| Commit message (Collapse) | Author | Age | Files | Lines |
... | |
|
|
|
| |
Bug: https://bugs.gentoo.org/645554
|
| |
|
|
|
|
| |
Bug: https://bugs.gentoo.org/645554
|
| |
|
|
|
|
|
|
|
|
|
| |
Some libcs do NOT do the favour of turning a NULL-pointer into the
string "(null)", but segfault instead. This is just a friendly guesture
of libcs which do, but it is nowhere standard or that this should result
in "(null)". While "(null)" is actually a developer concept, and
possibly out of context for users, check for printing NULL-pointers and
replace them with "<unset>" instead.
|
|
|
|
| |
Bug https://bugs.gentoo.org/645554
|
|
|
|
|
|
| |
commit fe42f2a99297fed36720e71ba2ed77cf7a9da804 changed the generated
file, this commit ensures that future regenerate operations don't lose
the changes
|
| |
|
| |
|
|
|
|
| |
backup files are suffixed by '~', not prefixed
|
|
|
|
|
| |
Bug: https://bugs.gentoo.org/335453
Bug: https://bugs.gentoo.org/367199
|
|
|
|
|
|
|
| |
When a merge gets aborted, ensure we forget about it if we find about
that same package later in another merge run.
Bug: https://bugs.gentoo.org/470168
|
| |
|
|
|
|
|
|
|
| |
In particular /etc/portage/make.conf can be a directory. If it is,
process it recursively in sorted order.
Bug: https://bugs.gentoo.org/558306
|
|
|
|
|
| |
in particular PORTDIR may get set, so ensure we print the value that's
going to be used
|
| |
|
| |
|
|
|
|
|
|
|
|
|
|
|
| |
There is two parts in here:
a) atom-explode.py: don't crash when Portage couldn't map an atom, just
print a descriptive message instead
b) dotest: filter out Manifest.* files to not feed in non-atoms
thanks tka for testing
Bug: https://bugs.gentoo.org/638970
|
|
|
|
|
|
|
| |
The strptime function is documented to require whitespace between any
two conversion specifications, so test like that.
Bug: https://bugs.gentoo.org/638032
|
| |
|
| |
|
| |
|
|
|
|
| |
Signed-off-by: Fabian Groffen <grobian@gentoo.org>
|
| |
|
|
|
|
|
|
|
|
|
|
| |
This is an somewhat alternative take to the patch by
Niklas Haas <bgo@nand.wakku.to>.
Also exclude /usr/src/debug by default, because it contains a lot of
clutter.
Bug: https://bugs.gentoo.org/646310
|
|
|
|
|
|
|
|
|
| |
This patch is based on the work by Pavel Balaev <pascal@unixdev.ru>.
Add support for color=color mappings. While at it, make color remapping
independent of the order in which they are defined in color.map.
Bug: https://bugs.gentoo.org/651546
|
| |
|
|
|
|
|
|
|
| |
When the argument is an atom-parsable thing, match whatever we find in
the vdb as atom, such that version and range can be taken into account.
Bug: https://bugs.gentoo.org/608960
|
|
|
|
|
|
| |
For bug #608960 we need to be able to handle odd atoms like
!<logrotate-2.3.0 (as used by sys-apps/portage). This adds support for
this construct.
|
| |
|
| |
|
|
|
|
| |
ownership isn't really an issue here, so don't fail on it
|
| |
|
| |
|
|
|
|
| |
This extracts the atom that matched the regular expression.
|
| |
|
|
|
|
|
| |
Just use an int to store getline return, then ensure it is positive,
such that a cast to size_t is guaranteed to result in the same value.
|
| |
|
|
|
|
|
| |
Just use an int to store getline return, then ensure it is positive,
such that a cast to size_t is guaranteed to result in the same value.
|
|
|
|
|
|
|
|
| |
Ex. packages are readeline, ncurses which can be installed
in several SLOTs
Signed-off-by: Joakim Tjernlund <joakim.tjernlund@infinera.com>
Signed-off-by: Fabian Groffen <grobian@gentoo.org>
|
|
|
|
|
| |
Signed-off-by: Joakim Tjernlund <joakim.tjernlund@infinera.com>
Signed-off-by: Fabian Groffen <grobian@gentoo.org>
|
|
|
|
|
|
|
| |
Some pkgs does needs it, dev-lang/python-exec is one.
Signed-off-by: Joakim Tjernlund <joakim.tjernlund@infinera.com>
Signed-off-by: Fabian Groffen <grobian@gentoo.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
* main.c (read_repos_conf): Exclude backup files (aka files with ~ as
postfix) from processing when reading the repos_conf directory.
The old behaviour was to process any file found in repos_conf (i.e.
/etc/portage/repos.conf/*), except those beginning with '.'. This
meant that '.', '..', and '.file' are rejected, but 'file.conf~' is
accepted. Since 'file.conf~' is a backup file, by default it should
rather not be processed.
Bug: https://bugs.gentoo.org/643820
Signed-off-by: Fabian Groffen <grobian@gentoo.org>
|
| |
|
|
|
|
| |
I managed to mix up two bugs, the bug this is all about is #639978
|
| |
|
|
|
|
|
| |
All things on top of PORTAGE_CONFIGROOT need to be relative, so don't
add EPREFIX on top of it.
|
| |
|
|
|
|
|
| |
Fallback to using sys/param.h when __BYTE_ORDER isn't set (which it
isn't on Darwin at least)
|
|
|
|
|
|
|
|
| |
Make sure we don't just eat a char at the end of the atom, because if
that char appears after a -, it will be seen as version part. An
example is xerces-c, where the parsed atom form would just be xerces.
Bug: https://bugs.gentoo.org/638816
|