snowykami/meilisearch
1
0
Fork 0
mirror of https://github.com/meilisearch/meilisearch.git synced 2025-07-31 02:40:01 +00:00
Code Issues Packages Projects Releases Wiki Activity

Fix more bugs + visual empty path cache logging

Browse Source
This commit is contained in:
Loïc Lecrenier
2023-02-27 15:04:40 +01:00
parent 0e1fbbf7c6
commit 6c85c0d95e
9 changed files with 107 additions and 68 deletions
Download Patch File Download Diff File Expand all files Collapse all files

12
milli/src/search/new/ranking_rule_graph/cheapest_paths.rs
View File

@ -81,7 +81,9 @@ impl KCheapestPathsState {
while self.kth_cheapest_path.cost <= cur_cost {
if let Some(next_self) = self.compute_next_cheapest_paths(graph, empty_paths_cache) {
self = next_self;
if self.kth_cheapest_path.cost == cur_cost {
if self.kth_cheapest_path.cost == cur_cost
&& !empty_paths_cache.path_is_empty(&self.kth_cheapest_path.edges)
{
into_map.add_path(&self.kth_cheapest_path);
} else {
break;
@ -148,7 +150,13 @@ impl KCheapestPathsState {
while let Some((next_cheapest_path, cost2)) = next_cheapest_paths.remove_first() {
assert_eq!(cost, cost2);
if empty_paths_cache.path_is_empty(&next_cheapest_path) {
// NOTE: it is important not to discard the paths that are forbidden due to a
// forbidden prefix, because the cheapest path algorithm (Dijkstra) cannot take
// this property into account.
if next_cheapest_path
.iter()
.any(|edge_index| graph.all_edges[*edge_index as usize].is_none())
{
continue;
} else {
self.cheapest_paths.insert(next_cheapest_path.iter().copied(), cost);

6
milli/src/search/new/ranking_rule_graph/empty_paths_cache.rs
View File

@ -4,12 +4,16 @@ use roaring::RoaringBitmap;
use super::paths_map::PathsMap;
#[derive(Default)]
#[derive(Default, Clone)]
pub struct EmptyPathsCache {
pub empty_edges: RoaringBitmap,
pub empty_prefixes: PathsMap<()>,
}
impl EmptyPathsCache {
pub fn forbid_edge(&mut self, edge_idx: u32) {
self.empty_edges.insert(edge_idx);
self.empty_prefixes.remove_edge(&edge_idx);
}
pub fn path_is_empty(&self, path: &[u32]) -> bool {
for edge in path {
if self.empty_edges.contains(*edge) {

2
milli/src/search/new/ranking_rule_graph/mod.rs
View File

@ -11,6 +11,7 @@ use std::ops::ControlFlow;
use heed::RoTxn;
use roaring::RoaringBitmap;
use self::empty_paths_cache::EmptyPathsCache;
use self::paths_map::PathsMap;
use super::db_cache::DatabaseCache;
@ -82,6 +83,7 @@ pub trait RankingRuleGraphTrait: Sized {
fn log_state(
graph: &RankingRuleGraph<Self>,
paths: &PathsMap<u64>,
empty_paths_cache: &EmptyPathsCache,
logger: &mut dyn SearchLogger<QueryGraph>,
);
}

4
milli/src/search/new/ranking_rule_graph/proximity/mod.rs
View File

@ -3,6 +3,7 @@ pub mod compute_docids;
use heed::RoTxn;
use super::empty_paths_cache::EmptyPathsCache;
use super::paths_map::PathsMap;
use super::{Edge, EdgeDetails, RankingRuleGraphTrait};
use crate::new::db_cache::DatabaseCache;
@ -67,8 +68,9 @@ impl RankingRuleGraphTrait for ProximityGraph {
fn log_state(
graph: &super::RankingRuleGraph<Self>,
paths: &PathsMap<u64>,
empty_paths_cache: &EmptyPathsCache,
logger: &mut dyn crate::new::logger::SearchLogger<crate::new::QueryGraph>,
) {
logger.log_proximity_state(graph, paths);
logger.log_proximity_state(graph, paths, empty_paths_cache);
}
}

2
milli/src/search/new/ranking_rule_graph/resolve_paths.rs
View File

@ -40,7 +40,7 @@ impl<G: RankingRuleGraphTrait> RankingRuleGraph<G> {
BitmapOrAllRef::Bitmap(edge_docids) => {
if edge_docids.is_disjoint(universe) {
// 1. Store in the cache that this edge is empty for this universe
empty_paths_cache.empty_edges.insert(edge_index);
empty_paths_cache.forbid_edge(edge_index);
// 2. remove all the paths that contain this edge for this universe
paths.remove_edge(&edge_index);
// 3. remove this edge from the proximity graph