ZipStrings doesn't honor the CodePage property when it reads file entries

[EvilBeaver]

Non-unicode filenames are extracted incorrectly because of behavior of ZipStrings.EncodingFromFlag function.

ZipStrings.EncodingFromFlag always uses SystemDefaultCodepage when it's not a Unicode. Here should be just a CodePage, which can be adjusted from client-side code.

I can fix this and send a Pull Request, but what is release cycle? Whether project abandoned? Maybe I'd better make a fork and use it?

Problem line:

https://github.com/icsharpcode/SharpZipLib/blob/master/src/ICSharpCode.SharpZipLib/Zip/ZipStrings.cs#L128

Steps to reproduce

1. Open non-unicode archive with ciryllic symbols
2. Enumerate entries
3. See that they are incorrect

Expected behavior

Symbols should be correct, when CodePage property is set to ciryllic one

Actual behavior

Symbols are incorrect, changing CodePage property doesn't help

Version of SharpZipLib

1.0

Obtained from (place an x between the brackets for all that apply)

• Compiled from source
• Package installed using:
  • NuGet

[piksel]

This is a duplicate of #274. The project is not abandoned, but I am the sole maintainer and have been very busy lately. A new version will be release pretty soon, so a PR would be welcomed.

[piksel]

I am closing this due to the changes being merged.