Add padpo to our CI

[JulienPalard]

Currently padpo gives way too much errors to add it to our CI, but by "padpo-ifying" files one by one, and enforcing a non-regression in our CI, we should be able to "conquier" the whole repo.

[JulienPalard]

I think it can be done like this:

JulienPalard@be51eff

But it requires https://github.com/vpoulailleau/padpo/issues/11

This way we can add files as they are clean in the PADPO_COVERED variable, until all files are in it, in which case we can drop this temporary variable.

[vpoulailleau]

Another way, with the existing padpo would be storing the padpo result somewhere for every file. Each time a file is modified, it is checked by padpo and compared with the previous result. Adding errors in a file being prohibited.

This would allow more incremental correction, not only file based padpoification 😉

[vpoulailleau]

@Seluj78 @JulienPalard padpo v0.7.0 is out!

It adds:

• --version flag (vpoulailleau/padpo#10)
• -i *.po (vpoulailleau/padpo#11)

[Seluj78]

🎉 ! I'll update poutils

[Seluj78]

poutils 0.2.2 is out ! 🚀

[JulienPalard]

Another way, with the existing padpo would be storing the padpo result somewhere for every file. Each time a file is modified, it is checked by padpo and compared with the previous result. Adding errors in a file being prohibited.

This would allow more incremental correction, not only file based padpoification wink

Oh yes a bit like in cpython Doc/tools/susp-ignored.csv, ignoring false positive from make suspicious.
I'd prefer it one shot in a single file though.

[vpoulailleau]

Maybe another possibility (but that's more work) is to add an --incremental flag to padpo to ignore messages that were already present in a provided commit (in case filenames didn't change).