Update app.py

app.py

@@ -185,7 +185,6 @@ def search(prompt: str):
     from haystack.components.generators import OpenAIGenerator
     from haystack.utils import Secret    
     from haystack.components.builders import PromptBuilder
-    from QueryMetadataExtractor import QueryMetadataExtractor
     
     start_time = time.time()
     
@@ -211,14 +210,12 @@ def search(prompt: str):
         model="meta-llama-3-8b-instruct",
         generation_kwargs = {"max_tokens": 512}
     )
-    metadata_extractor = QueryMetadataExtractor()
 
     querying = Pipeline()
     querying.add_component("sparse_text_embedder", FastembedSparseTextEmbedder(model="Qdrant/bm42-all-minilm-l6-v2-attentions"))
     querying.add_component("dense_text_embedder", FastembedTextEmbedder(
      model="sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2", prefix="Đại diện cho câu này để tìm kiếm các đoạn văn có liên quan: ")
      )
-    querying.add_component(instance=metadata_extractor, name="metadata_extractor")
     querying.add_component("retriever", QdrantHybridRetriever(document_store=document_store))
     querying.add_component("document_joiner", DocumentJoiner())
     querying.add_component("ranker", TransformersSimilarityRanker(model="BAAI/bge-m3"))    
@@ -227,18 +224,15 @@ def search(prompt: str):
     
     querying.connect("sparse_text_embedder.sparse_embedding", "retriever.query_sparse_embedding")
     querying.connect("dense_text_embedder.embedding", "retriever.query_embedding")
-    querying.connect("metadata_extractor.filters", "retriever.filters")
     querying.connect("retriever", "document_joiner")
     querying.connect("document_joiner", "ranker")
     querying.connect("ranker.documents", "prompt_builder.documents")
     querying.connect("prompt_builder", "llm")
     querying.debug=True
-    metadata_fields =  {"publish_date", "publisher", "document_type"}
     results = querying.run(
         {
             "dense_text_embedder": {"text": prompt},
             "sparse_text_embedder": {"text": prompt},
-            "metadata_extractor": {"query": prompt, "metadata_fields": metadata_fields},
             "ranker": {"query": prompt},
             "prompt_builder": {"question": prompt}
         }
@@ -282,6 +276,15 @@ def truncate_text(text: str) -> str:
     else:
         return text[:3000]
         
+@app.post("/query2metadata/")
+async def extract_metadata_from_query(query: str):
+    from QueryMetadataExtractor import QueryMetadataExtractor
+    
+    extractor = QueryMetadataExtractor()
+    metadata_fields = {"publisher", "publish_date", "document_type"}
+    
+    result = extractor.run(query, metadata_fields)
+    
 @app.post("/pdf2text/")
 async def convert_upload_file(file: UploadFile = File(...)):
     import pytesseract