io.archive.Zip.createArchive failed when (MBCS)Chinese file name

are you able to debug this so we can get a source line?

I debug the code above by ddbg, say:

Unhandled exception in Ddbg 0.11.3 beta:
test: 鎷掔粷璁块棶銆?

Please report this problem!
See the http://ddbg.mainia.de/releases.html for details.
Thank you!

the path(file name) in archive in local encode format maybe fix this issue.

I use this code to check the path is ASCII character only:

    // bugs with tango.io.archive.Zip, MBCS(e.g. Chinese) in path is not supported
    bool isAscii(char[] str)
    {
        foreach(c; str)
            if(c>127) 
                return false;
        return true;
    }    
    
    foreach (file; scan.files) 
        if(isAscii(file.toString))
            files ~= file.toString;

    createArchive(r"test.zip", Method.Deflate, files);

Hmm, it seems like the Zip spec implemented use Codepage 437 internally, and the problem then is that the chinese characters out of range - I doubt that really is unhandled though, but maybe it needs to go via the local codepage to work? It would be nice with some information/links on how this is meant to work - if anyone knows where to find it.

debug the code above: ddbg createzip.exe

output is:

Unhandled exception in Ddbg 0.11.3 beta:
test: Access is denied.

Please report this problem!
See the http://ddbg.mainia.de/releases.html for details.
Thank you!

How is ddbg's failure relevant to this ticket? It is not the information needed to resolve it. The error message you get is understandable, what is not is how to make non-ascii, non-western characters encodable to codepage 437, or some other format that zip files will accept. It must be possible to dig out information on this.

Seems like this issue is related to the UTF-8 TODO at the top of this file. Relevant information to resolve that would be this:

http://www.pkware.com/documents/casestudies/APPNOTE.TXT

Also, possibly relevant if wider codepage support is needed -

http://www.chilkatsoft.com/p/p_453.asp