UniqAI/open-webui
1
0
Fork 0
mirror of https://github.com/open-webui/open-webui synced 2025-04-30 03:02:06 +00:00
Code Issues Actions 4 Packages Projects Releases Wiki Activity

refac

Browse Source
This commit is contained in:
Timothy Jaeryang Baek 2025-04-06 15:45:48 -07:00
parent 9825d03602
commit 155dbd5a66
1 changed files with 11 additions and 10 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

21
backend/open_webui/routers/retrieval.py
View File

@ -174,7 +174,7 @@ class ProcessUrlForm(CollectionNameForm):
url: str url: str
class SearchForm(CollectionNameForm): class SearchForm(BaseModel):
query: str query: str
@ -1464,10 +1464,6 @@ async def process_web_search(
log.debug(f"web_results: {web_results}") log.debug(f"web_results: {web_results}")
try: try:
collection_basename = form_data.collection_name
if collection_basename == "" or collection_basename is None:
collection_basename = "web-search"
urls = [result.link for result in web_results] urls = [result.link for result in web_results]
loader = get_web_loader( loader = get_web_loader(
urls, urls,
@ -1476,7 +1472,9 @@ async def process_web_search(
trust_env=request.app.state.config.RAG_WEB_SEARCH_TRUST_ENV, trust_env=request.app.state.config.RAG_WEB_SEARCH_TRUST_ENV,
) )
docs = await loader.aload() docs = await loader.aload()
urls = [doc.metadata["source"] for doc in docs] # only keep URLs which could be retrieved urls = [
doc.metadata["source"] for doc in docs
] # only keep URLs which could be retrieved
if request.app.state.config.BYPASS_WEB_SEARCH_EMBEDDING_AND_RETRIEVAL: if request.app.state.config.BYPASS_WEB_SEARCH_EMBEDDING_AND_RETRIEVAL:
return { return {
@ -1495,14 +1493,17 @@ async def process_web_search(
else: else:
collection_names = [] collection_names = []
for doc_idx, doc in enumerate(docs): for doc_idx, doc in enumerate(docs):
collection_sha = calculate_sha256_string(f"{form_data.query}-{urls[doc_idx]}") collection_name = f"web-search-{calculate_sha256_string(
doc_collection_name = f"{collection_basename}-{collection_sha}"[:63] f"{form_data.query}-{urls[doc_idx]}"
collection_names.append(doc_collection_name) )}"[:63]
collection_names.append(collection_name)
await run_in_threadpool( await run_in_threadpool(
save_docs_to_vector_db, save_docs_to_vector_db,
request, request,
[doc], [doc],
doc_collection_name, collection_name,
overwrite=True, overwrite=True,
user=user, user=user,
) )