Thu, 05 Jun 2008 21:44:38 -0400
add web site to hg repo
-*- text -*- To do: * When there's partial success in extracting an archive, go ahead and use the partial version. To do this, when extracting, save the stderr of everything that gets run, along with its status code. Let the top-level loop decide what to do with that information. It should do this by looking in the extraction directory/file and seeing if there's anything actually there. If there is, it should go ahead and keep it, and report errors as warnings (including stderr). Otherwise, it should treat it as an unsuccessful extraction. * The text for when directory names don't match confused Paul. Consider alternatives. He thinks it would be cool if the actual/expected directory names aligned in the output; might need to do your own wrapping for this. * When we list how many archives are in the file, show a count, along with the number of regular files in the archive too. "foo.tar contains 5 other archives, out of 81 files total." * When we extract a compressed file, check to see if it itself is an archive. Follow all the usual rules for recursive extraction when we do this. * It seems like when you extract an archive that only has one file, with -f, you still get prompted about what to do with it. -f should always assume that everything gets extracted in the current directory. Track this bug down. Things which I have a use case/anti-use case for: * Support pisi packages (http://paketler.pardus.org.tr/pardus-2007/) * Steal ideas from <http://martin.ankerl.com/files/e>. * More consistently raise and handle exceptions. Things that are generally good: * Better tests. * Better error messages. Things I think might be good but can't prove: * Consider having options about whether or not to make sane directories, have tarbomb protection, etc. * Use zipfile instead of the zip commands. * Processing from stdin. * shar support.