Update main.py

main.py

@@ -27,27 +27,25 @@ class QueryRequest(BaseModel):
 
 def _unpack_faiss(src_path: str, extract_to: str) -> str:
     """
-    If src_path is a .zip, unzip to extract_to and return the directory
-    containing the .faiss file.  If it's already a folder, just return it.
+    If src_path is a valid .zip archive, unzip it into extract_to and
+    return the subdirectory that contains the .faiss index.
+    If src_path is already a directory, return it directly.
     """
-    # 1) ZIP case
-    if src_path.lower().endswith(".zip"):
-        if not os.path.isfile(src_path):
-            raise FileNotFoundError(f"Could not find zip file: {src_path}")
+    # 1) True ZIP file?
+    if zipfile.is_zipfile(src_path):
         with zipfile.ZipFile(src_path, "r") as zf:
             zf.extractall(extract_to)
-
-        # walk until we find any .faiss file
+        # scan until we find any .faiss file
         for root, _, files in os.walk(extract_to):
-            if any(fn.endswith(".faiss") for fn in files):
+            if any(f.endswith(".faiss") for f in files):
                 return root
-        raise RuntimeError(f"No .faiss index found inside {src_path}")
+        raise RuntimeError(f"No .faiss index found inside ZIP: {src_path}")
 
-    # 2) directory case
+    # 2) Already a folder?
     if os.path.isdir(src_path):
         return src_path
 
-    raise RuntimeError(f"Path is neither a .zip nor a directory: {src_path}")
+    raise RuntimeError(f"Path is neither a valid ZIP nor a directory: {src_path}")
 
 
 @app.on_event("startup")
@@ -61,7 +59,6 @@ def load_components():
         max_tokens=1024,
         api_key=os.getenv("api_key"),
     )
-
     embeddings = HuggingFaceEmbeddings(
         model_name="intfloat/multilingual-e5-large",
         model_kwargs={"device": "cpu"},
@@ -69,28 +66,24 @@ def load_components():
     )
 
     # --- 2) Load & merge two FAISS indexes ---
-    src1 = "faiss_index.zip"
-    src2 = "faiss_index_extra.zip"
+    # (these can be either real .zip files or existing folders)
+    src1 = "faiss_index.zip"          # or "faiss_index" if it's already a folder
+    src2 = "faiss_index_extra.zip"    # or "faiss_index_extra"
 
-    # Use TemporaryDirectory objects so they stick around until program exit
     tmp1 = tempfile.TemporaryDirectory()
     tmp2 = tempfile.TemporaryDirectory()
 
-    # Unpack & locate
     dir1 = _unpack_faiss(src1, tmp1.name)
     dir2 = _unpack_faiss(src2, tmp2.name)
 
-    # Load them
     vs1 = FAISS.load_local(dir1, embeddings, allow_dangerous_deserialization=True)
     vs2 = FAISS.load_local(dir2, embeddings, allow_dangerous_deserialization=True)
 
-    # Merge vs2 into vs1
     vs1.merge_from(vs2)
     vectorstore = vs1
 
     # --- 3) Build retriever & QA chain ---
     retriever = vectorstore.as_retriever(search_kwargs={"k": 5})
-
     prompt = PromptTemplate(
         template="""
 You are an expert assistant on Islamic knowledge.  
@@ -108,7 +101,6 @@ Your response:
 """,
         input_variables=["context", "question"],
     )
-
     chain = RetrievalQA.from_chain_type(
         llm=llm,
         chain_type="stuff",