AI/dify
1
0
Fork 0
mirror of https://git.mirrors.martin98.com/https://github.com/langgenius/dify.git synced 2025-08-14 21:35:53 +08:00
Code Issues Packages Projects Releases Wiki Activity

fix some tidb bugs (#4960)

Browse Source
This commit is contained in:
Jyong 2024-06-05 19:14:18 +08:00 committed by GitHub
parent 64c8093c1e
commit 02e4de5166
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 8 additions and 8 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

16
api/core/rag/datasource/vdb/tidb_vector/tidb_vector.py
View File

@ -79,8 +79,6 @@ class TiDBVector(BaseVector):
return
with Session(self._engine) as session:
session.begin()
drop_statement = sql_text(f"""DROP TABLE IF EXISTS {self._collection_name}; """)
session.execute(drop_statement)
create_statement = sql_text(f"""
CREATE TABLE IF NOT EXISTS {self._collection_name} (
id CHAR(36) PRIMARY KEY,
@ -123,7 +121,7 @@ class TiDBVector(BaseVector):
def text_exists(self, id: str) -> bool:
result = self.get_ids_by_metadata_field('doc_id', id)
return len(result) > 0
return bool(result)
def delete_by_ids(self, ids: list[str]) -> None:
with Session(self._engine) as session:
@ -184,14 +182,14 @@ class TiDBVector(BaseVector):
docs = []
if self._distance_func == 'l2':
tidb_func = 'Vec_l2_distance'
elif self._distance_func == 'l2':
elif self._distance_func == 'cosine':
tidb_func = 'Vec_Cosine_distance'
else:
tidb_func = 'Vec_Cosine_distance'
with Session(self._engine) as session:
select_statement = sql_text(
f"""SELECT meta, text FROM (
f"""SELECT meta, text, distance FROM (
SELECT meta, text, {tidb_func}(vector, "{query_vector_str}") as distance
FROM {self._collection_name}
ORDER BY distance
@ -199,9 +197,11 @@ class TiDBVector(BaseVector):
) t WHERE distance < {distance};"""
)
res = session.execute(select_statement)
results = [(row[0], row[1]) for row in res]
for meta, text in results:
docs.append(Document(page_content=text, metadata=json.loads(meta)))
results = [(row[0], row[1], row[2]) for row in res]
for meta, text, distance in results:
metadata = json.loads(meta)
metadata['score'] = 1 - distance
docs.append(Document(page_content=text, metadata=metadata))
return docs
def search_by_full_text(self, query: str, **kwargs: Any) -> list[Document]: