genkgo / camt

PHP Parser for CAMT messages
MIT License
81 stars 67 forks source link

Remove scrunitzr #139

Closed frederikbosch closed 1 year ago

frederikbosch commented 1 year ago

Shall we remove it? I see that test is failing.

PowerKiKi commented 1 year ago

Scrutinizer static analysis is indeed lagging behind PhpStan. So there are a few false positives. However I find that it still have some value, especially for code coverage statistics, and preventing that it goes down.

IMHO I'd keep it while tolerating that i might be failing for the wrong reasons from time to time. Since integration and usage is very simple, I see it as a nice to have with minimal cost. I would probably not keep it if it becomes too much of a burden though...

localheinz commented 1 year ago

@PowerKiKi

I use https://about.codecov.io for collecting and reporting code coverage - it is focussed and works like a charm. I vote for removing Scrutinizr.

PowerKiKi commented 1 year ago

Does it requires a login to view data ? I can't stand https://coveralls.io, because it hides everything behind a login :sob:

Does it provide a code quality badge that we can proudly show off in the README ? :sunglasses:

image

localheinz commented 1 year ago

@PowerKiKi

It does not show a code quality badge, no.

See #140!

PowerKiKi commented 1 year ago

@frederikbosch any opinion on the loss of code quality measurements ?

frederikbosch commented 1 year ago

No, I removed all style and code quality measurements from all my libraries. I found it a waste of time. I keep it with PHP-CS-Fixer and PHPStan.

localheinz commented 1 year ago

@PowerKiKi

Perhaps https://psalm.dev/docs/running_psalm/command_line_usage/#shepherd could be of interest? You could display type coverage when using vimeo/psalm.

Here is an example from ergebnis/phpunit-slow-test-detector:

CleanShot 2023-02-14 at 10 41 28@2x

PowerKiKi commented 1 year ago

I wouldn't replace 1 tool by 2 different tools. I guess only codecov will do then

PowerKiKi commented 1 year ago

Done via #140