SLaT

Simple Lambda Toolkit

Collection of reusable Python tools for lambda development.

This project is meant to :meat_on_bone: (or :corn:) these requirements

• As a lambda developer I want to write as little boilerplate as possible
• As a lambda developer I want to ensure I produce high value, discoverable logs
• As a lambda developer I want to ensure I sends proper, consistent, well formed responses
• As a lambda developer I want ToBeDetermined

Example hello world Lambda

from slat.log_util import LogUtil
from slat.response import LambdaProxyResponse
from slat.types import JsonapiBody

# initiate you logger outside the handler
log = LogUtil.init_logger(env_var_name='LOG_LEVEL', default_level='INFO')

def lambda_handler(event, context):
    # bind the request id to the logger
    LogUtil.init_request({'request_id': context.aws_request_id})
    # make log statements as normal
    log.debug('Lambda Handler Event', lambda_handler_event=event)
    # prepare a Jsonapi payload (https://jsonapi.org/)
    response_body = JsonapiBody(data={'hello': 'world!'}, errors=[], meta={})
    # respond in the expected APIGateway lambda proxy format (https://docs.aws.amazon.com/apigateway/latest/developerguide/set-up-lambda-proxy-integrations.html#api-gateway-simple-proxy-for-lambda-output-format)
    return LambdaProxyResponse.respond(200, response_body, context)

Slat Tools

Responses

TBDocumented

Logging

Structlog is used for structured JSON logging

Example Usage

import logging
from slat.log_util import LogUtil

log = LogUtil.init_logger(default_level='INFO',  correlation_id_key_val={'request_id': '999'})
log.info('is this JSON: {"answer": 42}')
log.error("the log message", some="value", extra_data=[1, 2, 3, "4"])
# only OUR logger will render as JSON
logging.getLogger("test").warning("hello")

output:

{"event": "is this JSON: {\"answer\": 42}", "level": "info", "logger": "slat.log_util", "request_id": "999", "timestamp": "2019-11-06T21:04:33.517295Z"}
{"event": "the log message", "extra_data": [1, 2, 3, "4"], "level": "error", "logger": "slat.log_util", "request_id": "999", "some": "value", "timestamp": "2019-11-06T21:04:33.517652Z"}
hello

For test runs you can add an ENV flag TESTING_RUN=true and log statements will be written to ./testing.log

TESTING_RUN=true pytest

Running tests

pytest

# with coverage 
pytest --cov=slat

# generage html report
coverage html
open ./htmlcov/index.html

Developing

create file ~/.pypirc

[distutils]
index-servers =
  pypi
  pypitest

[pypi]
repository: https://upload.pypi.org/legacy/
username:
password:

[pypitest]
repository: https://test.pypi.org/legacy/
username:
password:

build

# tick to semver `version` in setup.py
rm -rf dist
python setup.py bdist_wheel --universal

pypitest

twine upload --repository pypitest dist/*
pip install slat --index-url https://test.pypi.org/simple/ --upgrade

pypi

twine upload --repository pypi dist/*
pip install slat --upgrade