Enhance email processing: improve logging for existing email checks and add unstructured.pytesseract to requirements

app/controllers/mail.py

@@ -76,10 +76,12 @@ def list_emails(service, messages):
     for message in messages:
         msg = service.users().messages().get(userId="me", id=message["id"], format="full").execute()
         metadata = {}
-        logger.info("vectorstore.index_to_docstore_id: %s", vectorstore.index_to_docstore_id)
-        logger.info("type: %s", type(vectorstore.index_to_docstore_id))
-        if msg["id"] in vectorstore.index_to_docstore_id:
+        logger.info("vectorstore.index_to_docstore_id: %s", vectorstore.index_to_docstore_id.values())
+        logger.info("type: %s", type(vectorstore.index_to_docstore_id.values()))
+        if msg["id"] in list(vectorstore.index_to_docstore_id.values()):
             logger.info("Email already exists in the database.")
+            existing_ids = list(vectorstore.index_to_docstore_id.values())
+            logger.info("Existing email IDs in the database: %s", existing_ids)
             continue
         for header in msg["payload"]["headers"]:
             if header["name"] == "From":

app/requirements.txt

@@ -193,6 +193,7 @@ tzdata==2025.1
 unstructured==0.17.2
 unstructured-client==0.32.0
 unstructured-inference==0.8.10
+unstructured.pytesseract==0.3.15
 uritemplate==4.1.1
 urllib3==2.3.0
 uvicorn==0.34.0