Update app.py

app.py

@@ -12,16 +12,22 @@ def upload_pdf_to_unstract(pdf_file):
     headers = {
         "unstract-key": UNSTRACT_API_KEY,
     }
-    # Always reset file pointer and use "application/pdf"
+    # Always reset file pointer
     pdf_file.seek(0)
     file_bytes = pdf_file.read()
+
+    # Force correct .pdf extension and type
+    file_name = getattr(pdf_file, "name", None)
+    if not file_name or not file_name.lower().endswith(".pdf"):
+        file_name = "invoice.pdf"
+
     files = {
-        "file": (
-            pdf_file.name if pdf_file.name.lower().endswith(".pdf") else "invoice.pdf",
-            file_bytes,
-            "application/pdf"
-        ),
+        "file": (file_name, file_bytes, "application/pdf"),
     }
+
+    # Debug print for troubleshooting
+    # st.write("Uploading file with name:", file_name)
+
     with st.spinner("Uploading and starting OCR..."):
         resp = requests.post(url, headers=headers, files=files)
     if resp.status_code not in (200, 202):
@@ -73,15 +79,12 @@ st.title("Unstract OCR: PDF Invoice Text Extraction")
 uploaded_pdf = st.file_uploader("Upload Invoice PDF", type="pdf")
 
 if st.button("Extract Text from PDF") and uploaded_pdf:
-    # Step 1: Upload PDF and get whisper_hash
     whisper_hash = upload_pdf_to_unstract(uploaded_pdf)
     if not whisper_hash:
         st.stop()
     st.success(f"File accepted. Tracking hash: {whisper_hash}")
 
-    # Step 2: Poll until processed
     if poll_until_processed(whisper_hash):
-        # Step 3: Retrieve text
         text = retrieve_text(whisper_hash)
         if text:
             st.success("Text extraction complete!")