Creating a tarball out of a directory in a bare git repository

Hello,

I am developing a web interface, part of which allows the user to download a tarball of a directory (which is in a bare repostitory on the filesystem) on clicking on a link.

This is how I am planning on doing it. Just wanted to assess whether this is efficient enough, or if anyone knows of a better way. I haven't tried this out yet though, so I don't know if it will work at all.

- list all the files/folders in the directory (i have the sha1 hash of the
directory)
- foreach file in the directory, get the file contents (recurse if there are
further subdirectories). Probably store the result in an array. 
- create a file somewhere on the filesystem out of each file output above.
Store it under the directory name.
- create a tarball out of the 'working' directory as usual.

I was wondering if I can get away without having to create a working version out of the bare version of the directory (create each file in the directory and store it somewhere on the filesystem), and then compress it as usual.

I hope someone has a much simpler answer to this problem!

Thanks, Jawad.

On Mon, 1 Mar 2010 14:47:45 -0800 (PST), jateeq <jawad_atiq@hotmail.com> wrote:

download

filesystem)

[..snip..]

Have a look at "git archive":

http://www.kernel.org/pub/software/scm/git/docs/git-archive.html

jateeq> - list all the files/folders in the directory (i have the sha1 hash of the jateeq> directory) jateeq> - foreach file in the directory, get the file contents (recurse if there are jateeq> further subdirectories). Probably store the result in an array. jateeq> - create a file somewhere on the filesystem out of each file output above. jateeq> Store it under the directory name. jateeq> - create a tarball out of the 'working' directory as usual.

Sounds like you're reinventing "git archive". :)