Suggesting some minor changes.

[robkam]

• Usage should mention how to batch download wikis (Wikiteam had launcher.py with listfile).
• Also, include the wiki logo in the download folder. Including the logo in the IA upload puts the logo on the archive instead of the generic IA one.
• Instead of the user having to manually grep, when Wikiteam3dumpgenerator completes, it should check the dumped XML file and also check for a closing </mediawiki> tag.
• Fix the corrupt folder name, e.g. https://subdomain.miraheze.org/w/api.php gets a _w- in the folder name.

[yzqzss]

batch download

Adding a launcher is unnecessary for something that can be done via simple shell script.

wiki logo

OK. #18

xml check

Good suggestion, but parsing xmldump is not easy, especially since it comes in various formats. If anyone is willing to PR the big feature I'll take it.

corrupt folder name

I've repeated many times that this is not a bug. if you don't keep it consistent with the HTTP PATH, the identifier will not be consistent.

What do you think would happen if example.com, example.com/a/ and example.com/w/ were three different wikis?

[yzqzss]

please use sax to stream parse xml.

https://github.com/saveweb/wikiteam3/blob/v4-main/tools/xml2titles.py

[robkam]

To be honest, your code quality is not good

That's no surprise

Another suggestion, to use SevenZipFile from py7zr instead of relying on 7z in PATH?

[yzqzss]

https://github.com/miurahr/py7zr#performance

[robkam]

please use sax to stream parse xml.

https://github.com/robkam/checkxml/blob/main/check_xml_inegrity.py

[robkam]

wiki logo

OK. #18

Unnecesary - wikiteam3uploader fetches and uploads the logo.

[robkam]

Shouldn't --bypass-cdn-image-compression be enabled by default since the user probably wants to preserve the original image?

[yzqzss]

?

[robkam]

Shouldn't have to specify --bypass-cdn-image-compression should just fetch the uncompressed image(s) by default.

[robkam]

Never mind I've misunderstood.