Closed benoit74 closed 2 weeks ago
All modified and coverable lines are covered by tests :white_check_mark:
Project coverage is 100.00%. Comparing base (
baec9c1
) to head (f4176f0
).
:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.
It's OK except some naming for which I suggested alternatives.
I've started to build a cheatsheet to stop forgetting about these namings for exc
, fpath
, ... : https://github.com/openzim/overview/wiki/Python-naming-conventions
Do not hesitate to contribute, you're probably way more aware than I am about all these
Is this all necessary? I'm fine with it but I would have been satisfied with a single Exception and custom messages. This would fix the issue but with a smaller API extension.
This was "necessary" because in warc2zim we have different exit codes depending on the problem encountered. I tend to consider that multiple exception is not a high price to pay in term of development or maintenance, and that it is important to keep open all potential use cases open in a shared library.
I forgot to commit my changes, so none of your conversations are fixed in fact. Sorry, I've opened https://github.com/openzim/python-scraperlib/pull/171 with all fixes ...
I've started to build a cheatsheet to stop forgetting about these namings for
exc
,fpath
, ... : https://github.com/openzim/overview/wiki/Python-naming-conventionsDo not hesitate to contribute, you're probably way more aware than I am about all these
I will, but I only see them when it's not aligned 😉
Is this all necessary? I'm fine with it but I would have been satisfied with a single Exception and custom messages. This would fix the issue but with a smaller API extension.
This was "necessary" because in warc2zim we have different exit codes depending on the problem encountered.
Great.
I tend to consider that multiple exception is not a high price to pay in term of development or maintenance, and that it is important to keep open all potential use cases open in a shared library.
It's the little things that piles up and end-up slowing down the whole project. We've already decided long ago, from experience, that we only implement what's needed. There's always the possibility to extend and modify when need grows or change ; including for such libraries. This is just a general feedback on your general last sentence of course 😉
Fix #163