Skip iterating over documents when the faceted field list doesn't change

This commit is contained in:
ManyTheFish 2024-06-04 15:31:24 +02:00
parent 17c5ceeb9d
commit d29d4f88da

View File

@ -66,11 +66,6 @@ pub fn extract_fid_docid_facet_values<R: io::Read + io::Seek>(
max_memory.map(|m| m / 2), max_memory.map(|m| m / 2),
); );
let old_faceted_fids: BTreeSet<_> =
settings_diff.old.faceted_fields_ids.iter().copied().collect();
let new_faceted_fids: BTreeSet<_> =
settings_diff.new.faceted_fields_ids.iter().copied().collect();
// The tuples represents the Del and Add side for a bitmap // The tuples represents the Del and Add side for a bitmap
let mut facet_exists_docids = BTreeMap::<FieldId, (RoaringBitmap, RoaringBitmap)>::new(); let mut facet_exists_docids = BTreeMap::<FieldId, (RoaringBitmap, RoaringBitmap)>::new();
let mut facet_is_null_docids = BTreeMap::<FieldId, (RoaringBitmap, RoaringBitmap)>::new(); let mut facet_is_null_docids = BTreeMap::<FieldId, (RoaringBitmap, RoaringBitmap)>::new();
@ -80,6 +75,12 @@ pub fn extract_fid_docid_facet_values<R: io::Read + io::Seek>(
let mut numbers_key_buffer = Vec::new(); let mut numbers_key_buffer = Vec::new();
let mut strings_key_buffer = Vec::new(); let mut strings_key_buffer = Vec::new();
let old_faceted_fids: BTreeSet<_> =
settings_diff.old.faceted_fields_ids.iter().copied().collect();
let new_faceted_fids: BTreeSet<_> =
settings_diff.new.faceted_fields_ids.iter().copied().collect();
if !settings_diff.settings_update_only || old_faceted_fids != new_faceted_fids {
let mut cursor = obkv_documents.into_cursor()?; let mut cursor = obkv_documents.into_cursor()?;
while let Some((docid_bytes, value)) = cursor.move_on_next()? { while let Some((docid_bytes, value)) = cursor.move_on_next()? {
let obkv = obkv::KvReader::new(value); let obkv = obkv::KvReader::new(value);
@ -93,7 +94,9 @@ pub fn extract_fid_docid_facet_values<R: io::Read + io::Seek>(
}; };
// iterate over the faceted fields instead of over the whole document. // iterate over the faceted fields instead of over the whole document.
for eob in for eob in
merge_join_by(old_faceted_fids.iter(), new_faceted_fids.iter(), |old, new| old.cmp(new)) merge_join_by(old_faceted_fids.iter(), new_faceted_fids.iter(), |old, new| {
old.cmp(new)
})
{ {
let (field_id, del_value, add_value) = match eob { let (field_id, del_value, add_value) = match eob {
EitherOrBoth::Left(&field_id) => { EitherOrBoth::Left(&field_id) => {
@ -253,6 +256,7 @@ pub fn extract_fid_docid_facet_values<R: io::Read + io::Seek>(
} }
} }
} }
}
let mut buffer = Vec::new(); let mut buffer = Vec::new();
let mut facet_exists_docids_writer = create_writer( let mut facet_exists_docids_writer = create_writer(