Cast XML bytes to str to avoid encoding issues

[vgambier]

Intended to fix #29
Haven't tested it yet, wanted to double-check with you first because of the comments discouraging messing with the encoding.
Another alternative would be to add something like xml = xml.decode("utf-8)

[elsiehupp]

The unit tests are still broken in the python3 branch, so I can't test it that way, but I'll try running the command from this comment in the original issue. (I'll do this on elementary OS, since it's the closest thing I have to Linux Mint.)

[robkam]

The unit tests are still broken in the python3 branch,

Is this the fails mentioned at #7 and are these fails what's blocking progress on Wikiteam3?

[elsiehupp]

Is this the fails mentioned at #7 and are these fails what's blocking progress on Wikiteam3?

Among other things, yes. I'm also just super not on top of things. 😬 It doesn't help that I don't work in Python very much, so I'm extremely rusty.

I had been doing some work in the prepare-for-publication branch, but it has become increasingly out of date with the ad-hoc bug-fixes in the python3 branch. (The big thing I was doing was introducing mypy type-checking, which involved substantial rewrites. I mean you can fiddle around with that branch if you'd like...)

Anyway I just pushed a commit in this Pull Request that seems to fix the encoding issues, and I have repeatedly tested it using the following command:

dumpgenerator http://wiki.othing.xyz --xml

It works on my computer with elementary OS; could either of you—@vgambier or @robkam—try testing it on Windows? If it works on Windows at this point, then we can safely merge it!

As before, to check out this commit, you can do:

git pull && git checkout fix-bytes-regex

[elsiehupp]

Hi @yzqzss, @Dss0 & @RedSparr0w—could you give this Pull Request a try? Thanks!

(Also apologies that I have been extremely slow to respond here...)

[elsiehupp]

By the way, I have no idea why the Pre-Commit CI is failing. It looks like it's a configuration issue. I will have to dig into it further.

[vgambier]

I'm sorry, I don't have access to a Windows machine. Thank you for the response :)

[RedSparr0w]

This PR seems to be working well for me on the following:
Python 3.9.6 - Debian 11

dumpgenerator --xml

dumpgenerator --xml --curonly

[RedSparr0w]

Not sure if it's just this PR or not, but if I'm not using --xmlrevisions it seems to take a lot longer to fetch all the data.
Using the --xmlrevisions arg, takes about 10→15 minutes.
But without it, it's been about an hour and maybe 20% done.

[robkam]

dumpgenerator http://wiki.othing.xyz --xml works fine on Windows.
Thank you for supporting Windows!

[robkam]

I failed to get pip install --force-reinstall dist/*.whl to work with Python 3.11.1 from Python.org on Windows 10, in Git Bash in Windows Terminal. It worked when I used Python 3.10 from the Microsoft Store and using Command Prompt in Windows Terminal.