snowykami/meilisearch
1
0
Fork 0
mirror of https://github.com/meilisearch/meilisearch.git synced 2025-10-31 07:56:28 +00:00
Code Issues Packages Projects Releases Wiki Activity

Do not index a filed count when no word is counted

Browse Source
This commit is contained in:
ManyTheFish
2024-12-09 15:45:12 +01:00
parent 71f59749dc
commit 07f42e8057
1 changed files with 11 additions and 7 deletions
Download Patch File Download Diff File Expand all files Collapse all files

18
crates/milli/src/update/new/extract/searchable/extract_word_docids.rs
View File

@@ -28,7 +28,7 @@ pub struct WordDocidsBalancedCaches<'extractor> {
exact_word_docids: BalancedCaches<'extractor>,
word_position_docids: BalancedCaches<'extractor>,
fid_word_count_docids: BalancedCaches<'extractor>,
fid_word_count: HashMap<FieldId, (usize, usize)>,
fid_word_count: HashMap<FieldId, (Option<usize>, Option<usize>)>,
current_docid: Option<DocumentId>,
}
@@ -85,8 +85,8 @@ impl<'extractor> WordDocidsBalancedCaches<'extractor> {
self.fid_word_count
.entry(field_id)
.and_modify(|(_current_count, new_count)| *new_count += 1)
.or_insert((0, 1));
.and_modify(|(_current_count, new_count)| *new_count.get_or_insert(0) += 1)
.or_insert((None, Some(1)));
self.current_docid = Some(docid);
Ok(())
@@ -130,8 +130,8 @@ impl<'extractor> WordDocidsBalancedCaches<'extractor> {
self.fid_word_count
.entry(field_id)
.and_modify(|(current_count, _new_count)| *current_count += 1)
.or_insert((1, 0));
.and_modify(|(current_count, _new_count)| *current_count.get_or_insert(0) += 1)
.or_insert((Some(1), None));
self.current_docid = Some(docid);
@@ -141,14 +141,18 @@ impl<'extractor> WordDocidsBalancedCaches<'extractor> {
fn flush_fid_word_count(&mut self, buffer: &mut BumpVec<u8>) -> Result<()> {
for (fid, (current_count, new_count)) in self.fid_word_count.drain() {
if current_count != new_count {
if current_count <= MAX_COUNTED_WORDS {
if let Some(current_count) =
current_count.filter(|current_count| *current_count <= MAX_COUNTED_WORDS)
{
buffer.clear();
buffer.extend_from_slice(&fid.to_be_bytes());
buffer.push(current_count as u8);
self.fid_word_count_docids
.insert_del_u32(buffer, self.current_docid.unwrap())?;
}
if new_count <= MAX_COUNTED_WORDS {
if let Some(new_count) =
new_count.filter(|new_count| *new_count <= MAX_COUNTED_WORDS)
{
buffer.clear();
buffer.extend_from_slice(&fid.to_be_bytes());
buffer.push(new_count as u8);