PyMuPDFParser returning empty results on AWS Lambda, but locally it works fine

[ManffTee]

Checked other resources

• I added a very descriptive title to this question.
• I searched the LangChain documentation with the integrated search.
• I used the GitHub search to find a similar question and didn't find it.

Commit to Help

• I commit to help with one of those options 👆

Example Code

import os
import json
import time
import re
import base64
from io import BytesIO


from langchain_community.document_loaders.parsers.pdf import PyMuPDFParser
from langchain_community.document_loaders.blob_loaders.schema import Blob

def parse_multipart_formdata(body, content_type):
    """Parses multipart/form-data from raw event body."""
    
    # Extract boundary from content type
    boundary = content_type.split("boundary=")[-1]
    boundary_bytes = f"--{boundary}".encode()

    # Split the body using the boundary
    parts = body.split(boundary_bytes)
    
    file_content = None
    filename = None

    for part in parts:
        if b"Content-Disposition" in part and b"filename=" in part:
            # Extract file name
            disposition_match = re.search(rb'filename="(.+?)"', part)
            if disposition_match:
                filename = disposition_match.group(1).decode()

            # Extract file content (after double CRLF)
            file_content = part.split(b"\r\n\r\n", 1)[-1].rstrip(b"--\r\n")

            break  # Stop after finding the file
    
    if not file_content:
        raise ValueError("No file found in request body")

    return filename, BytesIO(file_content)






def lambda_handler(event, context):
    print("Upload call initiated")

    headers = {
        "Access-Control-Allow-Origin": "*",
        "Access-Control-Allow-Methods": "GET, POST, DELETE, OPTIONS",
        "Access-Control-Allow-Headers": "Content-Type,X-Amz-Date,Authorization,X-Api-Key",
    }

    start_time = time.time()

    try:
        if "body" not in event:
            raise ValueError("Missing request body")
        
        if "headers" not in event or "content-type" not in event["headers"]:
            raise ValueError("Missing Content-Type header")

        # Extract content type
        content_type = event["headers"]["content-type"]

        # Parse file from the raw multipart body
        filename, file_stream = parse_multipart_formdata(event["body"].encode(), content_type)
        
        print(f"Received file: {filename}, Size: {len(file_stream.getvalue())} bytes")

        
        blob = Blob(data=file_stream.getvalue()) 
        print(f"Type of blob: {type(blob)}")

        # Process with PyMuPDFParser
        parser = PyMuPDFParser()
        pages = parser.parse(blob)
        text = "\n".join([page.page_content for page in pages])

        return {
            'statusCode': 200,
            "body": json.dumps({"message": text}),
            "headers": headers,
        }

    except Exception as e:
        print(f"Error processing request: {e}")
        return {
            'statusCode': 400,
            'body': json.dumps({'error': f'Error processing file: {str(e)}'}),
            "headers": headers,
        }
        
    finally:
        print(f"Total execution time: {time.time() - start_time:.2f} seconds")

Description

Hello everyone,

I’m encountering an unusual issue while trying to extract text from a PDF in AWS Lambda. I have a function that works perfectly when run locally with a Flask API, but when uploaded to AWS Lambda, it returns an empty string ("") instead of the expected text from the PDF.

The function executes without errors, and the Lambda logs don't indicate any problems.

Has anyone encountered this issue before or have suggestions on what might be causing the discrepancy between local execution and Lambda?

System Info

langchain==0.3.14
langchain-community==0.3.14
langchain-core==0.3.31
langchain-groq==0.2.3
langchain-qdrant==0.2.0
langchain-text-splitters==0.3.5