searchengines-web/MeiliSearch
searchengines-web/MeiliSearch
1
0
Fork 0
mirror of https://github.com/meilisearch/MeiliSearch synced 2024-11-29 16:24:26 +01:00
Code Issues Releases Wiki Activity

Prefer using a smallstr instead of a real String to reduce allocations

Browse Source
This commit is contained in:
Clément Renault 2020-10-29 14:32:32 +01:00
parent 40993a0d25
commit b5d52b6b45
No known key found for this signature in database
GPG Key ID: 92ADA4E935E71FA4
3 changed files with 8 additions and 8 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

1
Cargo.lock generated
View File

@ -1838,6 +1838,7 @@ version = "0.2.0"
source = "registry+https://github.com/rust-lang/crates.io-index" source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "1e922794d168678729ffc7e07182721a14219c65814e66e91b839a272fe5ae4f" checksum = "1e922794d168678729ffc7e07182721a14219c65814e66e91b839a272fe5ae4f"
dependencies = [ dependencies = [
"serde",
"smallvec", "smallvec",
] ]

2
Cargo.toml
View File

@ -29,7 +29,7 @@ ringtail = "0.3.0"
roaring = "0.6.1" roaring = "0.6.1"
serde_json = "1.0.59" serde_json = "1.0.59"
slice-group-by = "0.2.6" slice-group-by = "0.2.6"
smallstr = "0.2.0" smallstr = { version = "0.2.0", features = ["serde"] }
smallvec = "1.4.0" smallvec = "1.4.0"
structopt = { version = "0.3.14", default-features = false, features = ["wrap_help"] } structopt = { version = "0.3.14", default-features = false, features = ["wrap_help"] }
tempfile = "3.1.0" tempfile = "3.1.0"

13
src/update/delete_documents.rs
View File

@ -4,7 +4,7 @@ use std::convert::TryFrom;
use fst::{IntoStreamer, Streamer}; use fst::{IntoStreamer, Streamer};
use roaring::RoaringBitmap; use roaring::RoaringBitmap;
use crate::{Index, BEU32}; use crate::{Index, BEU32, SmallString32};
use super::ClearDocuments; use super::ClearDocuments;
pub struct DeleteDocuments<'t, 'u, 'i> { pub struct DeleteDocuments<'t, 'u, 'i> {
@ -77,7 +77,6 @@ impl<'t, 'u, 'i> DeleteDocuments<'t, 'u, 'i> {
} = self.index; } = self.index;
// Retrieve the words and the users ids contained in the documents. // Retrieve the words and the users ids contained in the documents.
// TODO we must use a smallword instead of a string.
let mut words = Vec::new(); let mut words = Vec::new();
let mut users_ids = Vec::new(); let mut users_ids = Vec::new();
for docid in &documents_ids { for docid in &documents_ids {
@ -88,7 +87,7 @@ impl<'t, 'u, 'i> DeleteDocuments<'t, 'u, 'i> {
let mut iter = documents.range_mut(self.wtxn, &(key..=key))?; let mut iter = documents.range_mut(self.wtxn, &(key..=key))?;
if let Some((_key, obkv)) = iter.next().transpose()? { if let Some((_key, obkv)) = iter.next().transpose()? {
if let Some(content) = obkv.get(id_field) { if let Some(content) = obkv.get(id_field) {
let user_id: String = serde_json::from_slice(content).unwrap(); let user_id: SmallString32 = serde_json::from_slice(content).unwrap();
users_ids.push(user_id); users_ids.push(user_id);
} }
iter.del_current()?; iter.del_current()?;
@ -101,14 +100,14 @@ impl<'t, 'u, 'i> DeleteDocuments<'t, 'u, 'i> {
while let Some(result) = iter.next() { while let Some(result) = iter.next() {
let ((_docid, word), _positions) = result?; let ((_docid, word), _positions) = result?;
// This boolean will indicate if we must remove this word from the words FST. // This boolean will indicate if we must remove this word from the words FST.
words.push((String::from(word), false)); words.push((SmallString32::from(word), false));
iter.del_current()?; iter.del_current()?;
} }
} }
// We create the FST map of the users ids that we must delete. // We create the FST map of the users ids that we must delete.
users_ids.sort_unstable(); users_ids.sort_unstable();
let users_ids_to_delete = fst::Set::from_iter(users_ids)?; let users_ids_to_delete = fst::Set::from_iter(users_ids.iter().map(AsRef::as_ref))?;
let users_ids_to_delete = fst::Map::from(users_ids_to_delete.into_fst()); let users_ids_to_delete = fst::Map::from(users_ids_to_delete.into_fst());
let new_users_ids_documents_ids = { let new_users_ids_documents_ids = {
@ -143,7 +142,7 @@ impl<'t, 'u, 'i> DeleteDocuments<'t, 'u, 'i> {
// the LMDB B-Tree two times but only once. // the LMDB B-Tree two times but only once.
let mut iter = word_docids.prefix_iter_mut(self.wtxn, &word)?; let mut iter = word_docids.prefix_iter_mut(self.wtxn, &word)?;
if let Some((key, mut docids)) = iter.next().transpose()? { if let Some((key, mut docids)) = iter.next().transpose()? {
if key == word { if key == word.as_ref() {
docids.difference_with(&mut documents_ids); docids.difference_with(&mut documents_ids);
if docids.is_empty() { if docids.is_empty() {
iter.del_current()?; iter.del_current()?;
@ -156,7 +155,7 @@ impl<'t, 'u, 'i> DeleteDocuments<'t, 'u, 'i> {
} }
// We construct an FST set that contains the words to delete from the words FST. // We construct an FST set that contains the words to delete from the words FST.
let words_to_delete = words.iter().filter_map(|(w, d)| if *d { Some(w) } else { None }); let words_to_delete = words.iter().filter_map(|(w, d)| if *d { Some(w.as_ref()) } else { None });
let words_to_delete = fst::Set::from_iter(words_to_delete)?; let words_to_delete = fst::Set::from_iter(words_to_delete)?;
let new_words_fst = { let new_words_fst = {