feat(service): prefer recently/often used applications in search

This commit is contained in:
Thomas Ruprecht 2022-11-01 22:41:23 +01:00 committed by GitHub
parent f382690b28
commit 4eef0caae5
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
6 changed files with 225 additions and 25 deletions

90
service/src/recent.rs Normal file
View file

@ -0,0 +1,90 @@
use std::collections::{HashMap, hash_map::DefaultHasher};
use std::hash::{Hasher, Hash};
use serde::{Deserialize, Serialize, Serializer, Deserializer};
const SHORTTERM_CAP: usize = 20;
const LONGTERM_CAP: usize = 100;
// Holds a long term storage that tracks how often a search
// result was activated, and a short term storage that stores
// the order of recently activated search results (higher
// vales are more recent).
// Keys for both mappings are hashes of the acvtivated result's
// command string.
#[derive(Debug, Default)]
pub struct RecentUseStorage {
long_term: HashMap<usize, usize>,
short_term: HashMap<usize, usize>,
}
fn hash_key<K: Hash>(key: K) -> usize {
let mut hasher = DefaultHasher::new();
key.hash(&mut hasher);
hasher.finish() as usize
}
impl RecentUseStorage {
pub fn add<K: Hash>(&mut self, exec: &K) {
let key = hash_key(exec);
*self.long_term.entry(key).or_insert(0) += 1;
let short_term_idx = self.short_term.values().max().unwrap_or( &0)+1;
self.short_term.insert(key, short_term_idx);
self.trim()
}
fn trim(&mut self) {
while self.short_term.len() > SHORTTERM_CAP {
let key = *self.short_term.iter().min_by_key(|kv| kv.1).unwrap().0;
self.short_term.remove(&key);
}
while self.long_term.values().sum::<usize>() > LONGTERM_CAP {
let mut delete_keys = Vec::new();
for (k, v) in self.long_term.iter_mut() {
*v /= 2;
if *v == 0 {
delete_keys.push(*k);
}
}
for k in delete_keys {
self.long_term.remove(&k);
}
}
}
pub fn get_recent<K: Hash>(&self, exec: &K) -> usize {
self.short_term.get(&hash_key(exec)).copied().unwrap_or(0)
}
pub fn get_freq<K: Hash>(&self, exec: &K) -> usize {
self.long_term.get(&hash_key(exec)).copied().unwrap_or(0)
}
}
impl Serialize for RecentUseStorage {
fn serialize<S>(&self, serializer: S) -> Result<S::Ok, S::Error>
where
S: Serializer,
{
let mut stvec: Vec<_> = self.short_term.keys().copied().collect();
stvec.sort_by_key(|k| self.short_term[k]);
(&self.long_term, stvec).serialize(serializer)
}
}
impl<'de> Deserialize<'de> for RecentUseStorage {
fn deserialize<D>(deserializer: D) -> Result<RecentUseStorage, D::Error>
where
D: Deserializer<'de>,
{
type SerType = (HashMap<usize, usize>, Vec<usize>);
let (long_term, stv) = SerType::deserialize(deserializer)?;
let short_term: HashMap<_, _> = stv.into_iter().enumerate().map(|(v,k)| (k,v)).collect();
Ok(RecentUseStorage {
long_term,
short_term,
})
}
}