-
functions/modules are tight coupled
very high risk to brake code by implicit change
-
props drilling
env:namespaces < --- env config env:code -> lambda_handler(event, context) < --- entry point / front controller -> build_result(context, key, namespaces, code) < --- modules -> request_status(context, namespace, code) -> _sanitize_code(code)
-
tests are not adequate
it's hard to mock/stub dependencies of
count_items()
questions:
- are these tests isolated?
NO
- is source code loosely coupled?
NO
- are these tests maintainable?
NO
- do they really perform unittest?
NO
notes:
- requires to mock whole hierarchy
- there are some hidden non-easy testable resources (boto3 seesion/client)
- are these tests isolated?
p3.3.advanced-testing
Folders and files
Name | Name | Last commit date | ||
---|---|---|---|---|
parent directory.. | ||||