Currently, if setup_lambda() fails to validate a header then it will throw an error. We need to make sure this error is caught and returned to the initialisation error endpoint. This will mean changing the unit tests as well, now that they expect to PUSH to an endpoint that won't be defined.
This doesn't apply to situations in which the lambda_runtime_api environment variable is undefined. If this isn't available then there's no initialisation endpoint to push to, so we can skip that part.
Currently, if
setup_lambda()fails to validate a header then it will throw an error. We need to make sure this error is caught and returned to the initialisation error endpoint. This will mean changing the unit tests as well, now that they expect toPUSHto an endpoint that won't be defined.This doesn't apply to situations in which the
lambda_runtime_apienvironment variable is undefined. If this isn't available then there's no initialisation endpoint to push to, so we can skip that part.