DO NOT USE, THIS IS PRE-ALPHA, HAS NO TESTS, NO MONITORING, BAD LOGGING

tclambda

AWS Lambda calling library

Works together with functions defined in tc-sam-cli

Configuration of AWS Lambda

# app.py
import tclambda

from numpy.polynomial.polynomial import polyfit as np_polyfit

handler = tclambda.LambdaHandler()


@handler.register()
def polyfit(*args, **kwargs):
    return list(np_polyfit(*args, **kwargs))

The registered handler must return something that is json serializable.

Usage in other projects

Configuration is primarely done by environmental variables.

TC_NUMPY_QUEUE="https://sqs.eu-west-1.amazonaws.com/12345/NumpySqs"
TC_NUMPY_BUCKET="s3-result-bucket"

from tclambda.auto_functions import numpy

lambda_result = numpy.polyfit([1, 2], [2, 1], 1)
print(lambda_result.result())
# Output: [2.999999999999998, -0.9999999999999992]

Retry lambda

import tclambda
import requests

handler = tclambda.LambdaHandler()


@handler.register()
def web_data(*args, **kwargs):
    try:
        response = requests.get("https://example.com")
        response.raise_for_status()
    except Exception as e:
        raise tclambda.RetryException(e)
    return response.text

Error handling

Sentry

Install tclambda together with sentry pip install tclambda[sentry] Setting up sentry is done automatically through environmental variables.

• SENTRY_DSN
• SENTRY_RELEASE
• SENTRY_ENVIRONMENT

Tests

• Create an S3 bucket to test live against
• Add TC_THIS_BUCKET=my-test-s3-bucket to .env
• Run pipenv run test