Update auditqa/doc_process.py

auditqa/doc_process.py

@@ -27,6 +27,14 @@ def process_pdf():
             chunk_overlap=int(chunk_size / 10),
             add_start_index=True,
             strip_whitespace=True,
-            separators=["\n\n", "\n", ".", " ", ""],
+            separators=["\n\n", "\n"],
     )
-    
+    all_documents = {}
+    for file,value in docs.items():
+        doc_processed = text_splitter.split_documents(value)
+        for doc in doc_processed:
+            doc.metadata["source"] = file
+        all_documents[file] = doc_processed
+        
+    print(all_documents.keys())
+    print(all_documents['ABC'])