| Commit message (Collapse) | Author | Age |
|---|
| |
|
|
| |
Closes: #694404
|
| |
|
|
|
|
|
|
| |
Thanks to Eric Pozharski for the initial patch.
This version is slightly modified, as it provides a test case, and is
compliant with Policy 5.2, which states that comments must not have leading
whitespace before the hash.
|
| |
|
|
|
|
| |
We don't want magic environment variables to override what
we specified on the command line when asking dpkg-architecture
to compare architectures for us....
|
| | |
|
| |
|
|
| |
pbuilder-satisfydepends-funcs and add tests
|
| | |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
tags 620730 + patch pending
tags 626431 + patch pending
tags 627086 + patch pending
tags 603420 + patch pending
tags 601250 + patch pending
tags 603881 = unreproducible
tags 400848 = unreproducible
tags 622624 = unreproducible
thanks
Dear Junichi, Matt and others,
I’ve prepared an NMU for pbuilder (versioned as 0.199+nmu4) and uploaded
it to Debian unstable, with a two-day delay as per devref §5.11.1 since
this fixes an RC bug older than seven days. The debdiff is attached; it
closes five bugs. I’m also tagging three other bugs as unreproducible as
pbuilder itself (obviously) doesn’t FTBFS on my system and the test pak-
kage I used didn’t exhibit problems with either comment blocks or hori-
zontal tabulator characters in the control file. I’m also attaching the
test package (absolutely minimal, but that’s what you get).
I mostly used patches from the bugs closed as baseline for my changes,
but did a bit more usually. I’m a heavy user of cowbuilder, because on
m68k there’s currently no buildd and cowbuilder is all I know (and love,
really… someone should make a buildd flavour that uses cowbuilder in-
stead of sbuild, if it goes by my will).
Cheers.
|
| |
|
|
|
|
|
|
|
|
| |
Hi,
I talked to Junichi and he told me it was fine if I NMU'ed pbuilder to fix these
bugs. This is the debdiff for my NMU.
Regards,
Emilio
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| |
|
|
|
| |
Replace cat + awk + sed in get_control_re() with a sed script; rename
the function to get_source_control_field.
|
| |
|
|
|
|
|
|
|
| |
Set BINARY_ARCH in pbuilder-satisfydepends-checkparams just as in
pbuilder-checkparams instead of setting BC_REGEXP and BD_REGEXP and
change get_build_deps and get_build_conflicts to query for Build-Deps /
Build-Conflicts only, with an optional separate get_control_re call for
the -Indep deps if needed. Update testsuite to also test for
build-deps-indep.
|
| | |
|
| |
|
|
|
|
| |
package_versions(): drop useless subshell; pipes are already run in a
subshell. Also add LC_ALL=C for consistency when parsing output of
commands.
|
| |
|
|
| |
package_versions(), candidate_version(): cleanup regexp
|
| |
|
|
|
| |
Add get_build_deps() and get_build_conflicts() functions wrapping
get_control_re() for BD_REGEXP and BC_REGEXP over DEBIAN_CONTROL.
|
| | |
|
| |
|
|
|
|
|
| |
(closes: 519306)
It used to rely on ',' being there, which really package maintainers
are free to insert '|' etc.
|
| | |
|
| |
|
|
| |
Patch from Adeodato Simo.
|
| | |
|
| |
|
|
|
|
|
|
|
| |
value of source headers from a control file.
* Add function filter_arch_deps() to pbuilder-satisfydepends-aptitude to
filter out arch dependencies not for a certain arch.
* Source pbuilder-satisfydepends-funcs in pbuilder-satisfydepends-aptitude.
* Add support for arch specific build-deps and build-conflicts to
pbuilder-satisfydepends-aptitude.
|
| |
|
|
|
| |
enclosure early on and to permit spaces or slashes since some
pbuilder-satisfydepends* implementations use tr to map spaces to slashes.
|
| | |
|
| | |
|
| | |
|
|
|
pbuilder-satisfydepends-funcs.
|