Merge pull request #1292 from ddanat-smm/dev

Add htm/html support for RAG documents

backend/apps/rag/main.py

@@ -21,6 +21,7 @@ from langchain_community.document_loaders import (
     TextLoader,
     PyPDFLoader,
     CSVLoader,
+    BSHTMLLoader,
     Docx2txtLoader,
     UnstructuredEPubLoader,
     UnstructuredWordDocumentLoader,
@@ -114,6 +115,7 @@ class CollectionNameForm(BaseModel):
 class StoreWebForm(CollectionNameForm):
     url: str
 
+
 @app.get("/")
 async def get_status():
     return {
@@ -296,13 +298,18 @@ def store_web(form_data: StoreWebForm, user=Depends(get_current_user)):
 
 
 def store_data_in_vector_db(data, collection_name, overwrite: bool = False) -> bool:
+
     text_splitter = RecursiveCharacterTextSplitter(
         chunk_size=app.state.CHUNK_SIZE,
         chunk_overlap=app.state.CHUNK_OVERLAP,
         add_start_index=True,
     )
     docs = text_splitter.split_documents(data)
-    return store_docs_in_vector_db(docs, collection_name, overwrite)
+
+    if len(docs) > 0:
+        return store_docs_in_vector_db(docs, collection_name, overwrite), None
+    else:
+        raise ValueError(ERROR_MESSAGES.EMPTY_CONTENT)
 
 
 def store_text_in_vector_db(
@@ -318,6 +325,7 @@ def store_text_in_vector_db(
 
 
 def store_docs_in_vector_db(docs, collection_name, overwrite: bool = False) -> bool:
+
     texts = [doc.page_content for doc in docs]
     metadatas = [doc.metadata for doc in docs]
 
@@ -402,6 +410,8 @@ def get_loader(filename: str, file_content_type: str, file_path: str):
         loader = UnstructuredRSTLoader(file_path, mode="elements")
     elif file_ext == "xml":
         loader = UnstructuredXMLLoader(file_path)
+    elif file_ext in ["htm", "html"]:
+        loader = BSHTMLLoader(file_path, open_encoding="unicode_escape")
     elif file_ext == "md":
         loader = UnstructuredMarkdownLoader(file_path)
     elif file_content_type == "application/epub+zip":
@@ -452,6 +462,8 @@ def store_doc(
 
         loader, known_type = get_loader(file.filename, file.content_type, file_path)
         data = loader.load()
+
+        try:
             result = store_data_in_vector_db(data, collection_name)
 
             if result:
@@ -461,10 +473,10 @@ def store_doc(
                     "filename": filename,
                     "known_type": known_type,
                 }
-        else:
+        except Exception as e:
             raise HTTPException(
                 status_code=status.HTTP_500_INTERNAL_SERVER_ERROR,
-                detail=ERROR_MESSAGES.DEFAULT(),
+                detail=e,
             )
     except Exception as e:
         log.exception(e)
@@ -529,6 +541,7 @@ def scan_docs_dir(user=Depends(get_admin_user)):
                 )
                 data = loader.load()
 
+                try:
                     result = store_data_in_vector_db(data, collection_name)
 
                     if result:
@@ -561,6 +574,9 @@ def scan_docs_dir(user=Depends(get_admin_user)):
                                     }
                                 ),
                             )
+                except Exception as e:
+                    print(e)
+                    pass
 
         except Exception as e:
             log.exception(e)

backend/constants.py

@@ -60,3 +60,5 @@ class ERROR_MESSAGES(str, Enum):
     MODEL_NOT_FOUND = lambda name="": f"Model '{name}' was not found"
     OPENAI_NOT_FOUND = lambda name="": f"OpenAI API was not found"
     OLLAMA_NOT_FOUND = "WebUI could not connect to Ollama"
+
+    EMPTY_CONTENT = "The content provided is empty. Please ensure that there is text or data present before proceeding."

src/lib/constants.ts

@@ -22,6 +22,7 @@ export const SUPPORTED_FILE_TYPE = [
 	'text/plain',
 	'text/csv',
 	'text/xml',
+	'text/html',
 	'text/x-python',
 	'text/css',
 	'application/vnd.openxmlformats-officedocument.wordprocessingml.document',
@@ -50,6 +51,8 @@ export const SUPPORTED_FILE_EXTENSIONS = [
 	'h',
 	'c',
 	'cs',
+	'htm',
+	'html',
 	'sql',
 	'log',
 	'ini',