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

Fixes the stop words and words fst generation

Browse Source
This commit is contained in:
Clément Renault 2020-04-06 18:53:02 +02:00
parent eb28276923
commit 29d021ad4d
No known key found for this signature in database
GPG Key ID: 0151CDAB43460DAE
1 changed files with 42 additions and 41 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

83
meilisearch-core/src/update/settings_update.rs
View File

@ -135,52 +135,61 @@ pub fn apply_stop_words_update(
writer: &mut heed::RwTxn<MainT>, writer: &mut heed::RwTxn<MainT>,
index: &store::Index, index: &store::Index,
stop_words: BTreeSet<String>, stop_words: BTreeSet<String>,
) -> MResult<bool> { ) -> MResult<bool>
{
let mut must_reindex = false;
let old_stop_words: BTreeSet<String> = index.main let old_stop_words: BTreeSet<String> = index.main
.stop_words_fst(writer)? .stop_words_fst(writer)?
.unwrap_or_default() .unwrap_or_default()
.stream() .stream()
.into_strs().unwrap().into_iter().collect(); .into_strs()?
.into_iter()
.collect();
let deletion: BTreeSet<String> = old_stop_words.difference(&stop_words).cloned().collect(); let deletion: BTreeSet<String> = old_stop_words.difference(&stop_words).cloned().collect();
let addition: BTreeSet<String> = stop_words.difference(&old_stop_words).cloned().collect(); let addition: BTreeSet<String> = stop_words.difference(&old_stop_words).cloned().collect();
if !addition.is_empty() { if !addition.is_empty() {
apply_stop_words_addition( apply_stop_words_addition(writer, index, addition)?;
writer,
index,
addition
)?;
} }
if !deletion.is_empty() { if !deletion.is_empty() {
apply_stop_words_deletion( must_reindex = true;
writer, apply_stop_words_deletion(writer, index, deletion)?;
index,
deletion
)?;
return Ok(true)
} }
let stop_words_fst = fst::Set::from_iter(stop_words)?; if let Some(words_fst) = index.main.words_fst(writer)? {
index.main.put_words_fst(writer, &stop_words_fst)?; let stop_words = fst::Set::from_iter(stop_words)?;
Ok(false) let op = OpBuilder::new()
.add(&words_fst)
.add(&stop_words)
.difference();
let mut builder = fst::SetBuilder::memory();
builder.extend_stream(op)?;
let words_fst = builder.into_inner().and_then(fst::Set::from_bytes)?;
index.main.put_words_fst(writer, &words_fst)?;
index.main.put_stop_words_fst(writer, &stop_words)?;
}
Ok(must_reindex)
} }
fn apply_stop_words_addition( fn apply_stop_words_addition(
writer: &mut heed::RwTxn<MainT>, writer: &mut heed::RwTxn<MainT>,
index: &store::Index, index: &store::Index,
addition: BTreeSet<String>, addition: BTreeSet<String>,
) -> MResult<()> { ) -> MResult<()>
{
let main_store = index.main; let main_store = index.main;
let postings_lists_store = index.postings_lists; let postings_lists_store = index.postings_lists;
let mut stop_words_builder = SetBuilder::memory(); let mut stop_words_builder = SetBuilder::memory();
for word in addition { for word in addition {
stop_words_builder.insert(&word).unwrap(); stop_words_builder.insert(&word)?;
// we remove every posting list associated to a new stop word // we remove every posting list associated to a new stop word
postings_lists_store.del_postings_list(writer, word.as_bytes())?; postings_lists_store.del_postings_list(writer, word.as_bytes())?;
} }
@ -188,8 +197,7 @@ fn apply_stop_words_addition(
// create the new delta stop words fst // create the new delta stop words fst
let delta_stop_words = stop_words_builder let delta_stop_words = stop_words_builder
.into_inner() .into_inner()
.and_then(fst::Set::from_bytes) .and_then(fst::Set::from_bytes)?;
.unwrap();
// we also need to remove all the stop words from the main fst // we also need to remove all the stop words from the main fst
if let Some(word_fst) = main_store.words_fst(writer)? { if let Some(word_fst) = main_store.words_fst(writer)? {
@ -199,11 +207,10 @@ fn apply_stop_words_addition(
.difference(); .difference();
let mut word_fst_builder = SetBuilder::memory(); let mut word_fst_builder = SetBuilder::memory();
word_fst_builder.extend_stream(op).unwrap(); word_fst_builder.extend_stream(op)?;
let word_fst = word_fst_builder let word_fst = word_fst_builder
.into_inner() .into_inner()
.and_then(fst::Set::from_bytes) .and_then(fst::Set::from_bytes)?;
.unwrap();
main_store.put_words_fst(writer, &word_fst)?; main_store.put_words_fst(writer, &word_fst)?;
} }
@ -217,11 +224,10 @@ fn apply_stop_words_addition(
.r#union(); .r#union();
let mut stop_words_builder = SetBuilder::memory(); let mut stop_words_builder = SetBuilder::memory();
stop_words_builder.extend_stream(op).unwrap(); stop_words_builder.extend_stream(op)?;
let stop_words_fst = stop_words_builder let stop_words_fst = stop_words_builder
.into_inner() .into_inner()
.and_then(fst::Set::from_bytes) .and_then(fst::Set::from_bytes)?;
.unwrap();
main_store.put_stop_words_fst(writer, &stop_words_fst)?; main_store.put_stop_words_fst(writer, &stop_words_fst)?;
@ -237,14 +243,13 @@ fn apply_stop_words_deletion(
let mut stop_words_builder = SetBuilder::memory(); let mut stop_words_builder = SetBuilder::memory();
for word in deletion { for word in deletion {
stop_words_builder.insert(&word).unwrap(); stop_words_builder.insert(&word)?;
} }
// create the new delta stop words fst // create the new delta stop words fst
let delta_stop_words = stop_words_builder let delta_stop_words = stop_words_builder
.into_inner() .into_inner()
.and_then(fst::Set::from_bytes) .and_then(fst::Set::from_bytes)?;
.unwrap();
// now we delete all of these stop words from the main store // now we delete all of these stop words from the main store
let stop_words_fst = index.main.stop_words_fst(writer)?.unwrap_or_default(); let stop_words_fst = index.main.stop_words_fst(writer)?.unwrap_or_default();
@ -255,11 +260,8 @@ fn apply_stop_words_deletion(
.difference(); .difference();
let mut stop_words_builder = SetBuilder::memory(); let mut stop_words_builder = SetBuilder::memory();
stop_words_builder.extend_stream(op).unwrap(); stop_words_builder.extend_stream(op)?;
let stop_words_fst = stop_words_builder let stop_words_fst = stop_words_builder.into_inner().and_then(fst::Set::from_bytes)?;
.into_inner()
.and_then(fst::Set::from_bytes)
.unwrap();
Ok(index.main.put_stop_words_fst(writer, &stop_words_fst)?) Ok(index.main.put_stop_words_fst(writer, &stop_words_fst)?)
} }
@ -276,14 +278,14 @@ pub fn apply_synonyms_update(
let mut synonyms_builder = SetBuilder::memory(); let mut synonyms_builder = SetBuilder::memory();
synonyms_store.clear(writer)?; synonyms_store.clear(writer)?;
for (word, alternatives) in synonyms.clone() { for (word, alternatives) in synonyms.clone() {
synonyms_builder.insert(&word).unwrap(); synonyms_builder.insert(&word)?;
let alternatives = { let alternatives = {
let alternatives = SetBuf::from_dirty(alternatives); let alternatives = SetBuf::from_dirty(alternatives);
let mut alternatives_builder = SetBuilder::memory(); let mut alternatives_builder = SetBuilder::memory();
alternatives_builder.extend_iter(alternatives).unwrap(); alternatives_builder.extend_iter(alternatives)?;
let bytes = alternatives_builder.into_inner().unwrap(); let bytes = alternatives_builder.into_inner()?;
fst::Set::from_bytes(bytes).unwrap() fst::Set::from_bytes(bytes)?
}; };
synonyms_store.put_synonyms(writer, word.as_bytes(), &alternatives)?; synonyms_store.put_synonyms(writer, word.as_bytes(), &alternatives)?;
@ -291,8 +293,7 @@ pub fn apply_synonyms_update(
let synonyms_set = synonyms_builder let synonyms_set = synonyms_builder
.into_inner() .into_inner()
.and_then(fst::Set::from_bytes) .and_then(fst::Set::from_bytes)?;
.unwrap();
main_store.put_synonyms_fst(writer, &synonyms_set)?; main_store.put_synonyms_fst(writer, &synonyms_set)?;