contents
let adj_rect = userdata
.window_state_lock()
.window_flags
.adjust_rect(window, *rect)
.unwrap_or(*rect);
let deco_width = rect.left - adj_rect.left + adj_rect.right - rect.right;
let deco_height = rect.top - adj_rect.top + adj_rect.bottom - rect.bottom;
let width = rect.right - rect.left - deco_width;
let height = rect.bottom - rect.top - deco_height;
let mut width_delta = snap_to_nearest_increment_delta(width, inc.width);
let mut height_delta = snap_to_nearest_increment_delta(height, inc.height);
// Windows won't bound check the value of `rect` after we're done here, so we
// have to check manually. If the width/height we snap to would go out of bounds, just
// set it equal to the min/max bound.
let min_size =
userdata.window_state_lock().min_size.map(|size| size.to_physical(scale_factor));
let max_size =
userdata.window_state_lock().max_size.map(|size| size.to_physical(scale_factor));
let final_width = width + width_delta;
let final_height = height + height_delta;
if let Some(min_size) = min_size {
if final_width < min_size.width {
width_delta += min_size.width - final_width;
}
if final_height < min_size.height {
height_delta += min_size.height - final_height;
}
}
if let Some(max_size) = max_size {
if final_width > max_size.width {
width_delta -= final_width - max_size.width;
}
if final_height > max_size.height {
height_delta -= final_height - max_size.height;
}
}
match side {
WMSZ_LEFT | WMSZ_BOTTOMLEFT | WMSZ_TOPLEFT => {
rect.left -= width_delta;
},
WMSZ_RIGHT | WMSZ_BOTTOMRIGHT | WMSZ_TOPRIGHT => {
rect.right += width_delta;
},
_ => {},
}
match side {
WMSZ_TOP | WMSZ_TOPLEFT | WMSZ_TOPRIGHT => {
rect.top -= height_delta;
},
WMSZ_BOTTOM | WMSZ_BOTTOMLEFT | WMSZ_BOTTOMRIGHT => {
rect.bottom += height_delta;
},
_ => {},
}
result = ProcResult::DefWindowProc(wparam);
},
WM_MENUCHAR => {
result = ProcResult::Value((MNC_CLOSE << 16) as isize);
2019-06-24 12:14:55 -04:00
},
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
WM_IME_STARTCOMPOSITION => {
2022-08-31 18:32:19 +02:00
let ime_allowed = userdata.window_state_lock().ime_allowed;
if ime_allowed {
2022-08-31 18:32:19 +02:00
userdata.window_state_lock().ime_state = ImeState::Enabled;
userdata.send_event(Event::WindowEvent {
window_id: WindowId::from_raw(window as usize),
event: WindowEvent::Ime(Ime::Enabled),
});
}
result = ProcResult::DefWindowProc(wparam);
},
WM_IME_COMPOSITION => {
let ime_allowed_and_composing = {
2022-08-31 18:32:19 +02:00
let w = userdata.window_state_lock();
w.ime_allowed && w.ime_state != ImeState::Disabled
};
// Windows Hangul IME sends WM_IME_COMPOSITION after WM_IME_ENDCOMPOSITION, so
// check whether composing.
if ime_allowed_and_composing {
let ime_context = unsafe { ImeContext::current(window) };
if lparam == 0 {
userdata.send_event(Event::WindowEvent {
window_id: WindowId::from_raw(window as usize),
event: WindowEvent::Ime(Ime::Preedit(String::new(), None)),
});
}
// Google Japanese Input and ATOK have both flags, so
// first, receive composing result if exist.
if (lparam as u32 & GCS_RESULTSTR) != 0 {
if let Some(text) = unsafe { ime_context.get_composed_text() } {
2022-08-31 18:32:19 +02:00
userdata.window_state_lock().ime_state = ImeState::Enabled;
userdata.send_event(Event::WindowEvent {
window_id: WindowId::from_raw(window as usize),
event: WindowEvent::Ime(Ime::Preedit(String::new(), None)),
});
userdata.send_event(Event::WindowEvent {
window_id: WindowId::from_raw(window as usize),
event: WindowEvent::Ime(Ime::Commit(text)),
});
}
}
// Next, receive preedit range for next composing if exist.
if (lparam as u32 & GCS_COMPSTR) != 0 {
if let Some((text, first, last)) =
unsafe { ime_context.get_composing_text_and_cursor() }
{
2022-08-31 18:32:19 +02:00
userdata.window_state_lock().ime_state = ImeState::Preedit;
let cursor_range = first.map(|f| (f, last.unwrap_or(f)));
userdata.send_event(Event::WindowEvent {
window_id: WindowId::from_raw(window as usize),
event: WindowEvent::Ime(Ime::Preedit(text, cursor_range)),
});
}
}
}
// Not calling DefWindowProc to hide composing text drawn by IME.
result = ProcResult::Value(0);
},
WM_IME_ENDCOMPOSITION => {
let ime_allowed_or_composing = {
2022-08-31 18:32:19 +02:00
let w = userdata.window_state_lock();
w.ime_allowed || w.ime_state != ImeState::Disabled
};
if ime_allowed_or_composing {
2022-08-31 18:32:19 +02:00
if userdata.window_state_lock().ime_state == ImeState::Preedit {
// Windows Hangul IME sends WM_IME_COMPOSITION after WM_IME_ENDCOMPOSITION, so
// trying receiving composing result and commit if exists.
let ime_context = unsafe { ImeContext::current(window) };
if let Some(text) = unsafe { ime_context.get_composed_text() } {
userdata.send_event(Event::WindowEvent {
window_id: WindowId::from_raw(window as usize),
event: WindowEvent::Ime(Ime::Preedit(String::new(), None)),
});
userdata.send_event(Event::WindowEvent {
window_id: WindowId::from_raw(window as usize),
event: WindowEvent::Ime(Ime::Commit(text)),
});
}
}
2022-08-31 18:32:19 +02:00
userdata.window_state_lock().ime_state = ImeState::Disabled;
userdata.send_event(Event::WindowEvent {
window_id: WindowId::from_raw(window as usize),
event: WindowEvent::Ime(Ime::Disabled),
});
}
result = ProcResult::DefWindowProc(wparam);
},
WM_IME_SETCONTEXT => {
// Hide composing text drawn by IME.
let wparam = wparam & (!ISC_SHOWUICOMPOSITIONWINDOW as usize);
result = ProcResult::DefWindowProc(wparam);
},
// this is necessary for us to maintain minimize/restore state
2022-03-07 22:58:12 +01:00
WM_SYSCOMMAND => {
if wparam == SC_RESTORE as usize {
2022-08-31 18:32:19 +02:00
let mut w = userdata.window_state_lock();
w.set_window_flags_in_place(|f| f.set(WindowFlags::MINIMIZED, false));
}
2022-03-07 22:58:12 +01:00
if wparam == SC_MINIMIZE as usize {
2022-08-31 18:32:19 +02:00
let mut w = userdata.window_state_lock();
w.set_window_flags_in_place(|f| f.set(WindowFlags::MINIMIZED, true));
}
// Send `WindowEvent::Minimized` here if we decide to implement one
2022-03-07 22:58:12 +01:00
if wparam == SC_SCREENSAVE as usize {
2022-08-31 18:32:19 +02:00
let window_state = userdata.window_state_lock();
if window_state.fullscreen.is_some() {
result = ProcResult::Value(0);
return;
}
}
result = ProcResult::DefWindowProc(wparam);
},
2022-03-07 22:58:12 +01:00
WM_MOUSEMOVE => {
use crate::event::WindowEvent::{PointerEntered, PointerLeft, PointerMoved};
use crate::event::{PointerKind, PointerSource};
let x = super::get_x_lparam(lparam as u32) as i32;
let y = super::get_y_lparam(lparam as u32) as i32;
let position = PhysicalPosition::new(x as f64, y as f64);
let cursor_moved;
{
2022-08-31 18:32:19 +02:00
let mut w = userdata.window_state_lock();
let mouse_was_inside_window =
w.mouse.cursor_flags().contains(CursorFlags::IN_WINDOW);
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
match get_pointer_move_kind(window, mouse_was_inside_window, x, y) {
PointerMoveKind::Enter => {
w.mouse
.set_cursor_flags(window, |f| f.set(CursorFlags::IN_WINDOW, true))
.ok();
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
drop(w);
userdata.send_event(Event::WindowEvent {
window_id: WindowId::from_raw(window as usize),
event: PointerEntered {
device_id: None,
primary: true,
position,
kind: PointerKind::Mouse,
},
});
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
// Calling TrackMouseEvent in order to receive mouse leave events.
unsafe {
TrackMouseEvent(&mut TRACKMOUSEEVENT {
cbSize: mem::size_of::<TRACKMOUSEEVENT>() as u32,
dwFlags: TME_LEAVE,
hwndTrack: window,
dwHoverTime: HOVER_DEFAULT,
})
};
},
PointerMoveKind::Leave => {
w.mouse
.set_cursor_flags(window, |f| f.set(CursorFlags::IN_WINDOW, false))
.ok();
drop(w);
userdata.send_event(Event::WindowEvent {
window_id: WindowId::from_raw(window as usize),
event: PointerLeft {
device_id: None,
primary: true,
position: Some(position),
kind: PointerKind::Mouse,
},
});
},
PointerMoveKind::None => drop(w),
}
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
// handle spurious WM_MOUSEMOVE messages
// see https://devblogs.microsoft.com/oldnewthing/20031001-00/?p=42343
// and http://debugandconquer.blogspot.com/2015/08/the-cause-of-spurious-mouse-move.html
let mut w = userdata.window_state_lock();
cursor_moved = w.mouse.last_position != Some(position);
w.mouse.last_position = Some(position);
}
if cursor_moved {
update_modifiers(window, userdata);
Move `ModifiersChanged` variant to `WindowEvent` (#1381) * Move `ModifiersChanged` variant to `WindowEvent` * macos: Fix flags_changed for ModifiersChanged variant move I haven't look too deep at what this does internally, but at least cargo-check is fully happy now. :) Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * macos: Fire a ModifiersChanged event on window_did_resign_key From debugging, I determined that macOS' emission of a flagsChanged around window switching is inconsistent. It is fair to assume, I think, that when the user switches windows, they do not expect their former modifiers state to remain effective; so I think it's best to clear that state by sending a ModifiersChanged(ModifiersState::empty()). Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * windows: Fix build I don't know enough about the code to implement the fix as it is done on this branch, but this commit at least fixes the build. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * windows: Send ModifiersChanged(ModifiersState::empty) on KILLFOCUS Very similar to the changes made in [1], as focus is lost, send an event to the window indicating that the modifiers have been released. It's unclear to me (without a Windows device to test this on) whether this is necessary, but it certainly ensures that unfocused windows will have at least received this event, which is an improvement. [1]: f79f21641a31da3e4039d41be89047cdcc6028f7 Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * macos: Add a hook to update stale modifiers Sometimes, `ViewState` and `event` might have different values for their stored `modifiers` flags. These are internally stored as a bitmask in the latter and an enum in the former. We can check to see if they differ, and if they do, automatically dispatch an event to update consumers of modifier state as well as the stored `state.modifiers`. That's what the hook does. This hook is then called in the key_down, mouse_entered, mouse_exited, mouse_click, scroll_wheel, and pressure_change_with_event callbacks, which each will contain updated modifiers. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * Only call event_mods once when determining whether to update state Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * flags_changed: Memoize window_id collection Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * window_did_resign_key: Remove synthetic ModifiersChanged event We no longer need to emit this event, since we are checking the state of our modifiers before emitting most other events. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * mouse_motion: Add a call to update_potentially_stale_modifiers Now, cover all events (that I can think of, at least) where stale modifiers might affect how user programs behave. Effectively, every human-interface event (keypress, mouse click, keydown, etc.) will cause a ModifiersChanged event to be fired if something has changed. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * key_up: Add a call to update_potentially_stale_modifiers We also want to make sure modifiers state is synchronized here, too. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * mouse_motion: Remove update_potentially_stale_modifiers invocation Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * Retry CI * ViewState: Promote visibility of modifiers to the macos impl This is so that we can interact with the ViewState directly from the WindowDelegate. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * window_delegate: Synthetically set modifiers state to empty on resignKey This logic is implemented similarly on other platforms, so we wish to regain parity here. Originally this behavior was implemented to always fire an event with ModifiersState::empty(), but that was not the best as it was not necessarily correct and could be a duplicate event. This solution is perhaps the most elegant possible to implement the desired behavior of sending a synthetic empty modifiers event when a window loses focus, trading some safety for interoperation between the NSWindowDelegate and the NSView (as the objc runtime must now be consulted in order to acquire access to the ViewState which is "owned" by the NSView). Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * Check for modifiers change in window events * Fix modifier changed on macOS Since the `mouse_entered` function was generating a mouse motion, which updates the modifier state, a modifiers changed event was incorrectly generated. The updating of the modifier state has also been changed to make sure it consistently happens before events that have a modifier state attached to it, without happening on any other event. This of course means that no `CursorMoved` event is generated anymore when the user enters the window without it being focused, however I'd say that is consistent with how winit should behave. * Fix unused variable warning * Move changelog entry into `Unreleased` section Co-authored-by: Freya Gentz <zegentzy@protonmail.com> Co-authored-by: Kristofer Rye <kristofer.rye@gmail.com> Co-authored-by: Christian Duerr <contact@christianduerr.com>
2020-03-06 15:43:55 -07:00
userdata.send_event(Event::WindowEvent {
window_id: WindowId::from_raw(window as usize),
event: PointerMoved {
device_id: None,
primary: true,
position,
source: PointerSource::Mouse,
},
});
}
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
result = ProcResult::Value(0);
2019-06-24 12:14:55 -04:00
},
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
2022-03-07 22:58:12 +01:00
WM_MOUSELEAVE => {
use crate::event::PointerKind::Mouse;
use crate::event::WindowEvent::PointerLeft;
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
{
2022-08-31 18:32:19 +02:00
let mut w = userdata.window_state_lock();
w.mouse.set_cursor_flags(window, |f| f.set(CursorFlags::IN_WINDOW, false)).ok();
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
}
userdata.send_event(Event::WindowEvent {
window_id: WindowId::from_raw(window as usize),
event: PointerLeft { device_id: None, primary: true, position: None, kind: Mouse },
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
});
result = ProcResult::Value(0);
2019-06-24 12:14:55 -04:00
},
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
2022-03-07 22:58:12 +01:00
WM_MOUSEWHEEL => {
use crate::event::MouseScrollDelta::LineDelta;
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
let value = (wparam >> 16) as i16;
2022-03-07 22:58:12 +01:00
let value = value as f32 / WHEEL_DELTA as f32;
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
update_modifiers(window, userdata);
Move `ModifiersChanged` variant to `WindowEvent` (#1381) * Move `ModifiersChanged` variant to `WindowEvent` * macos: Fix flags_changed for ModifiersChanged variant move I haven't look too deep at what this does internally, but at least cargo-check is fully happy now. :) Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * macos: Fire a ModifiersChanged event on window_did_resign_key From debugging, I determined that macOS' emission of a flagsChanged around window switching is inconsistent. It is fair to assume, I think, that when the user switches windows, they do not expect their former modifiers state to remain effective; so I think it's best to clear that state by sending a ModifiersChanged(ModifiersState::empty()). Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * windows: Fix build I don't know enough about the code to implement the fix as it is done on this branch, but this commit at least fixes the build. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * windows: Send ModifiersChanged(ModifiersState::empty) on KILLFOCUS Very similar to the changes made in [1], as focus is lost, send an event to the window indicating that the modifiers have been released. It's unclear to me (without a Windows device to test this on) whether this is necessary, but it certainly ensures that unfocused windows will have at least received this event, which is an improvement. [1]: f79f21641a31da3e4039d41be89047cdcc6028f7 Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * macos: Add a hook to update stale modifiers Sometimes, `ViewState` and `event` might have different values for their stored `modifiers` flags. These are internally stored as a bitmask in the latter and an enum in the former. We can check to see if they differ, and if they do, automatically dispatch an event to update consumers of modifier state as well as the stored `state.modifiers`. That's what the hook does. This hook is then called in the key_down, mouse_entered, mouse_exited, mouse_click, scroll_wheel, and pressure_change_with_event callbacks, which each will contain updated modifiers. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * Only call event_mods once when determining whether to update state Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * flags_changed: Memoize window_id collection Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * window_did_resign_key: Remove synthetic ModifiersChanged event We no longer need to emit this event, since we are checking the state of our modifiers before emitting most other events. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * mouse_motion: Add a call to update_potentially_stale_modifiers Now, cover all events (that I can think of, at least) where stale modifiers might affect how user programs behave. Effectively, every human-interface event (keypress, mouse click, keydown, etc.) will cause a ModifiersChanged event to be fired if something has changed. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * key_up: Add a call to update_potentially_stale_modifiers We also want to make sure modifiers state is synchronized here, too. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * mouse_motion: Remove update_potentially_stale_modifiers invocation Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * Retry CI * ViewState: Promote visibility of modifiers to the macos impl This is so that we can interact with the ViewState directly from the WindowDelegate. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * window_delegate: Synthetically set modifiers state to empty on resignKey This logic is implemented similarly on other platforms, so we wish to regain parity here. Originally this behavior was implemented to always fire an event with ModifiersState::empty(), but that was not the best as it was not necessarily correct and could be a duplicate event. This solution is perhaps the most elegant possible to implement the desired behavior of sending a synthetic empty modifiers event when a window loses focus, trading some safety for interoperation between the NSWindowDelegate and the NSView (as the objc runtime must now be consulted in order to acquire access to the ViewState which is "owned" by the NSView). Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * Check for modifiers change in window events * Fix modifier changed on macOS Since the `mouse_entered` function was generating a mouse motion, which updates the modifier state, a modifiers changed event was incorrectly generated. The updating of the modifier state has also been changed to make sure it consistently happens before events that have a modifier state attached to it, without happening on any other event. This of course means that no `CursorMoved` event is generated anymore when the user enters the window without it being focused, however I'd say that is consistent with how winit should behave. * Fix unused variable warning * Move changelog entry into `Unreleased` section Co-authored-by: Freya Gentz <zegentzy@protonmail.com> Co-authored-by: Kristofer Rye <kristofer.rye@gmail.com> Co-authored-by: Christian Duerr <contact@christianduerr.com>
2020-03-06 15:43:55 -07:00
userdata.send_event(Event::WindowEvent {
window_id: WindowId::from_raw(window as usize),
event: WindowEvent::MouseWheel {
device_id: None,
delta: LineDelta(0.0, value),
phase: TouchPhase::Moved,
},
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
});
result = ProcResult::Value(0);
2019-06-24 12:14:55 -04:00
},
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
2022-03-07 22:58:12 +01:00
WM_MOUSEHWHEEL => {
use crate::event::MouseScrollDelta::LineDelta;
let value = (wparam >> 16) as i16;
let value = -value as f32 / WHEEL_DELTA as f32; // NOTE: inverted! See https://github.com/rust-windowing/winit/pull/2105/
update_modifiers(window, userdata);
Move `ModifiersChanged` variant to `WindowEvent` (#1381) * Move `ModifiersChanged` variant to `WindowEvent` * macos: Fix flags_changed for ModifiersChanged variant move I haven't look too deep at what this does internally, but at least cargo-check is fully happy now. :) Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * macos: Fire a ModifiersChanged event on window_did_resign_key From debugging, I determined that macOS' emission of a flagsChanged around window switching is inconsistent. It is fair to assume, I think, that when the user switches windows, they do not expect their former modifiers state to remain effective; so I think it's best to clear that state by sending a ModifiersChanged(ModifiersState::empty()). Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * windows: Fix build I don't know enough about the code to implement the fix as it is done on this branch, but this commit at least fixes the build. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * windows: Send ModifiersChanged(ModifiersState::empty) on KILLFOCUS Very similar to the changes made in [1], as focus is lost, send an event to the window indicating that the modifiers have been released. It's unclear to me (without a Windows device to test this on) whether this is necessary, but it certainly ensures that unfocused windows will have at least received this event, which is an improvement. [1]: f79f21641a31da3e4039d41be89047cdcc6028f7 Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * macos: Add a hook to update stale modifiers Sometimes, `ViewState` and `event` might have different values for their stored `modifiers` flags. These are internally stored as a bitmask in the latter and an enum in the former. We can check to see if they differ, and if they do, automatically dispatch an event to update consumers of modifier state as well as the stored `state.modifiers`. That's what the hook does. This hook is then called in the key_down, mouse_entered, mouse_exited, mouse_click, scroll_wheel, and pressure_change_with_event callbacks, which each will contain updated modifiers. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * Only call event_mods once when determining whether to update state Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * flags_changed: Memoize window_id collection Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * window_did_resign_key: Remove synthetic ModifiersChanged event We no longer need to emit this event, since we are checking the state of our modifiers before emitting most other events. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * mouse_motion: Add a call to update_potentially_stale_modifiers Now, cover all events (that I can think of, at least) where stale modifiers might affect how user programs behave. Effectively, every human-interface event (keypress, mouse click, keydown, etc.) will cause a ModifiersChanged event to be fired if something has changed. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * key_up: Add a call to update_potentially_stale_modifiers We also want to make sure modifiers state is synchronized here, too. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * mouse_motion: Remove update_potentially_stale_modifiers invocation Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * Retry CI * ViewState: Promote visibility of modifiers to the macos impl This is so that we can interact with the ViewState directly from the WindowDelegate. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * window_delegate: Synthetically set modifiers state to empty on resignKey This logic is implemented similarly on other platforms, so we wish to regain parity here. Originally this behavior was implemented to always fire an event with ModifiersState::empty(), but that was not the best as it was not necessarily correct and could be a duplicate event. This solution is perhaps the most elegant possible to implement the desired behavior of sending a synthetic empty modifiers event when a window loses focus, trading some safety for interoperation between the NSWindowDelegate and the NSView (as the objc runtime must now be consulted in order to acquire access to the ViewState which is "owned" by the NSView). Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * Check for modifiers change in window events * Fix modifier changed on macOS Since the `mouse_entered` function was generating a mouse motion, which updates the modifier state, a modifiers changed event was incorrectly generated. The updating of the modifier state has also been changed to make sure it consistently happens before events that have a modifier state attached to it, without happening on any other event. This of course means that no `CursorMoved` event is generated anymore when the user enters the window without it being focused, however I'd say that is consistent with how winit should behave. * Fix unused variable warning * Move changelog entry into `Unreleased` section Co-authored-by: Freya Gentz <zegentzy@protonmail.com> Co-authored-by: Kristofer Rye <kristofer.rye@gmail.com> Co-authored-by: Christian Duerr <contact@christianduerr.com>
2020-03-06 15:43:55 -07:00
userdata.send_event(Event::WindowEvent {
window_id: WindowId::from_raw(window as usize),
event: WindowEvent::MouseWheel {
device_id: None,
delta: LineDelta(value, 0.0),
phase: TouchPhase::Moved,
},
});
result = ProcResult::Value(0);
2019-06-24 12:14:55 -04:00
},
2022-03-07 22:58:12 +01:00
WM_KEYDOWN | WM_SYSKEYDOWN => {
if msg == WM_SYSKEYDOWN {
result = ProcResult::DefWindowProc(wparam);
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
}
2019-06-24 12:14:55 -04:00
},
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
2022-03-07 22:58:12 +01:00
WM_KEYUP | WM_SYSKEYUP => {
if msg == WM_SYSKEYUP && unsafe { GetMenu(window) != 0 } {
// let Windows handle event if the window has a native menu, a modal event loop
// is started here on Alt key up.
result = ProcResult::DefWindowProc(wparam);
}
2019-06-24 12:14:55 -04:00
},
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
2022-03-07 22:58:12 +01:00
WM_LBUTTONDOWN => {
use crate::event::ElementState::Pressed;
use crate::event::MouseButton::Left;
use crate::event::WindowEvent::PointerButton;
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
unsafe { capture_mouse(window, &mut userdata.window_state_lock()) };
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
update_modifiers(window, userdata);
Move `ModifiersChanged` variant to `WindowEvent` (#1381) * Move `ModifiersChanged` variant to `WindowEvent` * macos: Fix flags_changed for ModifiersChanged variant move I haven't look too deep at what this does internally, but at least cargo-check is fully happy now. :) Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * macos: Fire a ModifiersChanged event on window_did_resign_key From debugging, I determined that macOS' emission of a flagsChanged around window switching is inconsistent. It is fair to assume, I think, that when the user switches windows, they do not expect their former modifiers state to remain effective; so I think it's best to clear that state by sending a ModifiersChanged(ModifiersState::empty()). Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * windows: Fix build I don't know enough about the code to implement the fix as it is done on this branch, but this commit at least fixes the build. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * windows: Send ModifiersChanged(ModifiersState::empty) on KILLFOCUS Very similar to the changes made in [1], as focus is lost, send an event to the window indicating that the modifiers have been released. It's unclear to me (without a Windows device to test this on) whether this is necessary, but it certainly ensures that unfocused windows will have at least received this event, which is an improvement. [1]: f79f21641a31da3e4039d41be89047cdcc6028f7 Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * macos: Add a hook to update stale modifiers Sometimes, `ViewState` and `event` might have different values for their stored `modifiers` flags. These are internally stored as a bitmask in the latter and an enum in the former. We can check to see if they differ, and if they do, automatically dispatch an event to update consumers of modifier state as well as the stored `state.modifiers`. That's what the hook does. This hook is then called in the key_down, mouse_entered, mouse_exited, mouse_click, scroll_wheel, and pressure_change_with_event callbacks, which each will contain updated modifiers. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * Only call event_mods once when determining whether to update state Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * flags_changed: Memoize window_id collection Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * window_did_resign_key: Remove synthetic ModifiersChanged event We no longer need to emit this event, since we are checking the state of our modifiers before emitting most other events. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * mouse_motion: Add a call to update_potentially_stale_modifiers Now, cover all events (that I can think of, at least) where stale modifiers might affect how user programs behave. Effectively, every human-interface event (keypress, mouse click, keydown, etc.) will cause a ModifiersChanged event to be fired if something has changed. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * key_up: Add a call to update_potentially_stale_modifiers We also want to make sure modifiers state is synchronized here, too. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * mouse_motion: Remove update_potentially_stale_modifiers invocation Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * Retry CI * ViewState: Promote visibility of modifiers to the macos impl This is so that we can interact with the ViewState directly from the WindowDelegate. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * window_delegate: Synthetically set modifiers state to empty on resignKey This logic is implemented similarly on other platforms, so we wish to regain parity here. Originally this behavior was implemented to always fire an event with ModifiersState::empty(), but that was not the best as it was not necessarily correct and could be a duplicate event. This solution is perhaps the most elegant possible to implement the desired behavior of sending a synthetic empty modifiers event when a window loses focus, trading some safety for interoperation between the NSWindowDelegate and the NSView (as the objc runtime must now be consulted in order to acquire access to the ViewState which is "owned" by the NSView). Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * Check for modifiers change in window events * Fix modifier changed on macOS Since the `mouse_entered` function was generating a mouse motion, which updates the modifier state, a modifiers changed event was incorrectly generated. The updating of the modifier state has also been changed to make sure it consistently happens before events that have a modifier state attached to it, without happening on any other event. This of course means that no `CursorMoved` event is generated anymore when the user enters the window without it being focused, however I'd say that is consistent with how winit should behave. * Fix unused variable warning * Move changelog entry into `Unreleased` section Co-authored-by: Freya Gentz <zegentzy@protonmail.com> Co-authored-by: Kristofer Rye <kristofer.rye@gmail.com> Co-authored-by: Christian Duerr <contact@christianduerr.com>
2020-03-06 15:43:55 -07:00
let x = super::get_x_lparam(lparam as u32) as i32;
let y = super::get_y_lparam(lparam as u32) as i32;
let position = PhysicalPosition::new(x as f64, y as f64);
userdata.send_event(Event::WindowEvent {
window_id: WindowId::from_raw(window as usize),
event: PointerButton {
device_id: None,
primary: true,
state: Pressed,
position,
button: Left.into(),
},
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
});
result = ProcResult::Value(0);
2019-06-24 12:14:55 -04:00
},
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
2022-03-07 22:58:12 +01:00
WM_LBUTTONUP => {
use crate::event::ElementState::Released;
use crate::event::MouseButton::Left;
use crate::event::WindowEvent::PointerButton;
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
unsafe { release_mouse(userdata.window_state_lock()) };
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
update_modifiers(window, userdata);
Move `ModifiersChanged` variant to `WindowEvent` (#1381) * Move `ModifiersChanged` variant to `WindowEvent` * macos: Fix flags_changed for ModifiersChanged variant move I haven't look too deep at what this does internally, but at least cargo-check is fully happy now. :) Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * macos: Fire a ModifiersChanged event on window_did_resign_key From debugging, I determined that macOS' emission of a flagsChanged around window switching is inconsistent. It is fair to assume, I think, that when the user switches windows, they do not expect their former modifiers state to remain effective; so I think it's best to clear that state by sending a ModifiersChanged(ModifiersState::empty()). Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * windows: Fix build I don't know enough about the code to implement the fix as it is done on this branch, but this commit at least fixes the build. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * windows: Send ModifiersChanged(ModifiersState::empty) on KILLFOCUS Very similar to the changes made in [1], as focus is lost, send an event to the window indicating that the modifiers have been released. It's unclear to me (without a Windows device to test this on) whether this is necessary, but it certainly ensures that unfocused windows will have at least received this event, which is an improvement. [1]: f79f21641a31da3e4039d41be89047cdcc6028f7 Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * macos: Add a hook to update stale modifiers Sometimes, `ViewState` and `event` might have different values for their stored `modifiers` flags. These are internally stored as a bitmask in the latter and an enum in the former. We can check to see if they differ, and if they do, automatically dispatch an event to update consumers of modifier state as well as the stored `state.modifiers`. That's what the hook does. This hook is then called in the key_down, mouse_entered, mouse_exited, mouse_click, scroll_wheel, and pressure_change_with_event callbacks, which each will contain updated modifiers. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * Only call event_mods once when determining whether to update state Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * flags_changed: Memoize window_id collection Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * window_did_resign_key: Remove synthetic ModifiersChanged event We no longer need to emit this event, since we are checking the state of our modifiers before emitting most other events. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * mouse_motion: Add a call to update_potentially_stale_modifiers Now, cover all events (that I can think of, at least) where stale modifiers might affect how user programs behave. Effectively, every human-interface event (keypress, mouse click, keydown, etc.) will cause a ModifiersChanged event to be fired if something has changed. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * key_up: Add a call to update_potentially_stale_modifiers We also want to make sure modifiers state is synchronized here, too. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * mouse_motion: Remove update_potentially_stale_modifiers invocation Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * Retry CI * ViewState: Promote visibility of modifiers to the macos impl This is so that we can interact with the ViewState directly from the WindowDelegate. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * window_delegate: Synthetically set modifiers state to empty on resignKey This logic is implemented similarly on other platforms, so we wish to regain parity here. Originally this behavior was implemented to always fire an event with ModifiersState::empty(), but that was not the best as it was not necessarily correct and could be a duplicate event. This solution is perhaps the most elegant possible to implement the desired behavior of sending a synthetic empty modifiers event when a window loses focus, trading some safety for interoperation between the NSWindowDelegate and the NSView (as the objc runtime must now be consulted in order to acquire access to the ViewState which is "owned" by the NSView). Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * Check for modifiers change in window events * Fix modifier changed on macOS Since the `mouse_entered` function was generating a mouse motion, which updates the modifier state, a modifiers changed event was incorrectly generated. The updating of the modifier state has also been changed to make sure it consistently happens before events that have a modifier state attached to it, without happening on any other event. This of course means that no `CursorMoved` event is generated anymore when the user enters the window without it being focused, however I'd say that is consistent with how winit should behave. * Fix unused variable warning * Move changelog entry into `Unreleased` section Co-authored-by: Freya Gentz <zegentzy@protonmail.com> Co-authored-by: Kristofer Rye <kristofer.rye@gmail.com> Co-authored-by: Christian Duerr <contact@christianduerr.com>
2020-03-06 15:43:55 -07:00
let x = super::get_x_lparam(lparam as u32) as i32;
let y = super::get_y_lparam(lparam as u32) as i32;
let position = PhysicalPosition::new(x as f64, y as f64);
userdata.send_event(Event::WindowEvent {
window_id: WindowId::from_raw(window as usize),
event: PointerButton {
device_id: None,
primary: true,
state: Released,
position,
button: Left.into(),
},
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
});
result = ProcResult::Value(0);
2019-06-24 12:14:55 -04:00
},
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
2022-03-07 22:58:12 +01:00
WM_RBUTTONDOWN => {
use crate::event::ElementState::Pressed;
use crate::event::MouseButton::Right;
use crate::event::WindowEvent::PointerButton;
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
unsafe { capture_mouse(window, &mut userdata.window_state_lock()) };
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
update_modifiers(window, userdata);
Move `ModifiersChanged` variant to `WindowEvent` (#1381) * Move `ModifiersChanged` variant to `WindowEvent` * macos: Fix flags_changed for ModifiersChanged variant move I haven't look too deep at what this does internally, but at least cargo-check is fully happy now. :) Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * macos: Fire a ModifiersChanged event on window_did_resign_key From debugging, I determined that macOS' emission of a flagsChanged around window switching is inconsistent. It is fair to assume, I think, that when the user switches windows, they do not expect their former modifiers state to remain effective; so I think it's best to clear that state by sending a ModifiersChanged(ModifiersState::empty()). Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * windows: Fix build I don't know enough about the code to implement the fix as it is done on this branch, but this commit at least fixes the build. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * windows: Send ModifiersChanged(ModifiersState::empty) on KILLFOCUS Very similar to the changes made in [1], as focus is lost, send an event to the window indicating that the modifiers have been released. It's unclear to me (without a Windows device to test this on) whether this is necessary, but it certainly ensures that unfocused windows will have at least received this event, which is an improvement. [1]: f79f21641a31da3e4039d41be89047cdcc6028f7 Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * macos: Add a hook to update stale modifiers Sometimes, `ViewState` and `event` might have different values for their stored `modifiers` flags. These are internally stored as a bitmask in the latter and an enum in the former. We can check to see if they differ, and if they do, automatically dispatch an event to update consumers of modifier state as well as the stored `state.modifiers`. That's what the hook does. This hook is then called in the key_down, mouse_entered, mouse_exited, mouse_click, scroll_wheel, and pressure_change_with_event callbacks, which each will contain updated modifiers. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * Only call event_mods once when determining whether to update state Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * flags_changed: Memoize window_id collection Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * window_did_resign_key: Remove synthetic ModifiersChanged event We no longer need to emit this event, since we are checking the state of our modifiers before emitting most other events. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * mouse_motion: Add a call to update_potentially_stale_modifiers Now, cover all events (that I can think of, at least) where stale modifiers might affect how user programs behave. Effectively, every human-interface event (keypress, mouse click, keydown, etc.) will cause a ModifiersChanged event to be fired if something has changed. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * key_up: Add a call to update_potentially_stale_modifiers We also want to make sure modifiers state is synchronized here, too. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * mouse_motion: Remove update_potentially_stale_modifiers invocation Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * Retry CI * ViewState: Promote visibility of modifiers to the macos impl This is so that we can interact with the ViewState directly from the WindowDelegate. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * window_delegate: Synthetically set modifiers state to empty on resignKey This logic is implemented similarly on other platforms, so we wish to regain parity here. Originally this behavior was implemented to always fire an event with ModifiersState::empty(), but that was not the best as it was not necessarily correct and could be a duplicate event. This solution is perhaps the most elegant possible to implement the desired behavior of sending a synthetic empty modifiers event when a window loses focus, trading some safety for interoperation between the NSWindowDelegate and the NSView (as the objc runtime must now be consulted in order to acquire access to the ViewState which is "owned" by the NSView). Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * Check for modifiers change in window events * Fix modifier changed on macOS Since the `mouse_entered` function was generating a mouse motion, which updates the modifier state, a modifiers changed event was incorrectly generated. The updating of the modifier state has also been changed to make sure it consistently happens before events that have a modifier state attached to it, without happening on any other event. This of course means that no `CursorMoved` event is generated anymore when the user enters the window without it being focused, however I'd say that is consistent with how winit should behave. * Fix unused variable warning * Move changelog entry into `Unreleased` section Co-authored-by: Freya Gentz <zegentzy@protonmail.com> Co-authored-by: Kristofer Rye <kristofer.rye@gmail.com> Co-authored-by: Christian Duerr <contact@christianduerr.com>
2020-03-06 15:43:55 -07:00
let x = super::get_x_lparam(lparam as u32) as i32;
let y = super::get_y_lparam(lparam as u32) as i32;
let position = PhysicalPosition::new(x as f64, y as f64);
userdata.send_event(Event::WindowEvent {
window_id: WindowId::from_raw(window as usize),
event: PointerButton {
device_id: None,
primary: true,
state: Pressed,
position,
button: Right.into(),
},
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
});
result = ProcResult::Value(0);
2019-06-24 12:14:55 -04:00
},
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
2022-03-07 22:58:12 +01:00
WM_RBUTTONUP => {
use crate::event::ElementState::Released;
use crate::event::MouseButton::Right;
use crate::event::WindowEvent::PointerButton;
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
unsafe { release_mouse(userdata.window_state_lock()) };
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
update_modifiers(window, userdata);
Move `ModifiersChanged` variant to `WindowEvent` (#1381) * Move `ModifiersChanged` variant to `WindowEvent` * macos: Fix flags_changed for ModifiersChanged variant move I haven't look too deep at what this does internally, but at least cargo-check is fully happy now. :) Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * macos: Fire a ModifiersChanged event on window_did_resign_key From debugging, I determined that macOS' emission of a flagsChanged around window switching is inconsistent. It is fair to assume, I think, that when the user switches windows, they do not expect their former modifiers state to remain effective; so I think it's best to clear that state by sending a ModifiersChanged(ModifiersState::empty()). Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * windows: Fix build I don't know enough about the code to implement the fix as it is done on this branch, but this commit at least fixes the build. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * windows: Send ModifiersChanged(ModifiersState::empty) on KILLFOCUS Very similar to the changes made in [1], as focus is lost, send an event to the window indicating that the modifiers have been released. It's unclear to me (without a Windows device to test this on) whether this is necessary, but it certainly ensures that unfocused windows will have at least received this event, which is an improvement. [1]: f79f21641a31da3e4039d41be89047cdcc6028f7 Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * macos: Add a hook to update stale modifiers Sometimes, `ViewState` and `event` might have different values for their stored `modifiers` flags. These are internally stored as a bitmask in the latter and an enum in the former. We can check to see if they differ, and if they do, automatically dispatch an event to update consumers of modifier state as well as the stored `state.modifiers`. That's what the hook does. This hook is then called in the key_down, mouse_entered, mouse_exited, mouse_click, scroll_wheel, and pressure_change_with_event callbacks, which each will contain updated modifiers. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * Only call event_mods once when determining whether to update state Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * flags_changed: Memoize window_id collection Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * window_did_resign_key: Remove synthetic ModifiersChanged event We no longer need to emit this event, since we are checking the state of our modifiers before emitting most other events. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * mouse_motion: Add a call to update_potentially_stale_modifiers Now, cover all events (that I can think of, at least) where stale modifiers might affect how user programs behave. Effectively, every human-interface event (keypress, mouse click, keydown, etc.) will cause a ModifiersChanged event to be fired if something has changed. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * key_up: Add a call to update_potentially_stale_modifiers We also want to make sure modifiers state is synchronized here, too. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * mouse_motion: Remove update_potentially_stale_modifiers invocation Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * Retry CI * ViewState: Promote visibility of modifiers to the macos impl This is so that we can interact with the ViewState directly from the WindowDelegate. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * window_delegate: Synthetically set modifiers state to empty on resignKey This logic is implemented similarly on other platforms, so we wish to regain parity here. Originally this behavior was implemented to always fire an event with ModifiersState::empty(), but that was not the best as it was not necessarily correct and could be a duplicate event. This solution is perhaps the most elegant possible to implement the desired behavior of sending a synthetic empty modifiers event when a window loses focus, trading some safety for interoperation between the NSWindowDelegate and the NSView (as the objc runtime must now be consulted in order to acquire access to the ViewState which is "owned" by the NSView). Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * Check for modifiers change in window events * Fix modifier changed on macOS Since the `mouse_entered` function was generating a mouse motion, which updates the modifier state, a modifiers changed event was incorrectly generated. The updating of the modifier state has also been changed to make sure it consistently happens before events that have a modifier state attached to it, without happening on any other event. This of course means that no `CursorMoved` event is generated anymore when the user enters the window without it being focused, however I'd say that is consistent with how winit should behave. * Fix unused variable warning * Move changelog entry into `Unreleased` section Co-authored-by: Freya Gentz <zegentzy@protonmail.com> Co-authored-by: Kristofer Rye <kristofer.rye@gmail.com> Co-authored-by: Christian Duerr <contact@christianduerr.com>
2020-03-06 15:43:55 -07:00
let x = super::get_x_lparam(lparam as u32) as i32;
let y = super::get_y_lparam(lparam as u32) as i32;
let position = PhysicalPosition::new(x as f64, y as f64);
userdata.send_event(Event::WindowEvent {
window_id: WindowId::from_raw(window as usize),
event: PointerButton {
device_id: None,
primary: true,
state: Released,
position,
button: Right.into(),
},
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
});
result = ProcResult::Value(0);
2019-06-24 12:14:55 -04:00
},
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
2022-03-07 22:58:12 +01:00
WM_MBUTTONDOWN => {
use crate::event::ElementState::Pressed;
use crate::event::MouseButton::Middle;
use crate::event::WindowEvent::PointerButton;
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
unsafe { capture_mouse(window, &mut userdata.window_state_lock()) };
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
update_modifiers(window, userdata);
Move `ModifiersChanged` variant to `WindowEvent` (#1381) * Move `ModifiersChanged` variant to `WindowEvent` * macos: Fix flags_changed for ModifiersChanged variant move I haven't look too deep at what this does internally, but at least cargo-check is fully happy now. :) Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * macos: Fire a ModifiersChanged event on window_did_resign_key From debugging, I determined that macOS' emission of a flagsChanged around window switching is inconsistent. It is fair to assume, I think, that when the user switches windows, they do not expect their former modifiers state to remain effective; so I think it's best to clear that state by sending a ModifiersChanged(ModifiersState::empty()). Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * windows: Fix build I don't know enough about the code to implement the fix as it is done on this branch, but this commit at least fixes the build. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * windows: Send ModifiersChanged(ModifiersState::empty) on KILLFOCUS Very similar to the changes made in [1], as focus is lost, send an event to the window indicating that the modifiers have been released. It's unclear to me (without a Windows device to test this on) whether this is necessary, but it certainly ensures that unfocused windows will have at least received this event, which is an improvement. [1]: f79f21641a31da3e4039d41be89047cdcc6028f7 Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * macos: Add a hook to update stale modifiers Sometimes, `ViewState` and `event` might have different values for their stored `modifiers` flags. These are internally stored as a bitmask in the latter and an enum in the former. We can check to see if they differ, and if they do, automatically dispatch an event to update consumers of modifier state as well as the stored `state.modifiers`. That's what the hook does. This hook is then called in the key_down, mouse_entered, mouse_exited, mouse_click, scroll_wheel, and pressure_change_with_event callbacks, which each will contain updated modifiers. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * Only call event_mods once when determining whether to update state Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * flags_changed: Memoize window_id collection Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * window_did_resign_key: Remove synthetic ModifiersChanged event We no longer need to emit this event, since we are checking the state of our modifiers before emitting most other events. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * mouse_motion: Add a call to update_potentially_stale_modifiers Now, cover all events (that I can think of, at least) where stale modifiers might affect how user programs behave. Effectively, every human-interface event (keypress, mouse click, keydown, etc.) will cause a ModifiersChanged event to be fired if something has changed. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * key_up: Add a call to update_potentially_stale_modifiers We also want to make sure modifiers state is synchronized here, too. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * mouse_motion: Remove update_potentially_stale_modifiers invocation Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * Retry CI * ViewState: Promote visibility of modifiers to the macos impl This is so that we can interact with the ViewState directly from the WindowDelegate. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * window_delegate: Synthetically set modifiers state to empty on resignKey This logic is implemented similarly on other platforms, so we wish to regain parity here. Originally this behavior was implemented to always fire an event with ModifiersState::empty(), but that was not the best as it was not necessarily correct and could be a duplicate event. This solution is perhaps the most elegant possible to implement the desired behavior of sending a synthetic empty modifiers event when a window loses focus, trading some safety for interoperation between the NSWindowDelegate and the NSView (as the objc runtime must now be consulted in order to acquire access to the ViewState which is "owned" by the NSView). Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * Check for modifiers change in window events * Fix modifier changed on macOS Since the `mouse_entered` function was generating a mouse motion, which updates the modifier state, a modifiers changed event was incorrectly generated. The updating of the modifier state has also been changed to make sure it consistently happens before events that have a modifier state attached to it, without happening on any other event. This of course means that no `CursorMoved` event is generated anymore when the user enters the window without it being focused, however I'd say that is consistent with how winit should behave. * Fix unused variable warning * Move changelog entry into `Unreleased` section Co-authored-by: Freya Gentz <zegentzy@protonmail.com> Co-authored-by: Kristofer Rye <kristofer.rye@gmail.com> Co-authored-by: Christian Duerr <contact@christianduerr.com>
2020-03-06 15:43:55 -07:00
let x = super::get_x_lparam(lparam as u32) as i32;
let y = super::get_y_lparam(lparam as u32) as i32;
let position = PhysicalPosition::new(x as f64, y as f64);
userdata.send_event(Event::WindowEvent {
window_id: WindowId::from_raw(window as usize),
event: PointerButton {
device_id: None,
primary: true,
state: Pressed,
position,
button: Middle.into(),
},
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
});
result = ProcResult::Value(0);
2019-06-24 12:14:55 -04:00
},
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
2022-03-07 22:58:12 +01:00
WM_MBUTTONUP => {
use crate::event::ElementState::Released;
use crate::event::MouseButton::Middle;
use crate::event::WindowEvent::PointerButton;
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
unsafe { release_mouse(userdata.window_state_lock()) };
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
update_modifiers(window, userdata);
Move `ModifiersChanged` variant to `WindowEvent` (#1381) * Move `ModifiersChanged` variant to `WindowEvent` * macos: Fix flags_changed for ModifiersChanged variant move I haven't look too deep at what this does internally, but at least cargo-check is fully happy now. :) Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * macos: Fire a ModifiersChanged event on window_did_resign_key From debugging, I determined that macOS' emission of a flagsChanged around window switching is inconsistent. It is fair to assume, I think, that when the user switches windows, they do not expect their former modifiers state to remain effective; so I think it's best to clear that state by sending a ModifiersChanged(ModifiersState::empty()). Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * windows: Fix build I don't know enough about the code to implement the fix as it is done on this branch, but this commit at least fixes the build. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * windows: Send ModifiersChanged(ModifiersState::empty) on KILLFOCUS Very similar to the changes made in [1], as focus is lost, send an event to the window indicating that the modifiers have been released. It's unclear to me (without a Windows device to test this on) whether this is necessary, but it certainly ensures that unfocused windows will have at least received this event, which is an improvement. [1]: f79f21641a31da3e4039d41be89047cdcc6028f7 Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * macos: Add a hook to update stale modifiers Sometimes, `ViewState` and `event` might have different values for their stored `modifiers` flags. These are internally stored as a bitmask in the latter and an enum in the former. We can check to see if they differ, and if they do, automatically dispatch an event to update consumers of modifier state as well as the stored `state.modifiers`. That's what the hook does. This hook is then called in the key_down, mouse_entered, mouse_exited, mouse_click, scroll_wheel, and pressure_change_with_event callbacks, which each will contain updated modifiers. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * Only call event_mods once when determining whether to update state Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * flags_changed: Memoize window_id collection Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * window_did_resign_key: Remove synthetic ModifiersChanged event We no longer need to emit this event, since we are checking the state of our modifiers before emitting most other events. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * mouse_motion: Add a call to update_potentially_stale_modifiers Now, cover all events (that I can think of, at least) where stale modifiers might affect how user programs behave. Effectively, every human-interface event (keypress, mouse click, keydown, etc.) will cause a ModifiersChanged event to be fired if something has changed. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * key_up: Add a call to update_potentially_stale_modifiers We also want to make sure modifiers state is synchronized here, too. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * mouse_motion: Remove update_potentially_stale_modifiers invocation Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * Retry CI * ViewState: Promote visibility of modifiers to the macos impl This is so that we can interact with the ViewState directly from the WindowDelegate. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * window_delegate: Synthetically set modifiers state to empty on resignKey This logic is implemented similarly on other platforms, so we wish to regain parity here. Originally this behavior was implemented to always fire an event with ModifiersState::empty(), but that was not the best as it was not necessarily correct and could be a duplicate event. This solution is perhaps the most elegant possible to implement the desired behavior of sending a synthetic empty modifiers event when a window loses focus, trading some safety for interoperation between the NSWindowDelegate and the NSView (as the objc runtime must now be consulted in order to acquire access to the ViewState which is "owned" by the NSView). Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * Check for modifiers change in window events * Fix modifier changed on macOS Since the `mouse_entered` function was generating a mouse motion, which updates the modifier state, a modifiers changed event was incorrectly generated. The updating of the modifier state has also been changed to make sure it consistently happens before events that have a modifier state attached to it, without happening on any other event. This of course means that no `CursorMoved` event is generated anymore when the user enters the window without it being focused, however I'd say that is consistent with how winit should behave. * Fix unused variable warning * Move changelog entry into `Unreleased` section Co-authored-by: Freya Gentz <zegentzy@protonmail.com> Co-authored-by: Kristofer Rye <kristofer.rye@gmail.com> Co-authored-by: Christian Duerr <contact@christianduerr.com>
2020-03-06 15:43:55 -07:00
let x = super::get_x_lparam(lparam as u32) as i32;
let y = super::get_y_lparam(lparam as u32) as i32;
let position = PhysicalPosition::new(x as f64, y as f64);
userdata.send_event(Event::WindowEvent {
window_id: WindowId::from_raw(window as usize),
event: PointerButton {
device_id: None,
primary: true,
state: Released,
position,
button: Middle.into(),
},
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
});
result = ProcResult::Value(0);
2019-06-24 12:14:55 -04:00
},
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
2022-03-07 22:58:12 +01:00
WM_XBUTTONDOWN => {
use crate::event::ElementState::Pressed;
use crate::event::MouseButton::{Back, Forward, Other};
use crate::event::WindowEvent::PointerButton;
2022-03-07 22:58:12 +01:00
let xbutton = super::get_xbutton_wparam(wparam as u32);
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
unsafe { capture_mouse(window, &mut userdata.window_state_lock()) };
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
update_modifiers(window, userdata);
Move `ModifiersChanged` variant to `WindowEvent` (#1381) * Move `ModifiersChanged` variant to `WindowEvent` * macos: Fix flags_changed for ModifiersChanged variant move I haven't look too deep at what this does internally, but at least cargo-check is fully happy now. :) Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * macos: Fire a ModifiersChanged event on window_did_resign_key From debugging, I determined that macOS' emission of a flagsChanged around window switching is inconsistent. It is fair to assume, I think, that when the user switches windows, they do not expect their former modifiers state to remain effective; so I think it's best to clear that state by sending a ModifiersChanged(ModifiersState::empty()). Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * windows: Fix build I don't know enough about the code to implement the fix as it is done on this branch, but this commit at least fixes the build. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * windows: Send ModifiersChanged(ModifiersState::empty) on KILLFOCUS Very similar to the changes made in [1], as focus is lost, send an event to the window indicating that the modifiers have been released. It's unclear to me (without a Windows device to test this on) whether this is necessary, but it certainly ensures that unfocused windows will have at least received this event, which is an improvement. [1]: f79f21641a31da3e4039d41be89047cdcc6028f7 Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * macos: Add a hook to update stale modifiers Sometimes, `ViewState` and `event` might have different values for their stored `modifiers` flags. These are internally stored as a bitmask in the latter and an enum in the former. We can check to see if they differ, and if they do, automatically dispatch an event to update consumers of modifier state as well as the stored `state.modifiers`. That's what the hook does. This hook is then called in the key_down, mouse_entered, mouse_exited, mouse_click, scroll_wheel, and pressure_change_with_event callbacks, which each will contain updated modifiers. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * Only call event_mods once when determining whether to update state Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * flags_changed: Memoize window_id collection Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * window_did_resign_key: Remove synthetic ModifiersChanged event We no longer need to emit this event, since we are checking the state of our modifiers before emitting most other events. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * mouse_motion: Add a call to update_potentially_stale_modifiers Now, cover all events (that I can think of, at least) where stale modifiers might affect how user programs behave. Effectively, every human-interface event (keypress, mouse click, keydown, etc.) will cause a ModifiersChanged event to be fired if something has changed. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * key_up: Add a call to update_potentially_stale_modifiers We also want to make sure modifiers state is synchronized here, too. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * mouse_motion: Remove update_potentially_stale_modifiers invocation Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * Retry CI * ViewState: Promote visibility of modifiers to the macos impl This is so that we can interact with the ViewState directly from the WindowDelegate. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * window_delegate: Synthetically set modifiers state to empty on resignKey This logic is implemented similarly on other platforms, so we wish to regain parity here. Originally this behavior was implemented to always fire an event with ModifiersState::empty(), but that was not the best as it was not necessarily correct and could be a duplicate event. This solution is perhaps the most elegant possible to implement the desired behavior of sending a synthetic empty modifiers event when a window loses focus, trading some safety for interoperation between the NSWindowDelegate and the NSView (as the objc runtime must now be consulted in order to acquire access to the ViewState which is "owned" by the NSView). Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * Check for modifiers change in window events * Fix modifier changed on macOS Since the `mouse_entered` function was generating a mouse motion, which updates the modifier state, a modifiers changed event was incorrectly generated. The updating of the modifier state has also been changed to make sure it consistently happens before events that have a modifier state attached to it, without happening on any other event. This of course means that no `CursorMoved` event is generated anymore when the user enters the window without it being focused, however I'd say that is consistent with how winit should behave. * Fix unused variable warning * Move changelog entry into `Unreleased` section Co-authored-by: Freya Gentz <zegentzy@protonmail.com> Co-authored-by: Kristofer Rye <kristofer.rye@gmail.com> Co-authored-by: Christian Duerr <contact@christianduerr.com>
2020-03-06 15:43:55 -07:00
let x = super::get_x_lparam(lparam as u32) as i32;
let y = super::get_y_lparam(lparam as u32) as i32;
let position = PhysicalPosition::new(x as f64, y as f64);
userdata.send_event(Event::WindowEvent {
window_id: WindowId::from_raw(window as usize),
event: PointerButton {
device_id: None,
primary: true,
state: Pressed,
position,
button: match xbutton {
1 => Back,
2 => Forward,
_ => Other(xbutton),
}
.into(),
},
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
});
result = ProcResult::Value(0);
2019-06-24 12:14:55 -04:00
},
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
2022-03-07 22:58:12 +01:00
WM_XBUTTONUP => {
use crate::event::ElementState::Released;
use crate::event::MouseButton::{Back, Forward, Other};
use crate::event::WindowEvent::PointerButton;
2022-03-07 22:58:12 +01:00
let xbutton = super::get_xbutton_wparam(wparam as u32);
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
unsafe { release_mouse(userdata.window_state_lock()) };
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
update_modifiers(window, userdata);
Move `ModifiersChanged` variant to `WindowEvent` (#1381) * Move `ModifiersChanged` variant to `WindowEvent` * macos: Fix flags_changed for ModifiersChanged variant move I haven't look too deep at what this does internally, but at least cargo-check is fully happy now. :) Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * macos: Fire a ModifiersChanged event on window_did_resign_key From debugging, I determined that macOS' emission of a flagsChanged around window switching is inconsistent. It is fair to assume, I think, that when the user switches windows, they do not expect their former modifiers state to remain effective; so I think it's best to clear that state by sending a ModifiersChanged(ModifiersState::empty()). Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * windows: Fix build I don't know enough about the code to implement the fix as it is done on this branch, but this commit at least fixes the build. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * windows: Send ModifiersChanged(ModifiersState::empty) on KILLFOCUS Very similar to the changes made in [1], as focus is lost, send an event to the window indicating that the modifiers have been released. It's unclear to me (without a Windows device to test this on) whether this is necessary, but it certainly ensures that unfocused windows will have at least received this event, which is an improvement. [1]: f79f21641a31da3e4039d41be89047cdcc6028f7 Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * macos: Add a hook to update stale modifiers Sometimes, `ViewState` and `event` might have different values for their stored `modifiers` flags. These are internally stored as a bitmask in the latter and an enum in the former. We can check to see if they differ, and if they do, automatically dispatch an event to update consumers of modifier state as well as the stored `state.modifiers`. That's what the hook does. This hook is then called in the key_down, mouse_entered, mouse_exited, mouse_click, scroll_wheel, and pressure_change_with_event callbacks, which each will contain updated modifiers. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * Only call event_mods once when determining whether to update state Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * flags_changed: Memoize window_id collection Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * window_did_resign_key: Remove synthetic ModifiersChanged event We no longer need to emit this event, since we are checking the state of our modifiers before emitting most other events. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * mouse_motion: Add a call to update_potentially_stale_modifiers Now, cover all events (that I can think of, at least) where stale modifiers might affect how user programs behave. Effectively, every human-interface event (keypress, mouse click, keydown, etc.) will cause a ModifiersChanged event to be fired if something has changed. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * key_up: Add a call to update_potentially_stale_modifiers We also want to make sure modifiers state is synchronized here, too. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * mouse_motion: Remove update_potentially_stale_modifiers invocation Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * Retry CI * ViewState: Promote visibility of modifiers to the macos impl This is so that we can interact with the ViewState directly from the WindowDelegate. Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * window_delegate: Synthetically set modifiers state to empty on resignKey This logic is implemented similarly on other platforms, so we wish to regain parity here. Originally this behavior was implemented to always fire an event with ModifiersState::empty(), but that was not the best as it was not necessarily correct and could be a duplicate event. This solution is perhaps the most elegant possible to implement the desired behavior of sending a synthetic empty modifiers event when a window loses focus, trading some safety for interoperation between the NSWindowDelegate and the NSView (as the objc runtime must now be consulted in order to acquire access to the ViewState which is "owned" by the NSView). Signed-off-by: Kristofer Rye <kristofer.rye@gmail.com> * Check for modifiers change in window events * Fix modifier changed on macOS Since the `mouse_entered` function was generating a mouse motion, which updates the modifier state, a modifiers changed event was incorrectly generated. The updating of the modifier state has also been changed to make sure it consistently happens before events that have a modifier state attached to it, without happening on any other event. This of course means that no `CursorMoved` event is generated anymore when the user enters the window without it being focused, however I'd say that is consistent with how winit should behave. * Fix unused variable warning * Move changelog entry into `Unreleased` section Co-authored-by: Freya Gentz <zegentzy@protonmail.com> Co-authored-by: Kristofer Rye <kristofer.rye@gmail.com> Co-authored-by: Christian Duerr <contact@christianduerr.com>
2020-03-06 15:43:55 -07:00
let x = super::get_x_lparam(lparam as u32) as i32;
let y = super::get_y_lparam(lparam as u32) as i32;
let position = PhysicalPosition::new(x as f64, y as f64);
userdata.send_event(Event::WindowEvent {
window_id: WindowId::from_raw(window as usize),
event: PointerButton {
device_id: None,
primary: true,
state: Released,
position,
button: match xbutton {
1 => Back,
2 => Forward,
_ => Other(xbutton),
}
.into(),
},
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
});
result = ProcResult::Value(0);
2019-06-24 12:14:55 -04:00
},
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
2022-03-07 22:58:12 +01:00
WM_CAPTURECHANGED => {
// lparam here is a handle to the window which is gaining mouse capture.
// If it is the same as our window, then we're essentially retaining the capture. This
// can happen if `SetCapture` is called on our window when it already has the mouse
// capture.
2022-03-07 22:58:12 +01:00
if lparam != window {
2022-08-31 18:32:19 +02:00
userdata.window_state_lock().mouse.capture_count = 0;
}
result = ProcResult::Value(0);
},
2022-03-07 22:58:12 +01:00
WM_TOUCH => {
use crate::event::ButtonSource::Touch;
use crate::event::ElementState::{Pressed, Released};
use crate::event::{PointerKind, PointerSource};
2022-03-07 22:58:12 +01:00
let pcount = super::loword(wparam as u32) as usize;
let mut inputs = Vec::with_capacity(pcount);
2022-03-07 22:58:12 +01:00
let htouch = lparam;
if unsafe {
GetTouchInputInfo(
htouch,
pcount as u32,
inputs.as_mut_ptr(),
mem::size_of::<TOUCHINPUT>() as i32,
) > 0
} {
unsafe { inputs.set_len(pcount) };
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
for input in &inputs {
let mut position = POINT { x: input.x / 100, y: input.y / 100 };
if unsafe { ScreenToClient(window, &mut position) } == false.into() {
continue;
}
let x = position.x as f64 + (input.x % 100) as f64 / 100f64;
let y = position.y as f64 + (input.y % 100) as f64 / 100f64;
let position = PhysicalPosition::new(x, y);
let window_id = WindowId::from_raw(window as usize);
let finger_id = FingerId::from_raw(input.dwID as usize);
let primary = util::has_flag(input.dwFlags, TOUCHEVENTF_PRIMARY);
if util::has_flag(input.dwFlags, TOUCHEVENTF_DOWN) {
userdata.send_event(Event::WindowEvent {
window_id,
event: WindowEvent::PointerEntered {
device_id: None,
primary,
position,
kind: PointerKind::Touch(finger_id),
},
});
userdata.send_event(Event::WindowEvent {
window_id,
event: WindowEvent::PointerButton {
device_id: None,
primary,
state: Pressed,
position,
button: Touch { finger_id, force: None },
},
});
} else if util::has_flag(input.dwFlags, TOUCHEVENTF_UP) {
userdata.send_event(Event::WindowEvent {
window_id,
event: WindowEvent::PointerButton {
device_id: None,
primary,
state: Released,
position,
button: Touch { finger_id, force: None },
},
});
userdata.send_event(Event::WindowEvent {
window_id,
event: WindowEvent::PointerLeft {
device_id: None,
primary,
position: Some(position),
kind: PointerKind::Touch(finger_id),
},
});
} else if util::has_flag(input.dwFlags, TOUCHEVENTF_MOVE) {
userdata.send_event(Event::WindowEvent {
window_id,
event: WindowEvent::PointerMoved {
device_id: None,
primary,
position,
source: PointerSource::Touch { finger_id, force: None },
},
});
} else {
continue;
}
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
}
}
unsafe { CloseTouchInputHandle(htouch) };
result = ProcResult::Value(0);
2019-06-24 12:14:55 -04:00
},
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
2022-03-07 22:58:12 +01:00
WM_POINTERDOWN | WM_POINTERUPDATE | WM_POINTERUP => {
use crate::event::ElementState::{Pressed, Released};
use crate::event::{ButtonSource, PointerKind, PointerSource};
if let (
Some(GetPointerFrameInfoHistory),
Some(SkipPointerFrameMessages),
Some(GetPointerDeviceRects),
) = (
*util::GET_POINTER_FRAME_INFO_HISTORY,
*util::SKIP_POINTER_FRAME_MESSAGES,
*util::GET_POINTER_DEVICE_RECTS,
) {
2022-03-07 22:58:12 +01:00
let pointer_id = super::loword(wparam as u32) as u32;
let mut entries_count = 0u32;
let mut pointers_count = 0u32;
if unsafe {
GetPointerFrameInfoHistory(
pointer_id,
&mut entries_count,
&mut pointers_count,
ptr::null_mut(),
)
} == false.into()
{
result = ProcResult::Value(0);
return;
}
let pointer_info_count = (entries_count * pointers_count) as usize;
let mut pointer_infos = Vec::with_capacity(pointer_info_count);
if unsafe {
GetPointerFrameInfoHistory(
pointer_id,
&mut entries_count,
&mut pointers_count,
pointer_infos.as_mut_ptr(),
)
} == false.into()
{
result = ProcResult::Value(0);
return;
}
unsafe { pointer_infos.set_len(pointer_info_count) };
// https://docs.microsoft.com/en-us/windows/desktop/api/winuser/nf-winuser-getpointerframeinfohistory
// The information retrieved appears in reverse chronological order, with the most
// recent entry in the first row of the returned array
for pointer_info in pointer_infos.iter().rev() {
let mut device_rect = mem::MaybeUninit::uninit();
let mut display_rect = mem::MaybeUninit::uninit();
if unsafe {
GetPointerDeviceRects(
pointer_info.sourceDevice,
device_rect.as_mut_ptr(),
display_rect.as_mut_ptr(),
)
} == false.into()
{
continue;
}
let device_rect = unsafe { device_rect.assume_init() };
let display_rect = unsafe { display_rect.assume_init() };
// For the most precise himetric to pixel conversion we calculate the ratio
// between the resolution of the display device (pixel) and
// the touch device (himetric).
let himetric_to_pixel_ratio_x = (display_rect.right - display_rect.left) as f64
/ (device_rect.right - device_rect.left) as f64;
let himetric_to_pixel_ratio_y = (display_rect.bottom - display_rect.top) as f64
/ (device_rect.bottom - device_rect.top) as f64;
// ptHimetricLocation's origin is 0,0 even on multi-monitor setups.
// On multi-monitor setups we need to translate the himetric location to the
// rect of the display device it's attached to.
let x = display_rect.left as f64
+ pointer_info.ptHimetricLocation.x as f64 * himetric_to_pixel_ratio_x;
let y = display_rect.top as f64
+ pointer_info.ptHimetricLocation.y as f64 * himetric_to_pixel_ratio_y;
let mut location = POINT { x: x.floor() as i32, y: y.floor() as i32 };
if unsafe { ScreenToClient(window, &mut location) } == false.into() {
continue;
}
let force = if let PT_TOUCH = pointer_info.pointerType {
let mut touch_info = mem::MaybeUninit::uninit();
util::GET_POINTER_TOUCH_INFO.and_then(|GetPointerTouchInfo| {
match unsafe {
GetPointerTouchInfo(pointer_info.pointerId, touch_info.as_mut_ptr())
} {
0 => None,
_ => normalize_pointer_pressure(unsafe {
touch_info.assume_init().pressure
}),
}
})
} else {
None
};
let x = location.x as f64 + x.fract();
let y = location.y as f64 + y.fract();
let position = PhysicalPosition::new(x, y);
let window_id = WindowId::from_raw(window as usize);
let finger_id = FingerId::from_raw(pointer_info.pointerId as usize);
let primary = util::has_flag(pointer_info.pointerFlags, POINTER_FLAG_PRIMARY);
if util::has_flag(pointer_info.pointerFlags, POINTER_FLAG_DOWN) {
userdata.send_event(Event::WindowEvent {
window_id,
event: WindowEvent::PointerEntered {
device_id: None,
primary,
position,
kind: if let PT_TOUCH = pointer_info.pointerType {
PointerKind::Touch(finger_id)
} else {
PointerKind::Unknown
},
},
});
userdata.send_event(Event::WindowEvent {
window_id,
event: WindowEvent::PointerButton {
device_id: None,
primary,
state: Pressed,
position,
button: if let PT_TOUCH = pointer_info.pointerType {
ButtonSource::Touch { finger_id, force }
} else {
ButtonSource::Unknown(0)
},
},
});
} else if util::has_flag(pointer_info.pointerFlags, POINTER_FLAG_UP) {
userdata.send_event(Event::WindowEvent {
window_id,
event: WindowEvent::PointerButton {
device_id: None,
primary,
state: Released,
position,
button: if let PT_TOUCH = pointer_info.pointerType {
ButtonSource::Touch { finger_id, force }
} else {
ButtonSource::Unknown(0)
},
},
});
userdata.send_event(Event::WindowEvent {
window_id,
event: WindowEvent::PointerLeft {
device_id: None,
primary,
position: Some(position),
kind: if let PT_TOUCH = pointer_info.pointerType {
PointerKind::Touch(finger_id)
} else {
PointerKind::Unknown
},
},
});
} else if util::has_flag(pointer_info.pointerFlags, POINTER_FLAG_UPDATE) {
userdata.send_event(Event::WindowEvent {
window_id,
event: WindowEvent::PointerMoved {
device_id: None,
primary,
position,
source: if let PT_TOUCH = pointer_info.pointerType {
PointerSource::Touch { finger_id, force }
} else {
PointerSource::Unknown
},
},
});
} else {
continue;
}
}
unsafe { SkipPointerFrameMessages(pointer_id) };
}
result = ProcResult::Value(0);
},
WM_NCACTIVATE => {
let is_active = wparam != false.into();
2022-08-31 18:32:19 +02:00
let active_focus_changed = userdata.window_state_lock().set_active(is_active);
if active_focus_changed {
if is_active {
unsafe { gain_active_focus(window, userdata) };
} else {
unsafe { lose_active_focus(window, userdata) };
}
}
result = ProcResult::DefWindowProc(wparam);
},
WM_SETFOCUS => {
2022-08-31 18:32:19 +02:00
let active_focus_changed = userdata.window_state_lock().set_focused(true);
if active_focus_changed {
unsafe { gain_active_focus(window, userdata) };
}
result = ProcResult::Value(0);
2019-06-24 12:14:55 -04:00
},
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
2022-03-07 22:58:12 +01:00
WM_KILLFOCUS => {
2022-08-31 18:32:19 +02:00
let active_focus_changed = userdata.window_state_lock().set_focused(false);
if active_focus_changed {
unsafe { lose_active_focus(window, userdata) };
}
result = ProcResult::Value(0);
2019-06-24 12:14:55 -04:00
},
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
2022-03-07 22:58:12 +01:00
WM_SETCURSOR => {
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
let set_cursor_to = {
2022-08-31 18:32:19 +02:00
let window_state = userdata.window_state_lock();
// The return value for the preceding `WM_NCHITTEST` message is conveniently
// provided through the low-order word of lParam. We use that here since
// `WM_MOUSEMOVE` seems to come after `WM_SETCURSOR` for a given cursor movement.
2022-03-07 22:58:12 +01:00
let in_client_area = super::loword(lparam as u32) as u32 == HTCLIENT;
if in_client_area {
Some(window_state.mouse.selected_cursor.clone())
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
} else {
None
}
};
match set_cursor_to {
Some(selected_cursor) => {
let hcursor = match selected_cursor {
SelectedCursor::Named(cursor_icon) => unsafe {
LoadCursorW(0, util::to_windows_cursor(cursor_icon))
},
SelectedCursor::Custom(cursor) => cursor.as_raw_handle(),
};
unsafe { SetCursor(hcursor) };
result = ProcResult::Value(0);
2019-06-24 12:14:55 -04:00
},
None => result = ProcResult::DefWindowProc(wparam),
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
}
2019-06-24 12:14:55 -04:00
},
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
2022-03-07 22:58:12 +01:00
WM_GETMINMAXINFO => {
let mmi = lparam as *mut MINMAXINFO;
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
2022-08-31 18:32:19 +02:00
let window_state = userdata.window_state_lock();
let window_flags = window_state.window_flags;
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
if window_state.min_size.is_some() || window_state.max_size.is_some() {
if let Some(min_size) = window_state.min_size {
let min_size = min_size.to_physical(window_state.scale_factor);
let (width, height): (u32, u32) =
window_flags.adjust_size(window, min_size).into();
unsafe { (*mmi).ptMinTrackSize = POINT { x: width as i32, y: height as i32 } };
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
}
if let Some(max_size) = window_state.max_size {
let max_size = max_size.to_physical(window_state.scale_factor);
let (width, height): (u32, u32) =
window_flags.adjust_size(window, max_size).into();
unsafe { (*mmi).ptMaxTrackSize = POINT { x: width as i32, y: height as i32 } };
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
}
}
result = ProcResult::Value(0);
2019-06-24 12:14:55 -04:00
},
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
// Only sent on Windows 8.1 or newer. On Windows 7 and older user has to log out to change
// DPI, therefore all applications are closed while DPI is changing.
2022-03-07 22:58:12 +01:00
WM_DPICHANGED => {
2020-01-05 16:57:32 -05:00
use crate::event::WindowEvent::ScaleFactorChanged;
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
// This message actually provides two DPI values - x and y. However MSDN says that
// "you only need to use either the X-axis or the Y-axis value when scaling your
// application since they are the same".
// https://msdn.microsoft.com/en-us/library/windows/desktop/dn312083(v=vs.85).aspx
2022-03-07 22:58:12 +01:00
let new_dpi_x = super::loword(wparam as u32) as u32;
let new_scale_factor = dpi_to_scale_factor(new_dpi_x);
let old_scale_factor: f64;
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
let (allow_resize, window_flags) = {
2022-08-31 18:32:19 +02:00
let mut window_state = userdata.window_state_lock();
old_scale_factor = window_state.scale_factor;
window_state.scale_factor = new_scale_factor;
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
if new_scale_factor == old_scale_factor {
result = ProcResult::Value(0);
return;
}
let allow_resize = window_state.fullscreen.is_none()
&& !window_state.window_flags().contains(WindowFlags::MAXIMIZED);
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
(allow_resize, window_state.window_flags)
};
// New size as suggested by Windows.
let suggested_rect = unsafe { *(lparam as *const RECT) };
// The window rect provided is the window's outer size, not it's surface size. However,
// win32 doesn't provide an `UnadjustWindowRectEx` function to get the client rect from
// the outer rect, so we instead adjust the window rect to get the decoration margins
// and remove them from the outer size.
let margin_left: i32;
let margin_top: i32;
// let margin_right: i32;
// let margin_bottom: i32;
{
let adjusted_rect =
window_flags.adjust_rect(window, suggested_rect).unwrap_or(suggested_rect);
margin_left = suggested_rect.left - adjusted_rect.left;
margin_top = suggested_rect.top - adjusted_rect.top;
// margin_right = adjusted_rect.right - suggested_rect.right;
// margin_bottom = adjusted_rect.bottom - suggested_rect.bottom;
}
let old_physical_inner_rect = util::WindowArea::Inner
.get_rect(window)
.expect("failed to query (old) inner window area");
let old_physical_surface_size = PhysicalSize::new(
(old_physical_inner_rect.right - old_physical_inner_rect.left) as u32,
(old_physical_inner_rect.bottom - old_physical_inner_rect.top) as u32,
);
// `allow_resize` prevents us from re-applying DPI adjustment to the restored size after
// exiting fullscreen (the restored size is already DPI adjusted).
let new_physical_surface_size = match allow_resize {
// We calculate our own size because the default suggested rect doesn't do a great
// job of preserving the window's logical size.
true => old_physical_surface_size
.to_logical::<f64>(old_scale_factor)
.to_physical::<u32>(new_scale_factor),
false => old_physical_surface_size,
};
let new_surface_size = Arc::new(Mutex::new(new_physical_surface_size));
2022-06-10 13:43:33 +03:00
userdata.send_event(Event::WindowEvent {
window_id: WindowId::from_raw(window as usize),
2020-01-05 16:57:32 -05:00
event: ScaleFactorChanged {
scale_factor: new_scale_factor,
surface_size_writer: SurfaceSizeWriter::new(Arc::downgrade(&new_surface_size)),
},
});
let new_physical_surface_size = *new_surface_size.lock().unwrap();
drop(new_surface_size);
let dragging_window: bool;
{
2022-08-31 18:32:19 +02:00
let window_state = userdata.window_state_lock();
dragging_window =
window_state.window_flags().contains(WindowFlags::MARKER_IN_SIZE_MOVE);
// Unset maximized if we're changing the window's size.
if new_physical_surface_size != old_physical_surface_size {
WindowState::set_window_flags(window_state, window, |f| {
f.set(WindowFlags::MAXIMIZED, false)
});
}
}
let new_outer_rect: RECT;
{
let suggested_ul =
(suggested_rect.left + margin_left, suggested_rect.top + margin_top);
let mut conservative_rect = RECT {
left: suggested_ul.0,
top: suggested_ul.1,
right: suggested_ul.0 + new_physical_surface_size.width as i32,
bottom: suggested_ul.1 + new_physical_surface_size.height as i32,
};
conservative_rect = window_flags
.adjust_rect(window, conservative_rect)
.unwrap_or(conservative_rect);
// If we're dragging the window, offset the window so that the cursor's
// relative horizontal position in the title bar is preserved.
if dragging_window {
let bias = {
let cursor_pos = {
let mut pos = unsafe { mem::zeroed() };
unsafe { GetCursorPos(&mut pos) };
pos
};
let suggested_cursor_horizontal_ratio = (cursor_pos.x - suggested_rect.left)
as f64
/ (suggested_rect.right - suggested_rect.left) as f64;
(cursor_pos.x
- (suggested_cursor_horizontal_ratio
* (conservative_rect.right - conservative_rect.left) as f64)
2022-03-07 22:58:12 +01:00
as i32)
- conservative_rect.left
};
conservative_rect.left += bias;
conservative_rect.right += bias;
}
// Check to see if the new window rect is on the monitor with the new DPI factor.
// If it isn't, offset the window so that it is.
let new_dpi_monitor = unsafe { MonitorFromWindow(window, MONITOR_DEFAULTTONULL) };
2022-03-07 22:58:12 +01:00
let conservative_rect_monitor =
unsafe { MonitorFromRect(&conservative_rect, MONITOR_DEFAULTTONULL) };
new_outer_rect = if conservative_rect_monitor == new_dpi_monitor {
conservative_rect
} else {
let get_monitor_rect = |monitor| {
2022-03-07 22:58:12 +01:00
let mut monitor_info = MONITORINFO {
cbSize: mem::size_of::<MONITORINFO>() as _,
..unsafe { mem::zeroed() }
};
unsafe { GetMonitorInfoW(monitor, &mut monitor_info) };
monitor_info.rcMonitor
};
let wrong_monitor = conservative_rect_monitor;
let wrong_monitor_rect = get_monitor_rect(wrong_monitor);
let new_monitor_rect = get_monitor_rect(new_dpi_monitor);
// The direction to nudge the window in to get the window onto the monitor with
// the new DPI factor. We calculate this by seeing which monitor edges are
// shared and nudging away from the wrong monitor based on those.
#[allow(clippy::bool_to_int_with_if)]
let delta_nudge_to_dpi_monitor = (
if wrong_monitor_rect.left == new_monitor_rect.right {
-1
} else if wrong_monitor_rect.right == new_monitor_rect.left {
1
} else {
0
},
if wrong_monitor_rect.bottom == new_monitor_rect.top {
1
} else if wrong_monitor_rect.top == new_monitor_rect.bottom {
-1
} else {
0
},
);
let abort_after_iterations = new_monitor_rect.right - new_monitor_rect.left
+ new_monitor_rect.bottom
- new_monitor_rect.top;
for _ in 0..abort_after_iterations {
conservative_rect.left += delta_nudge_to_dpi_monitor.0;
conservative_rect.right += delta_nudge_to_dpi_monitor.0;
conservative_rect.top += delta_nudge_to_dpi_monitor.1;
conservative_rect.bottom += delta_nudge_to_dpi_monitor.1;
if unsafe { MonitorFromRect(&conservative_rect, MONITOR_DEFAULTTONULL) }
2022-03-07 22:58:12 +01:00
== new_dpi_monitor
{
break;
}
}
conservative_rect
};
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
}
unsafe {
SetWindowPos(
window,
0,
new_outer_rect.left,
new_outer_rect.top,
new_outer_rect.right - new_outer_rect.left,
new_outer_rect.bottom - new_outer_rect.top,
SWP_NOZORDER | SWP_NOACTIVATE,
)
};
result = ProcResult::Value(0);
2019-06-24 12:14:55 -04:00
},
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
2022-03-07 22:58:12 +01:00
WM_SETTINGCHANGE => {
use crate::event::WindowEvent::ThemeChanged;
2022-08-31 18:32:19 +02:00
let preferred_theme = userdata.window_state_lock().preferred_theme;
if preferred_theme.is_none() {
let new_theme = try_theme(window, preferred_theme);
2022-08-31 18:32:19 +02:00
let mut window_state = userdata.window_state_lock();
if window_state.current_theme != new_theme {
window_state.current_theme = new_theme;
drop(window_state);
userdata.send_event(Event::WindowEvent {
window_id: WindowId::from_raw(window as usize),
event: ThemeChanged(new_theme),
});
}
}
result = ProcResult::DefWindowProc(wparam);
},
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
_ => {
if msg == DESTROY_MSG_ID.get() {
unsafe { DestroyWindow(window) };
result = ProcResult::Value(0);
} else if msg == SET_RETAIN_STATE_ON_SIZE_MSG_ID.get() {
2022-08-31 18:32:19 +02:00
let mut window_state = userdata.window_state_lock();
window_state.set_window_flags_in_place(|f| {
f.set(WindowFlags::MARKER_RETAIN_STATE_ON_SIZE, wparam != 0)
});
result = ProcResult::Value(0);
} else if msg == TASKBAR_CREATED.get() {
2022-08-31 18:32:19 +02:00
let window_state = userdata.window_state_lock();
unsafe { set_skip_taskbar(window, window_state.skip_taskbar) };
result = ProcResult::DefWindowProc(wparam);
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
} else {
result = ProcResult::DefWindowProc(wparam);
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
}
2019-06-24 12:14:55 -04:00
},
};
userdata
.event_loop_runner
.catch_unwind(callback)
.unwrap_or_else(|| result = ProcResult::Value(-1));
match result {
ProcResult::DefWindowProc(wparam) => unsafe { DefWindowProcW(window, msg, wparam, lparam) },
ProcResult::Value(val) => val,
}
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
}
Windows: Make `EventLoopWindowTarget` independent of UserEvent type (#3061) * make `EventLoopWindowTarget` independent of UserEvent type the `EventLoopWindowTarget` is needed for window creation. conceptually, only `EventLoop` and `EventLoopProxy` need to be parameterized, and all other parts of the backend should be agnostic about the user event type, parallel to how `Event<T>` is parameterized, but `WindowEvent` is not. this change removes the dependency on the type of user events from the `EventLoopWindowTarget` for the Windows backend, but keep a phantom data to keep the API intact. to achieve this, I moved the `Receiver` end of the mpsc channel from `ThreadMsgTargetData` into `EventLoop` itself, so the `UserEvent` is only passed between `EventLoop` and `EventLoopProxy`, all other part of the backend just use unit type as a placeholder for user events. it's similar to the macos backend where an erased `EventHandler` trait object is used so all component except `EventLoop` and `EventLoopProxy` need to be parameterized. however `EventLoop` of the Windows backend already use an `Box<dyn FnMut>` to wrap the user provided event handler callback, so no need for an dedicated trait object, I just modified the wrapper to replace the placeholder user event with real value pulled from the channel. I find this is the approach which need minimum change to be made to existing code. but it does the job and could serve as a starting point to future Windows backend re-works. * fix CI clippy failure. * make UserEventPlaceholder a new type instead of alias * invariance is maintained by top-level EventLoopWindowTarget<T> this field is transitional and her to keep API compatibility only. the correct variance and such is already ensured by the top-level `EventLoopWindowTarget`, just use `PhantomData<T>` here.
2024-01-04 23:47:07 +08:00
unsafe extern "system" fn thread_event_target_callback(
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
window: HWND,
2022-03-07 22:58:12 +01:00
msg: u32,
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
wparam: WPARAM,
lparam: LPARAM,
) -> LRESULT {
let userdata_ptr =
Windows: Make `EventLoopWindowTarget` independent of UserEvent type (#3061) * make `EventLoopWindowTarget` independent of UserEvent type the `EventLoopWindowTarget` is needed for window creation. conceptually, only `EventLoop` and `EventLoopProxy` need to be parameterized, and all other parts of the backend should be agnostic about the user event type, parallel to how `Event<T>` is parameterized, but `WindowEvent` is not. this change removes the dependency on the type of user events from the `EventLoopWindowTarget` for the Windows backend, but keep a phantom data to keep the API intact. to achieve this, I moved the `Receiver` end of the mpsc channel from `ThreadMsgTargetData` into `EventLoop` itself, so the `UserEvent` is only passed between `EventLoop` and `EventLoopProxy`, all other part of the backend just use unit type as a placeholder for user events. it's similar to the macos backend where an erased `EventHandler` trait object is used so all component except `EventLoop` and `EventLoopProxy` need to be parameterized. however `EventLoop` of the Windows backend already use an `Box<dyn FnMut>` to wrap the user provided event handler callback, so no need for an dedicated trait object, I just modified the wrapper to replace the placeholder user event with real value pulled from the channel. I find this is the approach which need minimum change to be made to existing code. but it does the job and could serve as a starting point to future Windows backend re-works. * fix CI clippy failure. * make UserEventPlaceholder a new type instead of alias * invariance is maintained by top-level EventLoopWindowTarget<T> this field is transitional and her to keep API compatibility only. the correct variance and such is already ensured by the top-level `EventLoopWindowTarget`, just use `PhantomData<T>` here.
2024-01-04 23:47:07 +08:00
unsafe { super::get_window_long(window, GWL_USERDATA) } as *mut ThreadMsgTargetData;
if userdata_ptr.is_null() {
// `userdata_ptr` will always be null for the first `WM_GETMINMAXINFO`, as well as
// `WM_NCCREATE` and `WM_CREATE`.
return unsafe { DefWindowProcW(window, msg, wparam, lparam) };
}
let userdata = unsafe { Box::from_raw(userdata_ptr) };
2022-03-07 22:58:12 +01:00
if msg != WM_PAINT {
unsafe { RedrawWindow(window, ptr::null(), 0, RDW_INTERNALPAINT) };
}
let mut userdata_removed = false;
// I decided to bind the closure to `callback` and pass it to catch_unwind rather than passing
// the closure to catch_unwind directly so that the match body indentation wouldn't change and
// the git blame and history would be preserved.
let callback = || match msg {
2022-03-07 22:58:12 +01:00
WM_NCDESTROY => {
unsafe { super::set_window_long(window, GWL_USERDATA, 0) };
userdata_removed = true;
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
0
2019-06-24 12:14:55 -04:00
},
WM_PAINT => unsafe {
2022-03-07 22:58:12 +01:00
ValidateRect(window, ptr::null());
// Default WM_PAINT behaviour. This makes sure modals and popups are shown immediately
// when opening them.
2022-03-07 22:58:12 +01:00
DefWindowProcW(window, msg, wparam, lparam)
},
2022-03-07 22:58:12 +01:00
WM_INPUT => {
if let Some(data) = raw_input::get_raw_input_data(lparam as _) {
unsafe { handle_raw_input(&userdata, data) };
}
unsafe { DefWindowProcW(window, msg, wparam, lparam) }
},
_ if msg == USER_EVENT_MSG_ID.get() => {
Windows: Make `EventLoopWindowTarget` independent of UserEvent type (#3061) * make `EventLoopWindowTarget` independent of UserEvent type the `EventLoopWindowTarget` is needed for window creation. conceptually, only `EventLoop` and `EventLoopProxy` need to be parameterized, and all other parts of the backend should be agnostic about the user event type, parallel to how `Event<T>` is parameterized, but `WindowEvent` is not. this change removes the dependency on the type of user events from the `EventLoopWindowTarget` for the Windows backend, but keep a phantom data to keep the API intact. to achieve this, I moved the `Receiver` end of the mpsc channel from `ThreadMsgTargetData` into `EventLoop` itself, so the `UserEvent` is only passed between `EventLoop` and `EventLoopProxy`, all other part of the backend just use unit type as a placeholder for user events. it's similar to the macos backend where an erased `EventHandler` trait object is used so all component except `EventLoop` and `EventLoopProxy` need to be parameterized. however `EventLoop` of the Windows backend already use an `Box<dyn FnMut>` to wrap the user provided event handler callback, so no need for an dedicated trait object, I just modified the wrapper to replace the placeholder user event with real value pulled from the channel. I find this is the approach which need minimum change to be made to existing code. but it does the job and could serve as a starting point to future Windows backend re-works. * fix CI clippy failure. * make UserEventPlaceholder a new type instead of alias * invariance is maintained by top-level EventLoopWindowTarget<T> this field is transitional and her to keep API compatibility only. the correct variance and such is already ensured by the top-level `EventLoopWindowTarget`, just use `PhantomData<T>` here.
2024-01-04 23:47:07 +08:00
// synthesis a placeholder UserEvent, so that if the callback is
// re-entered it can be buffered for later delivery. the real
// user event is still in the mpsc channel and will be pulled
// once the placeholder event is delivered to the wrapper
// `event_handler`
userdata.send_event(Event::UserWakeUp);
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
0
2019-06-24 12:14:55 -04:00
},
_ if msg == EXEC_MSG_ID.get() => {
let mut function: ThreadExecFn = unsafe { Box::from_raw(wparam as *mut _) };
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
function();
0
},
_ => unsafe { DefWindowProcW(window, msg, wparam, lparam) },
};
let result = userdata.event_loop_runner.catch_unwind(callback).unwrap_or(-1);
if userdata_removed {
drop(userdata);
} else {
Box::leak(userdata);
}
result
Event Loop 2.0 API and Windows implementation (#638) * Rename EventsLoop and associated types to EventLoop * Rename WindowEvent::Refresh to WindowEvent::Redraw * Remove second thread from win32 backend * Update run_forever to hijack thread * Replace windows Mutex with parking_lot Mutex * Implement new ControlFlow and associated events * Add StartCause::Init support, timer example * Add ability to send custom user events * Fully invert windows control flow so win32 calls into winit's callback * Add request_redraw * Rename platform to platform_impl * Rename os to platform, add Ext trait postfixes * Add platform::desktop module with EventLoopExt::run_return * Re-organize into module structure * Improve documentation * Small changes to examples * Improve docs for run and run_return * Change instances of "events_loop" to "event_loop" * Rename MonitorId to MonitorHandle * Add CHANGELOG entry * Improve WaitUntil timer precision * When SendEvent is called during event closure, buffer events * Fix resize lag when waiting in some situations * Update send test and errors that broke some examples/APIs * Improve clarity/fix typos in docs * Fix unreachable panic after setting ControlFlow to Poll during some RedrawRequested events. * Fix crash when running in release mode * Remove crossbeam dependency and make drop events work again * Remove serde implementations from ControlFlow * Fix 1.24.1 build * Fix freeze when setting decorations * Replace &EventLoop in callback with &EventLoopWindowTarget * Document and implement Debug for EventLoopWindowTarget * Fix some deadlocks that could occur when changing window state * Fix thread executor not executing closure when called from non-loop thread * Fix buffered events not getting dispatched * Fix crash with runner refcell not getting dropped * Address review feedback * Fix CHANGELOG typo * Catch panics in user callback
2019-02-05 10:30:33 -05:00
}
Windows: Make `EventLoopWindowTarget` independent of UserEvent type (#3061) * make `EventLoopWindowTarget` independent of UserEvent type the `EventLoopWindowTarget` is needed for window creation. conceptually, only `EventLoop` and `EventLoopProxy` need to be parameterized, and all other parts of the backend should be agnostic about the user event type, parallel to how `Event<T>` is parameterized, but `WindowEvent` is not. this change removes the dependency on the type of user events from the `EventLoopWindowTarget` for the Windows backend, but keep a phantom data to keep the API intact. to achieve this, I moved the `Receiver` end of the mpsc channel from `ThreadMsgTargetData` into `EventLoop` itself, so the `UserEvent` is only passed between `EventLoop` and `EventLoopProxy`, all other part of the backend just use unit type as a placeholder for user events. it's similar to the macos backend where an erased `EventHandler` trait object is used so all component except `EventLoop` and `EventLoopProxy` need to be parameterized. however `EventLoop` of the Windows backend already use an `Box<dyn FnMut>` to wrap the user provided event handler callback, so no need for an dedicated trait object, I just modified the wrapper to replace the placeholder user event with real value pulled from the channel. I find this is the approach which need minimum change to be made to existing code. but it does the job and could serve as a starting point to future Windows backend re-works. * fix CI clippy failure. * make UserEventPlaceholder a new type instead of alias * invariance is maintained by top-level EventLoopWindowTarget<T> this field is transitional and her to keep API compatibility only. the correct variance and such is already ensured by the top-level `EventLoopWindowTarget`, just use `PhantomData<T>` here.
2024-01-04 23:47:07 +08:00
unsafe fn handle_raw_input(userdata: &ThreadMsgTargetData, data: RAWINPUT) {
use crate::event::DeviceEvent::{Button, Key, MouseWheel, PointerMotion};
use crate::event::ElementState::{Pressed, Released};
use crate::event::MouseScrollDelta::LineDelta;
let device_id = Some(wrap_device_id(data.header.hDevice as _));
if data.header.dwType == RIM_TYPEMOUSE {
let mouse = unsafe { data.data.mouse };
if util::has_flag(mouse.usFlags as u32, MOUSE_MOVE_RELATIVE) {
let x = mouse.lLastX as f64;
let y = mouse.lLastY as f64;
if x != 0.0 || y != 0.0 {
userdata.send_event(Event::DeviceEvent {
device_id,
event: PointerMotion { delta: (x, y) },
});
}
}
let button_flags = unsafe { mouse.Anonymous.Anonymous.usButtonFlags };
if util::has_flag(button_flags as u32, RI_MOUSE_WHEEL) {
let button_data = unsafe { mouse.Anonymous.Anonymous.usButtonData } as i16;
let delta = button_data as f32 / WHEEL_DELTA as f32;
userdata.send_event(Event::DeviceEvent {
device_id,
event: MouseWheel { delta: LineDelta(0.0, delta) },
});
}
if util::has_flag(button_flags as u32, RI_MOUSE_HWHEEL) {
let button_data = unsafe { mouse.Anonymous.Anonymous.usButtonData } as i16;
let delta = -button_data as f32 / WHEEL_DELTA as f32;
userdata.send_event(Event::DeviceEvent {
device_id,
event: MouseWheel { delta: LineDelta(delta, 0.0) },
});
}
let button_state = raw_input::get_raw_mouse_button_state(button_flags as u32);
for (button, state) in button_state.iter().enumerate() {
if let Some(state) = *state {
userdata.send_event(Event::DeviceEvent {
device_id,
event: Button { button: button as _, state },
});
}
}
} else if data.header.dwType == RIM_TYPEKEYBOARD {
let keyboard = unsafe { data.data.keyboard };
let pressed = keyboard.Message == WM_KEYDOWN || keyboard.Message == WM_SYSKEYDOWN;
let released = keyboard.Message == WM_KEYUP || keyboard.Message == WM_SYSKEYUP;
if !pressed && !released {
return;
}
if let Some(physical_key) = raw_input::get_keyboard_physical_key(keyboard) {
let state = if pressed { Pressed } else { Released };
userdata.send_event(Event::DeviceEvent {
device_id,
event: Key(RawKeyEvent { physical_key, state }),
});
}
}
}
enum PointerMoveKind {
/// Pointer enterd to the window.
Enter,
/// Pointer leaved the window client area.
Leave,
/// Pointer is inside the window or `GetClientRect` failed.
None,
}
fn get_pointer_move_kind(
window: HWND,
mouse_was_inside_window: bool,
x: i32,
y: i32,
) -> PointerMoveKind {
let rect: RECT = unsafe {
let mut rect: RECT = mem::zeroed();
if GetClientRect(window, &mut rect) == false.into() {
return PointerMoveKind::None; // exit early if GetClientRect failed
}
rect
};
let x = (rect.left..rect.right).contains(&x);
let y = (rect.top..rect.bottom).contains(&y);
if !mouse_was_inside_window && x && y {
PointerMoveKind::Enter
} else if mouse_was_inside_window && !(x && y) {
PointerMoveKind::Leave
} else {
PointerMoveKind::None
}
}