Chinese characters are corrupted after unpacking HTML docset

[zjzdy]

My system is Windows 7 zh_CN.
System default code is GBK.
In HTML.tgz\HTML.docset\Contents\Resources\Documents\mdn.mozillademos.org\zh-CN\docs\Web\HTML\Element\datalist$samples\ have a file is 示例426e.html.But file name is ê?ày426e.html in system.

Then std::ifstream can't find file and QIODevice::read: device not open.

[DerekTurtleRoe]

@zjzdy Does this happen on an English Windows 7?

[zjzdy]

@vgturtle127 I don't know.Because I hadn't installed English Windows.You can try it.

[DerekTurtleRoe]

@zjzdy I am using Windows 10, so I don't know if the problem will exist since localization has changed a lot since Windows 7...

[trollixx]

Don't look in the commit above, it's not relevant, and does not include my attempt to fix the issue.

It looks like that there's no good way to fix this problem until an issue in libarchive is resolved.

[trollixx]

Looks like I've missed archive_entry_pathname_w(), which is hacky, but should fix the issue on Windows.

[DerekTurtleRoe]

@trollixx Oh, thanks. I had no idea!

[trollixx]

@zjzdy where exactly does your HTML.docset come from?