Update app.py via AI Editor

app.py

@@ -8,6 +8,8 @@ import pandas as pd
 import openai
 import logging
 from docx import Document
+import requests
+import mimetypes
 
 logging.basicConfig(
     level=logging.INFO,
@@ -44,15 +46,56 @@ def decode_document(decoded_bytes):
             logging.error("Document decode failed for both utf-8 and latin-1: %s", e)
             return None
 
+def guess_mime_type(filename):
+    mime_type, _ = mimetypes.guess_type(filename)
+    if not mime_type:
+        ext = filename.lower().split('.')[-1]
+        if ext == "pdf":
+            mime_type = "application/pdf"
+        elif ext == "docx":
+            mime_type = "application/vnd.openxmlformats-officedocument.wordprocessingml.document"
+        elif ext == "doc":
+            mime_type = "application/msword"
+        elif ext == "xlsx":
+            mime_type = "application/vnd.openxmlformats-officedocument.spreadsheetml.sheet"
+        elif ext == "xls":
+            mime_type = "application/vnd.ms-excel"
+        else:
+            mime_type = "application/octet-stream"
+    return mime_type
+
 def upload_to_openai_file(decoded_bytes, filename):
     try:
-        memfile = io.BytesIO(decoded_bytes)
-        memfile.seek(0)
-        resp = openai.files.create(file=memfile, purpose="assistants", file_name=filename)
-        logging.info(f"File uploaded to OpenAI: {filename}, file_id: {resp.id}")
-        return resp.id
+        url = "https://api.openai.com/v1/uploads"
+        mime_type = guess_mime_type(filename)
+        headers = {
+            "Authorization": f"Bearer {OPENAI_API_KEY}"
+        }
+        files = {
+            "file": (filename, io.BytesIO(decoded_bytes), mime_type)
+        }
+        data = {
+            "bytes": str(len(decoded_bytes)),
+            "filename": filename,
+            "mime_type": mime_type,
+            "purpose": "assistants"
+        }
+        logging.info(f"Uploading file to OpenAI /v1/uploads: {filename}, size={len(decoded_bytes)}, mime_type={mime_type}")
+        resp = requests.post(url, headers=headers, files=files, data=data)
+        if resp.status_code == 200:
+            resp_json = resp.json()
+            file_id = resp_json.get("id")
+            if file_id:
+                logging.info(f"File uploaded to OpenAI uploads: {filename}, file_id: {file_id}")
+                return file_id
+            else:
+                logging.error(f"OpenAI /v1/uploads did not return file id. Response: {resp_json}")
+                return None
+        else:
+            logging.error(f"Failed uploading file to OpenAI API /v1/uploads: {resp.status_code}, {resp.text}")
+            return None
     except Exception as e:
-        logging.error(f"Failed uploading file to OpenAI API: {e}")
+        logging.error(f"Exception during file upload to OpenAI /v1/uploads: {e}")
         return None
 
 def openai_stream_generate(prompt, file_id=None):