Update app.py

app.py

@@ -75,33 +75,59 @@ def set_custom_css():
     """, unsafe_allow_html=True)
 
 def get_docparser_data(file, api_key, parser_id) -> Optional[dict]:
-    upload_url = f"https://api.docparser.com/v1/document/upload/{parser_id}"
-    auth = (api_key, '')  # Use HTTP Basic Auth with the API key as username and empty password
-    files = {'file': file}
+    # First, upload the document
+    upload_url = "https://api.docparser.com/v1/document/upload"
+    headers = {
+        'Authorization': f'Basic {api_key}'
+    }
+    files = {
+        'file': (file.name, file, 'application/pdf')
+    }
+    params = {
+        'parser_id': parser_id
+    }
+    
     try:
-        # Upload the document
-        response = requests.post(upload_url, auth=auth, files=files)
-        response.raise_for_status()
-        document_id = response.json().get('id')
-
-        # Ensure document ID is valid
+        # Upload document
+        upload_response = requests.post(
+            upload_url, 
+            headers=headers,
+            files=files,
+            params=params
+        )
+        upload_response.raise_for_status()
+        
+        # Get document ID from upload response
+        upload_data = upload_response.json()
+        if not isinstance(upload_data, list) or len(upload_data) == 0:
+            st.error("Invalid response from Docparser upload")
+            return None
+            
+        document_id = upload_data[0].get('id')
         if not document_id:
-            st.error("Failed to retrieve document ID from Docparser.")
+            st.error("Failed to get document ID from upload response")
             return None
 
-        # Fetch parsed data
-        result_url = f"https://api.docparser.com/v1/results/{parser_id}/{document_id}"
-        result_response = requests.get(result_url, auth=auth)
-        result_response.raise_for_status()
-        data = result_response.json()
-
-        # Check if the response is a list and handle accordingly
-        if isinstance(data, list) and len(data) > 0:
-            data = data[0]  # Assuming you want the first result
+        # Get parsed results
+        results_url = f"https://api.docparser.com/v1/results/{parser_id}/{document_id}"
+        results_response = requests.get(
+            results_url,
+            headers=headers
+        )
+        results_response.raise_for_status()
+        
+        # Handle results
+        results_data = results_response.json()
+        if isinstance(results_data, list) and len(results_data) > 0:
+            return results_data[0]
+        else:
+            st.error("No parsed data received from Docparser")
+            return None
 
-        return data
     except requests.exceptions.HTTPError as http_err:
         st.error(f"HTTP error occurred: {http_err}")
+        if http_err.response is not None:
+            st.error(f"Response content: {http_err.response.content}")
     except Exception as e:
         st.error(f"Error fetching data from Docparser: {e}")
     return None