Bug#646400: git-buildpackage: qbp-pq truncates a valid patch

2011-11-21 Thread Guido Günther
On Tue, Oct 25, 2011 at 12:46:16AM +0200, Robert Luberda wrote: > Guido Günther writes: > > > > > Signatures start with '-- \n', I've fixed this. > > Why don't you want to use the --no-signature option of format-patch > (and eventually add the '-- \n' line at the end afterwards)? > > As I wrot

Bug#646400: git-buildpackage: qbp-pq truncates a valid patch

2011-10-24 Thread Robert Luberda
Guido Günther writes: > > Signatures start with '-- \n', I've fixed this. Why don't you want to use the --no-signature option of format-patch (and eventually add the '-- \n' line at the end afterwards)? As I wrote in my second mail, '-- \n' can also be part of a valid patch. Imagine the quite

Bug#646400: git-buildpackage: qbp-pq truncates a valid patch

2011-10-24 Thread Guido Günther
On Sun, Oct 23, 2011 at 11:29:07PM +0200, Robert Luberda wrote: > Package: git-buildpackage > Version: 0.5.32 > Severity: important > > gbp-pq command contains the following code: > if in_patch: > if line.startswith('-- '): > # Found final signature, we're done: >

Bug#646400: git-buildpackage: qbp-pq truncates a valid patch

2011-10-23 Thread Robert Luberda
Just to be sure - changing if line.startswith('-- '): into something like: if line == '-- ': won't fix anything. Patch generated by commands like: echo '- ' > a; echo '+ '> b; diff -u a b will still be truncated. I can see that git format-patch has --no-signature option. Probably

Bug#646400: git-buildpackage: qbp-pq truncates a valid patch

2011-10-23 Thread Robert Luberda
Package: git-buildpackage Version: 0.5.32 Severity: important gbp-pq command contains the following code: if in_patch: if line.startswith('-- '): # Found final signature, we're done: tmp.write(line) break The assumption written in the