Signed-off-by: NLuca Falavigna <dktrkranz@debian.org>

Signed-off-by: NLuca Falavigna <dktrkranz@debian.org>

Signed-off-by: NLuca Falavigna <dktrkranz@debian.org>

as we also parse dsc files here, only do the rejection on missing keywords
when we parse a changes file.
adjust all calls for dsc files to tell us they are dsc, by default we do
the check, caller has to turn it off...
Signed-off-by: NJoerg Jaspert <joerg@debian.org>

get_poolfile_like_name is pickier when parsing file names, so
adhere to the new schema to avoid wrong matches.
Signed-off-by: NLuca Falavigna <dktrkranz@debian.org>

Signed-off-by: NJoerg Jaspert <joerg@debian.org>

fix up a few wrong docstrings
Signed-off-by: NJoerg Jaspert <joerg@debian.org>

only bail out if we have a non-source upload.
do not bail out if self.pkg.orig_files is empty.
We either have a bug filling orig_files where we only set this when the files
already have been in the archive, or we intended to use orig_files to show
us files already in the archive. Currently it looks like the latter, and then
it is wrong to check on that in check_source, as that would mean we skip one
check (and also extraction of .debinfo information for the BTS!) for every
upload that comes with totally new files. (Imagine a native package).
Signed-off-by: NJoerg Jaspert <joerg@debian.org>

partly back out the previous commit,
keep the conditional on source, we only extract the info on source
uploads.
Signed-off-by: NJoerg Jaspert <joerg@debian.org>

drop the condition on bts-changelog and source upload.
bts -> old old comment says it should have been done years ago
source -> don says bts has no problem with it getting the files
always.

maybe this "fixes" the bug where we not write .versions for all
uploads, just some. seemingly random.
maybe it just gets is to backtrace lots.
lets see, a surprise for the new year.
Signed-off-by: NJoerg Jaspert <joerg@debian.org>

Signed-off-by: NJoerg Jaspert <joerg@debian.org>

no longer try to rampage through a morgue/ directory, it has no
business there. Instead just claim the next name in the reject/ subdir
of queue/
Signed-off-by: NJoerg Jaspert <joerg@debian.org>

Signed-off-by: NMark Hymers <mhy@debian.org>

deal with orig's we have in the DB which are references in the changes file but not there (e.g. security uploads)
Signed-off-by: NMark Hymers <mhy@debian.org>

Signed-off-by: NMark Hymers <mhy@debian.org>

Signed-off-by: NMark Hymers <mhy@debian.org>

Signed-off-by: NMark Hymers <mhy@debian.org>

do it in determine_new() (queue.py) instead of do_new() (process_new.py)
Signed-off-by: NTorsten Werner <twerner@debian.org>

Signed-off-by: NJoerg Jaspert <joerg@debian.org>

Signed-off-by: NJoerg Jaspert <joerg@debian.org>

Signed-off-by: NJoerg Jaspert <joerg@debian.org>

dont fillin automated reject message when doing a manual NEW reject.
Signed-off-by: NJoerg Jaspert <joerg@debian.org>

Signed-off-by: NJoerg Jaspert <joerg@debian.org>

Signed-off-by: NMark Hymers <mhy@debian.org>

Signed-off-by: NChris Lamb <lamby@debian.org>

Signed-off-by: NChris Lamb <lamby@debian.org>

Signed-off-by: NChris Lamb <lamby@debian.org>

Signed-off-by: NChris Lamb <lamby@debian.org>

Signed-off-by: NChris Lamb <lamby@debian.org>

Signed-off-by: NChris Lamb <lamby@debian.org>

Signed-off-by: NChris Lamb <lamby@debian.org>

Signed-off-by: NMark Hymers <mhy@debian.org>

Signed-off-by: NMark Hymers <mhy@debian.org>

Signed-off-by: NMark Hymers <mhy@debian.org>

Signed-off-by: NMark Hymers <mhy@debian.org>

We don't actually have any multi-archive machines at the moment
and removing it makes test-setups easier to do.
Signed-off-by: NFrank Lichtenheld <djpig@debian.org>

Signed-off-by: NFrank Lichtenheld <djpig@debian.org>

Signed-off-by: NMark Hymers <mhy@debian.org>