fix: url knowledge multiple brain (#3145)

# Description

- Find knowledge path in storage

backend/api/quivr_api/modules/knowledge/controller/knowledge_routes.py

@@ -95,13 +95,13 @@ async def generate_signed_url_endpoint(
 
     knowledge = await knowledge_service.get_knowledge(knowledge_id)
 
-    if len(knowledge.brains_ids) == 0:
+    if len(knowledge.brain_ids) == 0:
         raise HTTPException(
             status_code=HTTPStatus.NOT_FOUND,
             detail="knowledge not associated with brains yet.",
         )
 
-    brain_id = knowledge.brains_ids[0]
+    brain_id = knowledge.brain_ids[0]
 
     validate_brain_authorization(brain_id=brain_id, user_id=current_user.id)
 

backend/api/quivr_api/modules/knowledge/repository/knowledges.py

@@ -100,6 +100,22 @@ class KnowledgeRepository(BaseRepository):
 
         return knowledge
 
+    async def get_knowledge_by_file_name_brain_id(
+        self, file_name: str, brain_id: UUID
+    ) -> KnowledgeDB:
+        query = (
+            select(KnowledgeDB)
+            .where(KnowledgeDB.file_name == file_name)
+            .where(KnowledgeDB.brains.any(brain_id=brain_id))  # type: ignore
+        )
+
+        result = await self.session.exec(query)
+        knowledge = result.first()
+        if not knowledge:
+            raise NoResultFound("Knowledge not found")
+
+        return knowledge
+
     async def get_knowledge_by_sha1(self, sha1: str) -> KnowledgeDB:
         query = select(KnowledgeDB).where(KnowledgeDB.file_sha1 == sha1)
         result = await self.session.exec(query)

backend/api/quivr_api/modules/knowledge/service/knowledge_service.py

@@ -18,6 +18,7 @@ from quivr_api.modules.sync.entity.sync_models import (
     DownloadedSyncFile,
     SyncFile,
 )
+from quivr_api.modules.upload.service.upload_file import check_file_exists
 
 logger = get_logger(__name__)
 
@@ -35,6 +36,24 @@ class KnowledgeService(BaseService[KnowledgeRepository]):
         km = await km.to_dto()
         return km
 
+    # TODO: this is temporary fix for getting knowledge path.
+    # KM storage path should be unrelated to brain
+    async def get_knowledge_storage_path(
+        self, file_name: str, brain_id: UUID
+    ) -> str | None:
+        try:
+            km = await self.repository.get_knowledge_by_file_name_brain_id(
+                file_name, brain_id
+            )
+            brains = await km.awaitable_attrs.brains
+            return next(
+                f"{b.brain_id}/{file_name}"
+                for b in brains
+                if check_file_exists(b.brain_id, file_name)
+            )
+        except NoResultFound:
+            raise FileNotFoundError(f"No knowledge for file_name: {file_name}")
+
     async def get_knowledge(self, knowledge_id: UUID) -> Knowledge:
         inserted_knowledge_db_instance = await self.repository.get_knowledge_by_id(
             knowledge_id

backend/api/quivr_api/modules/knowledge/tests/test_knowledges.py

@@ -12,6 +12,7 @@ from quivr_api.modules.knowledge.entity.knowledge import KnowledgeDB
 from quivr_api.modules.knowledge.entity.knowledge_brain import KnowledgeBrain
 from quivr_api.modules.knowledge.repository.knowledges import KnowledgeRepository
 from quivr_api.modules.knowledge.service.knowledge_service import KnowledgeService
+from quivr_api.modules.upload.service.upload_file import upload_file_storage
 from quivr_api.modules.user.entity.user_identity import User
 from quivr_api.vector.entity.vector import Vector
 from sqlalchemy.exc import IntegrityError, NoResultFound
@@ -96,8 +97,8 @@ async def test_data(session: AsyncSession) -> TestData:
     )
 
     knowledge_brain_1 = KnowledgeDB(
-        file_name="test_file_1",
-        extension="txt",
+        file_name="test_file_1.txt",
+        extension=".txt",
         status="UPLOADED",
         source="test_source",
         source_link="test_source_link",
@@ -108,8 +109,8 @@ async def test_data(session: AsyncSession) -> TestData:
     )
 
     knowledge_brain_2 = KnowledgeDB(
-        file_name="test_file_2",
-        extension="txt",
+        file_name="test_file_2.txt",
+        extension=".txt",
         status="UPLOADED",
         source="test_source",
         source_link="test_source_link",
@@ -349,7 +350,7 @@ async def test_should_process_knowledge_link_brain(
     assert brain.brain_id
     prev = KnowledgeDB(
         file_name="prev",
-        extension="txt",
+        extension=".txt",
         status=KnowledgeStatus.UPLOADED,
         source="test_source",
         source_link="test_source_link",
@@ -465,3 +466,29 @@ async def test_should_process_knowledge_prev_error(
     assert new.id
     new = await service.repository.get_knowledge_by_id(new.id)
     assert new.file_sha1
+
+
+@pytest.mark.asyncio
+async def test_get_knowledge_storage_path(session: AsyncSession, test_data: TestData):
+    brain, [knowledge, _] = test_data
+    assert knowledge.file_name
+    repository = KnowledgeRepository(session)
+    service = KnowledgeService(repository)
+    brain_2 = Brain(
+        name="test_brain",
+        description="this is a test brain",
+        brain_type=BrainType.integration,
+    )
+    session.add(brain_2)
+    await session.commit()
+    await session.refresh(brain_2)
+    assert brain_2.brain_id
+    km_data = os.urandom(128)
+    km_path = f"{str(knowledge.brains[0].brain_id)}/{knowledge.file_name}"
+    await upload_file_storage(km_data, km_path)
+    # Link knowledge to two brains
+    await repository.link_to_brain(knowledge, brain_2.brain_id)
+    storage_path = await service.get_knowledge_storage_path(
+        knowledge.file_name, brain_2.brain_id
+    )
+    assert storage_path == km_path

backend/api/quivr_api/modules/rag_service/rag_service.py

@@ -263,10 +263,11 @@ class RAGService:
             streamed_chat_history.metadata["snippet_emoji"] = (
                 self.brain.snippet_emoji if self.brain else None
             )
-        sources_urls = generate_source(
-            response.metadata.sources,
-            self.brain.brain_id,
-            (
+        sources_urls = await generate_source(
+            knowledge_service=self.knowledge_service,
+            brain_id=self.brain.brain_id,
+            source_documents=response.metadata.sources,
+            citations=(
                 streamed_chat_history.metadata["citations"]
                 if streamed_chat_history.metadata
                 else None

backend/api/quivr_api/modules/rag_service/utils.py

@@ -3,6 +3,7 @@ from typing import Any, List
 from uuid import UUID
 
 from quivr_api.modules.chat.dto.chats import Sources
+from quivr_api.modules.knowledge.service.knowledge_service import KnowledgeService
 from quivr_api.modules.upload.service.generate_file_signed_url import (
     generate_file_signed_url,
 )
@@ -11,9 +12,10 @@ logger = logging.getLogger(__name__)
 
 
 # TODO: REFACTOR THIS, it does call the DB , so maybe in a service
-def generate_source(
-    source_documents: List[Any] | None,
+async def generate_source(
+    knowledge_service: KnowledgeService,
     brain_id: UUID,
+    source_documents: List[Any] | None,
     citations: List[int] | None = None,
 ) -> List[Sources]:
     """
@@ -62,8 +64,11 @@ def generate_source(
             if is_url:
                 source_url = doc.metadata["original_file_name"]
             else:
-                file_path = f"{brain_id}/{doc.metadata['file_name']}"
                 # Check if the URL has already been generated
+                file_name = doc.metadata["file_name"]
+                file_path = await knowledge_service.get_knowledge_storage_path(
+                    file_name=file_name, brain_id=brain_id
+                )
                 if file_path in generated_urls:
                     source_url = generated_urls[file_path]
                 else:

backend/api/quivr_api/modules/sync/tests/conftest.py

@@ -77,6 +77,16 @@ from quivr_api.modules.user.entity.user_identity import User
 pg_database_base_url = "postgres:postgres@localhost:54322/postgres"
 
 
+@pytest.fixture(scope="module")
+def event_loop():
+    try:
+        loop = asyncio.get_running_loop()
+    except RuntimeError:
+        loop = asyncio.new_event_loop()
+    yield loop
+    loop.close()
+
+
 @pytest.fixture(scope="function")
 def page_response() -> dict[str, Any]:
     json_path = (
@@ -182,17 +192,7 @@ def fetch_response():
 
 
 @pytest.fixture(scope="session")
-def event_loop():
-    try:
-        loop = asyncio.get_running_loop()
-    except RuntimeError:
-        loop = asyncio.new_event_loop()
-    yield loop
-    loop.close()
-
-
-@pytest_asyncio.fixture(scope="session")
-async def sync_engine():
+def sync_engine():
     engine = create_engine(
         "postgresql://" + pg_database_base_url,
         echo=True if os.getenv("ORM_DEBUG") else False,
@@ -204,8 +204,8 @@ async def sync_engine():
     yield engine
 
 
-@pytest_asyncio.fixture()
-async def sync_session(sync_engine):
+@pytest.fixture
+def sync_session(sync_engine):
     with sync_engine.connect() as conn:
         conn.begin()
         conn.begin_nested()
@@ -273,7 +273,9 @@ def search_result():
     ]
 
 
-@pytest_asyncio.fixture(scope="session")
+@pytest_asyncio.fixture(
+    scope="session",
+)
 async def async_engine():
     engine = create_async_engine(
         "postgresql+asyncpg://" + pg_database_base_url,