On Sun, 2009-05-24 at 22:31 +0200, Peter De Wachter wrote: > I've set up a search engine which indexes Debian's source code: > > http://walrus.rave.org/source/ > > Currently, unstable main and contrib are indexed, in total 70 gigabytes > of source. (It would probably be nice to have stable and testing as > well, but I don't have enough disk space available. I skipped non-free > in fear of annoying licenses.) [...]
Cool - that looks really useful. However, it looks like you're just running "dpkg-source -x" to unpack packages. This misses any Debian changes made using a patch system. Unfortunately there is no standard mechanism to apply patches in version 1 source patches, but it should be easy enough to support the standard patch queue formats. Ideally, you would also unpack multi-tarball packages such as gcc before indexing. When I wrote a program to search through Debian source (without indexing) I recursively unpacked all tarballs. Note that there is a tar implementation that includes some weird tarballs as test cases. Ben. -- Ben Hutchings Teamwork is essential - it allows you to blame someone else.
signature.asc
Description: This is a digitally signed message part
