Unicode path not supported?

Question

Unicode path not supported?

qiuzhanghua opened this issue 2 years ago · 3 comments

I use this crate to unzip a file with archive path including unicode name, It seems not correct.

cargo run

should be

File 0 extracted to "中文/"
File 1 extracted to "中文/.DS_Store" (6148 bytes)
File 2 extracted to "__MACOSX/中文/._.DS_Store" (120 bytes)
File 3 extracted to "中文/folder/"
File 4 extracted to "中文/目录/"

instead of

File 0 extracted to "Σ╕¡µûç/"
File 1 extracted to "Σ╕¡µûç/.DS_Store" (6148 bytes)
File 2 extracted to "__MACOSX/Σ╕¡µûç/._.DS_Store" (120 bytes)
File 3 extracted to "Σ╕¡µûç/folder/"
File 4 extracted to "Σ╕¡µûç/τ¢«σ╜ò/"

Answer 1 · 2023-05-08T20:46:06.000Z

Thankyou for the example archive! I'll have a look at how it's representing the filenames (encoding in zip files is a little troublesome, I imagine this is using the unicode Extra Data field, which we don't support)

Answer 2 · 2023-08-11T03:35:40.000Z

See also #188

Answer 3 · 2024-06-02T21:34:55.000Z

This archive is not using the Unicode Extra Data field. It's encoding the file names in UTF-8, but then not setting the utf8 bit in flags, which normally means the filename is CP437. This isn't a bug in zip, it's a bug in the program that created the archive.