Age | Commit message (Collapse) | Author |
|
Also fixes the bug in the previous commit, in which the
arch-independent packages would be deleted from the staging
directory after only one iteration.
|
|
Signed-off-by: Aaron Griffin <aaronmgriffin@gmail.com>
|
|
This could be useful if we move community to a separate
SVN repo
Signed-off-by: Aaron Griffin <aaronmgriffin@gmail.com>
|
|
ARCHES is defined in config. db-update should now loop
over all arches on each run and update whatever is in the
staging dir.
Remove the db-*64 scripts due to this
Signed-off-by: Aaron Griffin <aaronmgriffin@gmail.com>
|
|
Conflicts:
db-update
Signed-off-by: Aaron Griffin <aaronmgriffin@gmail.com>
|
|
Signed-off-by: Aaron Griffin <aaronmgriffin@gmail.com>
|
|
This prevents some errors where packages where removed
from the staging dir when they shouldn't have been.
Original-idea-by: Pierre Schmitz <pierre@archlinux.org>
Signed-off-by: Aaron Griffin <aaronmgriffin@gmail.com>
|
|
This assumes the -q flag patch is accepted for pacman 8)
Signed-off-by: Aaron Griffin <aaronmgriffin@gmail.com>
|
|
repo-add works fine if no db file exists
Signed-off-by: Aaron Griffin <aaronmgriffin@gmail.com>
|
|
We were copying our DB file from the repo to
our temp build dir all the time. We should only
do this if there are packages to add
Signed-off-by: Aaron Griffin <aaronmgriffin@gmail.com>
|
|
If add/del dirs exist, we try to fix it all up and it can
error quite loudly
Signed-off-by: Aaron Griffin <aaronmgriffin@gmail.com>
|
|
The chmod doesn't work unless the user is the owner of the file
Resorting to umask changes here
Signed-off-by: Aaron Griffin <aaronmgriffin@gmail.com>
|
|
The local variable 'arch' was being overwritten when we
source PKGBUILDs. Change it to _arch wherever we deal
with PKGBUILD files
Signed-off-by: Aaron Griffin <aaronmgriffin@gmail.com>
|
|
db-remove is superior and requiring package files is dumb.
Kill off this dir and clean up some code, yay!
Signed-off-by: Aaron Griffin <aaronmgriffin@gmail.com>
|
|
This prevents us from trying to remove the lockfile
twice, and calling cleanup multiple times
Signed-off-by: Aaron Griffin <aaronmgriffin@gmail.com>
|
|
Additionally, make TMPDIR configurable so we can move that
Signed-off-by: Aaron Griffin <aaronmgriffin@gmail.com>
|
|
This should allow us to move FTP and SVN placement
around
Signed-off-by: Aaron Griffin <aaronmgriffin@gmail.com>
|
|
Signed-off-by: Aaron Griffin <aaronmgriffin@gmail.com>
|
|
This will force all files copied back and forth to have
0664 permissions so that we can attempt to do away with
this permission adjusting cron job
Signed-off-by: Aaron Griffin <aaronmgriffin@gmail.com>
|
|
Signed-off-by: Aaron Griffin <aaronmgriffin@gmail.com>
|
|
This is to allow moves to /home/ftp/ to remain on the same
filesystem, thus making the final moves atomic
Signed-off-by: Aaron Griffin <aaronmgriffin@gmail.com>
|
|
Apparently, people seem to be renaming files as a "hack" and
breaking things. Let's make sure they edited the PKGINFO too
Signed-off-by: Aaron Griffin <aaronmgriffin@gmail.com>
|
|
Common functions, such as sourcing makepkg, and locking a repo
Additionally, repo-locking has been unified to prevent multiple
users from corrupting a repo (i.e. db-move run while db-remove
in progress)
Signed-off-by: Aaron Griffin <aaronmgriffin@gmail.com>
|
|
This stops files being removed from the staging area if they were not
successfully copied to the repo. This can happen due to permission
issues with the database when multiple people update the repo in quick
succession.
Original-work-by: Allan McRae <allan@archlinux.org>
Signed-off-by: Aaron Griffin <aaronmgriffin@gmail.com>
|
|
This way, all package files get cleaned up properly
at the end of the process.
Signed-off-by: Aaron Griffin <aaronmgriffin@gmail.com>
|
|
This would break all adds if there were any deletes.
We definitely don't want that.
Also, cleanup of cron scripts in the same commit because
I'm lazy
Signed-off-by: Aaron Griffin <aaronmgriffin@gmail.com>
|
|
Signed-off-by: Aaron Griffin <aaronmgriffin@gmail.com>
|
|
Left in a comment while debugging
Signed-off-by: Aaron Griffin <aaronmgriffin@gmail.com>
|
|
We don't use fakeroot anymore in these scripts
Signed-off-by: Aaron Griffin <aaronmgriffin@gmail.com>
|
|
Leftovers - accidental commit
Signed-off-by: Aaron Griffin <aaronmgriffin@gmail.com>
|
|
Remove all package file we added or removed
Signed-off-by: Aaron Griffin <aaronmgriffin@gmail.com>
|
|
This allows us to intuitively handle on-the-fly repos
by calling `db-update foobar i686` for the foobar-i686
tag, and let us fail on the server side if the repo
hasn't been approved/created by someone with admin rights.
Signed-off-by: Aaron Griffin <aaronmgriffin@gmail.com>
|
|
We accidentally were calling repo-add/repo-remove
with invalid (empty) params.
Signed-off-by: Aaron Griffin <aaronmgriffin@gmail.com>
|
|
Signed-off-by: Aaron Griffin <aaronmgriffin@gmail.com>
|
|
I sure didn't
Signed-off-by: Aaron Griffin <aaronmgriffin@gmail.com>
|
|
Signed-off-by: Aaron Griffin <aaronmgriffin@gmail.com>
|
|
This is silly and broken anyway. We can do this with a
cron script later
Signed-off-by: Aaron Griffin <aaronmgriffin@gmail.com>
|
|
Also, added a pretty comment. It's so pretty!
Signed-off-by: Aaron Griffin <aaronmgriffin@gmail.com>
|
|
Packages can be differentiated now by $CARCH in the
filename. No need for this extra dir.
Also, using absolute paths for binaries
Signed-off-by: Aaron Griffin <aaronmgriffin@gmail.com>
|
|
These snuck in in a previous commit
Signed-off-by: Aaron Griffin <aaronmgriffin@gmail.com>
|
|
Signed-off-by: Aaron Griffin <aaronmgriffin@gmail.com>
|
|
Yeah too much to write out. I basically rewrote it from
scratch to use repo-add and all that fun stuff.
|
|
Updating a SQL database should NOT be a task of these
scripts. They should deal ONLY with the act of maintaining
a pacman repo
Signed-off-by: Aaron Griffin <aaronmgriffin@gmail.com>
|
|
Switch to a callable script in place of an include file
Signed-off-by: Aaron Griffin <aaronmgriffin@gmail.com>
|