Skip to content
Open
Show file tree
Hide file tree
Changes from 1 commit
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
Prev Previous commit
efficient filtering instead of slow
  • Loading branch information
Jay Bhambhani committed Jun 17, 2025
commit 8fcb623a4088f98301ee1cad020a981176e6acbc
56 changes: 39 additions & 17 deletions docarray/index/backends/opensearchv2.py
Original file line number Diff line number Diff line change
Expand Up @@ -163,10 +163,6 @@ def build(self, *args, **kwargs) -> Any:

if len(self._query['query']) == 0:
del self._query['query']
elif 'knn' in self._query:
self._query['knn']['filter'] = self._query['query']
del self._query['query']

return self._query

def find(
Expand All @@ -191,15 +187,11 @@ def find(
else:
query_vec = query
query_vec_np = BaseDocIndex._to_numpy(self._outer_instance, query_vec)
self._query['query']['bool']['must'].append(
{
'knn': self._outer_instance._form_search_body(
query_vec_np,
search_field,
limit,
)['query']['knn']
}
)
self._query['query']['knn'] = self._outer_instance._form_search_body(
query_vec_np,
search_field,
limit,
)['query']['knn']
self._query['size'] = limit

return self
Expand All @@ -216,7 +208,22 @@ def filter(self, query: Dict[str, Any], limit: int = 10):
self._outer_instance._logger.debug('Executing filter query')

self._query['size'] = limit
self._query['query']['bool']['filter'].append(query)
if self._query['query'].get('knn'):
search_field = next(iter(self._query['query']['knn']))
if self._query['query']['knn'][search_field].get('filter') is not None:
self._query['query']['knn'][search_field]['filter']['bool'][
'must'
].append(query)
else:
self._query['query']['knn'][search_field]['filter'] = {}
self._query['query']['knn'][search_field]['filter']['bool'] = {
'must': []
}
self._query['query']['knn'][search_field]['filter']['bool'][
'must'
].append(query)
else:
self._query['query']['bool']['filter'].append(query)
return self

def text_search(self, query: str, search_field: str = 'text', limit: int = 10):
Expand All @@ -231,9 +238,24 @@ def text_search(self, query: str, search_field: str = 'text', limit: int = 10):

self._outer_instance._validate_search_field(search_field)
self._query['size'] = limit
self._query['query']['bool']['must'].append(
{'match': {search_field: query}}
)
if self._query['query'].get('knn'):
vector_field = next(iter(self._query['query']['knn']))
if self._query['query']['knn'][vector_field].get('filter') is not None:
self._query['query']['knn'][vector_field]['filter']['bool'][
'must'
].append({'term': {search_field: query}})
else:
self._query['query']['knn'][vector_field]['filter'] = {}
self._query['query']['knn'][vector_field]['filter']['bool'] = {
'must': []
}
self._query['query']['knn'][vector_field]['filter']['bool'][
'must'
].append({'term': {search_field: query}})
else:
self._query['query']['bool']['must'].append(
{'match': {search_field: query}}
)
return self

find_batched = _raise_not_composable('find_batched')
Expand Down
53 changes: 25 additions & 28 deletions tests/index/opensearch/test_find_os.py
Original file line number Diff line number Diff line change
Expand Up @@ -306,44 +306,41 @@ class MyDoc(BaseDoc):
# combination
q = (
index.build_query()
.filter({'range': {'num': {'lte': 3}}})
.find(index_docs[-1], search_field='tens')
.filter({'range': {'num': {'lte': 3}}})
.text_search('0', search_field='text')
.build()
)
print('Q', q)
docs, _ = index.execute_query(q)
assert [doc['id'] for doc in docs] == ['0', '1']
assert [doc['id'] for doc in docs] == ['1', '0']

# direct
query = {
'query': {
'bool': {
'filter': [
{'range': {'num': {'gte': 2}}},
{'range': {'num': {'lte': 3}}},
],
'must': [
{
'knn': {
'tens': {
'vector': [
9.0,
9.0,
9.0,
9.0,
9.0,
9.0,
9.0,
9.0,
9.0,
9.0,
],
'k': 10,
}
'knn': {
'tens': {
'vector': [
9.0,
9.0,
9.0,
9.0,
9.0,
9.0,
9.0,
9.0,
9.0,
9.0,
],
'k': 10,
'filter': {
'bool': {
'must': [
{'range': {'num': {'gte': 2}}},
{'range': {'num': {'lte': 3}}},
]
}
}
],
},
},
}
}
}
Expand Down
Loading