lambda_function.py

#!/usr/bin/env python
"""
pdf2txt on AWS Lambda
"""


from io import StringIO, BytesIO
from os import environ
import boto3

from pdfminer.pdfinterp import PDFResourceManager, PDFPageInterpreter
from pdfminer.converter import TextConverter
from pdfminer.layout import LAParams
from pdfminer.pdfpage import PDFPage

s3 = boto3.resource('s3')

PDF_BUCKET = environ['PDF_BUCKET']
TXT_BUCKET = environ['TXT_BUCKET']

def pdf2txt(pdf_obj):
    """
    return pdf contents
    """

    rsrcmgr = PDFResourceManager()
    codec = "utf-8"
    text = ""

    with StringIO() as output:
        device = TextConverter(rsrcmgr, output, codec=codec, laparams=LAParams())
        with BytesIO(pdf_obj.read()) as inp:
            interpreter = PDFPageInterpreter(rsrcmgr, device)
            for page in PDFPage.get_pages(inp):
                interpreter.process_page(page)
                text += output.getvalue()

        device.close()
        return text

def lambda_handler(event, context):
    """
    aws lambda handler endpoint
    """
    print(event)
    print(context)

    s3_object_name = event["s3_object_name"]
    pdf = s3.Object(PDF_BUCKET, s3_object_name).get()['Body']
    txt = pdf2txt(pdf)

    output_csv_filename = f'{s3_object_name}.txt'
    s3.Object(TXT_BUCKET, output_csv_filename).put(Body = txt)

    return ({"filename": f'{output_csv_filename}', "content": txt})