Using Python Functionality in Flow Scripts

You can use almost all standard Python functionality in flow scripts by importing the standard Python module of your choice. There are several third-party modules available. See available third-party Python modules for more information. On this page we will give you some examples of common usage patterns.

Available Third-Party Python Modules

The following third-party Python modules are available to be used in your flow scripts:

  • async-timeout
  • html5lib
  • lxml
  • passlib
  • pdftotext
  • pycryptodome
  • pyjwt
  • pyquery
  • pystache
  • python-dateutil
  • pyyaml
  • ujson

Contact us at to request additional Python modules.


Parse a PDF File

You can access the content of PDF files directly in your flow script.

Example: Access a date written in a PDF file

import re
import base64
import io
import pdftotext

def handler(system, this):
    pdf_name = 'example.pdf'
    # Open the Cloudomation file object
    pdf_file = system.file(pdf_name)
    # Per default Cloudomation tries to parse files using UTF-8 and returns a Python string.
    # Here we want to access the binary content of the file,
    # so we need to set `convert_binary` to `False`
    file_base64 = pdf_file.get('content', convert_binary=False)
    file_bytes = base64.b64decode(file_base64)
    # Construct a `PDF` object from the bytes
    pdf = pdftotext.PDF(io.BytesIO(file_bytes))
    # Access the text of the first page of the PDF file
    pdf_text = pdf[0]
    # Clear the PDF object, so that the file object is closed
    pdf = None
    # Use a Regex to find the date
    the_date ='Date:\s*(\d{4}-[01]\d-[0-3]\d)', pdf_text).group(1)
    return this.success('all done')

Work with CSV Files

You can use the csv module to read CSV files into Python objects or write Python objects into CSV files.

Example: Create a CSV file and send it by email.

import base64
import csv
import io

def handler(system, this):
    fake_data_list = [
        {'name': 'spam', 'number': 1},
        {'name': 'eggs', 'number': 2},
        {'name': 'foo', 'number': 3},
        {'name': 'bar', 'number': 4},
    result_file = io.StringIO()
    csv_writer = csv.DictWriter(
        fieldnames=['name', 'number'],
    for data in fake_data_list:
    csv_content = result_file.getvalue()
    result_file = None
    csv_writer = None


            'from': '',
            'to': '',
            'subject': 'Result',
            'text': 'See the attached file',
            'login': '',
            'password': '****',
            'smtp_host': '',
            'smtp_port': 587,
            'use_tls': True,
            'attachments': [

    return this.success('all done')