chiark / gitweb /
Carsten Hey [Mon, 26 Apr 2010 08:19:00 +0000 (10:19 +0200)]
Merge commit 'refs/top-bases/fixes/tg--r-require-arg' into fixes/tg--r-require-arg
* commit 'refs/top-bases/fixes/tg--r-require-arg': (31 commits)
tg patch: treat all files as ascii
tg mail: document config setting to prevent unintended mass mail sending
Don't call pre-commit hook if tg is not installed
tg-delete: Handle the case where the branch has been removed already, but the base is still left.
tg-mail: Properly quote argument to --in-reply-to
tg-push: Filter out plain SHA1s from being pushed.
Make 'tg patch' work in subdirectories
quote $LESS in case it includes spaces
deal with single quotes in from/to headers
bump version number to 0.8
Fix tg export --quilt --numbered w/ > 8 patches
tg-push: add bash completion
help: use pager
tg-push: handle non-tgish branches
tg-push: prevent handing branches more than once to git-push
tg-push: remove useless warning about missing bases for non-tgish branches
tg-update: fix user visible typo in info msg
tg-remote: don't add push specs but warn about existing ones.
tg-push: new command pushing a tgish branch
recurse_deps: if no_remotes is set don't recurse into remote deps
...
Uwe Kleine-König [Wed, 25 Mar 2009 17:21:04 +0000 (18:21 +0100)]
tg patch: treat all files as ascii
This fixes patching binary files, for a problem case check out
http://bugs.debian.org/521173 .
Closes: http://bugs.debian.org/521173
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Uwe Kleine-König [Sat, 10 Apr 2010 13:24:47 +0000 (15:24 +0200)]
tg mail: document config setting to prevent unintended mass mail sending
Addresses: http://bugs.debian.org/506306
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Carsten Hey [Sat, 10 Apr 2010 13:20:10 +0000 (15:20 +0200)]
Don't call pre-commit hook if tg is not installed
Signed-off-by: Carsten Hey <carsten@debian.org>
Closes: http://bugs.debian.org/576221
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Thomas Schwinge [Sun, 28 Feb 2010 11:39:34 +0000 (12:39 +0100)]
tg-delete: Handle the case where the branch has been removed already, but the base is still left.
A user might have used 'git branch -D B' to remove the branch B, and then
certain TopGit commands complain, because the base of branch B is still there.
Signed-off-by: Thomas Schwinge <thomas@schwinge.name>
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Thomas Schwinge [Sun, 28 Feb 2010 13:24:08 +0000 (14:24 +0100)]
tg-mail: Properly quote argument to --in-reply-to
Message-Ids usually contain < and >. These need to be quoted.
Signed-off-by: Thomas Schwinge <thomas@schwinge.name>
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Thomas Schwinge [Sun, 28 Feb 2010 13:57:41 +0000 (14:57 +0100)]
tg-push: Filter out plain SHA1s from being pushed.
Otherwise, problems arise when branches depend (via .topdeps) on SHA1s:
$ tg push -r tmp t/____longjmp_chk_cleanup_linux_x86_64
fatal:
77c84aeb81808c3109665949448dba59965c391e cannot be resolved to branch.
fatal: The remote end hung up unexpectedly
Signed-off-by: Thomas Schwinge <thomas@schwinge.name>
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Michal Sojka [Tue, 2 Feb 2010 10:31:12 +0000 (11:31 +0100)]
Make 'tg patch' work in subdirectories
When 'tg patch' is called from a subdirectory, it outputs only the
commit message and no diff.
Signed-off-by: Michal Sojka <sojkam1@fel.cvut.cz>
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
martin f. krafft [Tue, 27 Oct 2009 18:43:15 +0000 (19:43 +0100)]
quote $LESS in case it includes spaces
Signed-off-by: martin f. krafft <madduck@debian.org>
martin f. krafft [Tue, 27 Oct 2009 18:43:37 +0000 (19:43 +0100)]
deal with single quotes in from/to headers
Signed-off-by: martin f. krafft <madduck@debian.org>
Uwe Kleine-König [Wed, 23 Sep 2009 09:08:14 +0000 (11:08 +0200)]
bump version number to 0.8
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Jon Ringle [Wed, 16 Sep 2009 15:39:24 +0000 (11:39 -0400)]
Fix tg export --quilt --numbered w/ > 8 patches
Creating a quilt patch series with more than 8 patches with:
tg export --quilt --numbered patchdir
Causes the following error to be produced:
eval: 1: arithmetic expression: expecting EOF: "0008 + 1"
This is due to the leading 0 and the $(()) evaluates as octal rather
than decimal.
Add the leading zeros only to the filename, not the number itself.
Signed-off-by: Jon Ringle <jon@ringle.org>
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Bert Wesarg [Tue, 26 May 2009 23:31:34 +0000 (01:31 +0200)]
tg-push: add bash completion
Include all options.
Signed-off-by: Bert Wesarg <bert.wesarg@googlemail.com>
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Bert Wesarg [Tue, 2 Jun 2009 20:35:12 +0000 (22:35 +0200)]
help: use pager
Signed-off-by: Bert Wesarg <bert.wesarg@googlemail.com>
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Bert Wesarg [Wed, 27 May 2009 08:17:01 +0000 (10:17 +0200)]
tg-push: handle non-tgish branches
tg push died if you gave it a non-tgish branch:
$ tg push -r push-test master
fatal: Not a valid object name refs/top-bases/master
fatal: ambiguous argument '^{tree}': unknown revision or path not in the working tree.
Use '--' to separate paths from revisions
fatal: Not a valid object name master:.topdeps
Signed-off-by: Bert Wesarg <bert.wesarg@googlemail.com>
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Bert Wesarg [Wed, 27 May 2009 08:16:35 +0000 (10:16 +0200)]
tg-push: prevent handing branches more than once to git-push
Doing otherwise can result in a lot of warnings and errors:
warning, no base found top-bases/master
warning, no base found top-bases/master
warning, no base found top-bases/master
warning, no base found top-bases/master
warning, no base found top-bases/master
warning, no base found top-bases/master
warning, no base found top-bases/master
warning, no base found top-bases/master
warning, no base found top-bases/master
warning, no base found top-bases/master
warning, no base found top-bases/master
error: Ref refs/heads/master is at
b725fc951a37854fc65945ac5bcab3bc61ccbd94 but expected
0000000000000000000000000000000000000000
error: failed to lock refs/heads/master
error: Ref refs/heads/master is at
b725fc951a37854fc65945ac5bcab3bc61ccbd94 but expected
0000000000000000000000000000000000000000
error: failed to lock refs/heads/master
error: Ref refs/heads/master is at
b725fc951a37854fc65945ac5bcab3bc61ccbd94 but expected
0000000000000000000000000000000000000000
error: failed to lock refs/heads/master
error: Ref refs/heads/master is at
b725fc951a37854fc65945ac5bcab3bc61ccbd94 but expected
0000000000000000000000000000000000000000
error: failed to lock refs/heads/master
error: Ref refs/heads/master is at
b725fc951a37854fc65945ac5bcab3bc61ccbd94 but expected
0000000000000000000000000000000000000000
error: failed to lock refs/heads/master
error: Ref refs/heads/master is at
b725fc951a37854fc65945ac5bcab3bc61ccbd94 but expected
0000000000000000000000000000000000000000
error: failed to lock refs/heads/master
error: Ref refs/heads/master is at
b725fc951a37854fc65945ac5bcab3bc61ccbd94 but expected
0000000000000000000000000000000000000000
error: failed to lock refs/heads/master
error: Ref refs/heads/master is at
b725fc951a37854fc65945ac5bcab3bc61ccbd94 but expected
0000000000000000000000000000000000000000
error: failed to lock refs/heads/master
error: Ref refs/heads/master is at
b725fc951a37854fc65945ac5bcab3bc61ccbd94 but expected
0000000000000000000000000000000000000000
error: failed to lock refs/heads/master
error: Ref refs/heads/master is at
b725fc951a37854fc65945ac5bcab3bc61ccbd94 but expected
0000000000000000000000000000000000000000
error: failed to lock refs/heads/master
error: Ref refs/heads/bw/log is at
a16df35cc7009b36f6f71717ae3d9a3dc29987da but expected
0000000000000000000000000000000000000000
error: failed to lock refs/heads/bw/log
error: Ref refs/top-bases/bw/log is at
b725fc951a37854fc65945ac5bcab3bc61ccbd94 but expected
0000000000000000000000000000000000000000
error: failed to lock refs/top-bases/bw/log
! [remote rejected] master -> master (failed to lock)
! [remote rejected] master -> master (failed to lock)
! [remote rejected] master -> master (failed to lock)
! [remote rejected] master -> master (failed to lock)
! [remote rejected] master -> master (failed to lock)
! [remote rejected] master -> master (failed to lock)
! [remote rejected] master -> master (failed to lock)
! [remote rejected] master -> master (failed to lock)
! [remote rejected] master -> master (failed to lock)
! [remote rejected] master -> master (failed to lock)
! [remote rejected] bw/log -> bw/log (failed to lock)
! [remote rejected] refs/top-bases/bw/log -> refs/top-bases/bw/log (failed to lock)
Note, the result is OK only the output is irritating.
Signed-off-by: Bert Wesarg <bert.wesarg@googlemail.com>
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Bert Wesarg [Wed, 27 May 2009 08:16:22 +0000 (10:16 +0200)]
tg-push: remove useless warning about missing bases for non-tgish branches
The call to ref_exists() isn't needed because $_dep_is_tgish from
recurse_deps() gives us exactly this information.
Signed-off-by: Bert Wesarg <bert.wesarg@googlemail.com>
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Ilpo Järvinen [Thu, 28 May 2009 06:04:33 +0000 (09:04 +0300)]
tg-update: fix user visible typo in info msg
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi>
Uwe Kleine-König [Wed, 13 May 2009 09:59:51 +0000 (11:59 +0200)]
tg-remote: don't add push specs but warn about existing ones.
topgit used to add some push specs to assert that topbases are pushed,
too. This should now be handled by tg-push.
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Cc: Marc Weber <marco-oweber@gmx.de>
Tested-and-acked-by: martin f. krafft <madduck@debian.org>
Cc: 528442@bugs.debian.org
Marc Weber [Tue, 12 May 2009 08:54:51 +0000 (10:54 +0200)]
tg-push: new command pushing a tgish branch
It pushes recursively all deps and bases together with the branch.
Signed-off-by: Marc Weber <marco-oweber@gmx.de>
Tested-and-acked-by: martin f. krafft <madduck@madduck.net>
[ukleinek: fixed some nitpicks and changed remote handling]
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Acked-by: Marc Weber <marco-oweber@gmx.de>
Uwe Kleine-König [Mon, 11 May 2009 20:32:54 +0000 (22:32 +0200)]
recurse_deps: if no_remotes is set don't recurse into remote deps
This is based on a patch by Marc Weber. It will be used by tg-push.
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Acked-by: Marc Weber <marco-oweber@gmx.de>
Uwe Kleine-König [Sun, 10 May 2009 21:59:46 +0000 (23:59 +0200)]
remove test for git send-email
This test didn't work as expected with certain pagers (e.g. vimpager)
when stdout is redirected to a non-terminal. Now it's the actual call
to git send-email that generates a very similar error message.
Reported-by: Antonio Ospite <ospite@studenti.unina.it>
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Acked-and-tested-by: Antonio Ospite <ospite@studenti.unina.it>
Uwe Kleine-König [Mon, 20 Apr 2009 09:42:53 +0000 (11:42 +0200)]
depend: don't disallow adding a dependency on a non-TopGit-branch
there is no techical reason not to depend on more than one
non-TopGit-branch. You can argue if you want it though. This patch
lets you decide if you want it or not.
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Uwe Kleine-König [Thu, 26 Mar 2009 08:52:26 +0000 (09:52 +0100)]
hooks/pre-commit.sh: fix bashism
This was introduced in
fcb488d51e72c7414f9beb40ad06bf529b8b38dc.
A similar fix was suggested by martin f krafft, too.
Reported-by: Bert Wesarg <bert.wesarg@googlemail.com>
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Tested-by: Marc Kleine-Budde <mkl@pengutronix.de>
Acked-by: Bert Wesarg <bert.wesarg@googlemail.com>
Tested-by: Bert Wesarg <bert.wesarg@googlemail.com>
Uwe Kleine-König [Tue, 24 Mar 2009 21:16:48 +0000 (22:16 +0100)]
hooks/pre-commit.sh: don't stop commits on non-topgit branches
This is a regression introduced by
c32d1dd4fcafad3f9c570adeda580a90d56
"${head_#refs/heads/#refs/top-bases#}" doesn't substitute refs/heads by
refs/top-bases, but strips the prefix of $head_ if it starts with
"refs/heads/#refs/top-bases#".
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Uwe Kleine-König [Tue, 24 Mar 2009 20:46:23 +0000 (21:46 +0100)]
don't use return when being sourced
Fabien Thomas reports that a Bourne (not Again) sh on FreeBSD just
ignores the return and continues executing with the rest of the script.
So just wrap everything that was skipped by the return using other
shells in a big if instead.
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Closes: http://bugs.debian.org/516188
Uwe Kleine-König [Fri, 20 Mar 2009 19:40:07 +0000 (20:40 +0100)]
hooks/pre-commit.sh: don't wail on detached head for missing .top* files
This fixes a warning
fatal: ref HEAD is not a symbolic ref
when your HEAD is detached.
And branches under refs/top-bases/ don't need these files either.
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Uwe Kleine-König [Fri, 13 Mar 2009 09:18:17 +0000 (10:18 +0100)]
Add GPLv2 and description for S-o-b
The intend is to clearify the copyright situation. I'm not a lawyer,
but it should now be better than before.
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Uwe Kleine-König [Sun, 8 Mar 2009 21:15:16 +0000 (22:15 +0100)]
Call git-rerere after a failing merge as git-merge does
This maybe does all the hard work for you ...
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Uwe Kleine-König [Sun, 8 Mar 2009 20:52:14 +0000 (21:52 +0100)]
Fix merge failure handling in tg export
There exist two problems that show that up to now the error handling was
completely untested:
- tg.sh uses `set -e`, so `git merge $branch; retmerge=$?` obviously
doesn't do the right thing.
- in recurse_deps stdin is redirected from $_depsfile. Starting a
shell with that doesn't give the user any chance to do something.
Reported-by: "Bernhard R. Link" <brlink@debian.org>
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Uwe Kleine-König [Sat, 7 Mar 2009 20:01:30 +0000 (21:01 +0100)]
Pass -- to diff-tree for branch/filename disambiguation
Script to show the need of this patch:
mkdir ty;
cd ty;
git init-db;
touch a;
git add a;
git commit -m "...";
tg create bla;
git commit -m "...";
mkdir bla;
touch bla/blub;
git add bla/blub;
git commit -m "...";
tg export --quilt tt;
Reported-by: "Bernhard R. Link" <brlink@debian.org>
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Uwe Kleine-König [Thu, 5 Mar 2009 10:29:07 +0000 (11:29 +0100)]
bump version number to 0.7
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Uwe Kleine-König [Tue, 3 Mar 2009 10:23:22 +0000 (11:23 +0100)]
Merge commit 'refs/top-bases/fixes/tg--r-require-arg' into fixes/tg--r-require-arg
Uwe Kleine-König [Tue, 3 Mar 2009 09:31:22 +0000 (10:31 +0100)]
Fix typo s/emmail/email/
This was introduced by a tired coder in
d479474440713.
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Uwe Kleine-König [Tue, 3 Mar 2009 09:30:10 +0000 (10:30 +0100)]
Merge branch 'upstream' of git.debian.org:/git/collab-maint/topgit
Uwe Kleine-König [Tue, 3 Mar 2009 09:01:23 +0000 (10:01 +0100)]
Add documentation for tg export --linearize
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Uwe Kleine-König [Sun, 1 Mar 2009 19:55:40 +0000 (20:55 +0100)]
Don't throw away already started base on resumed create.
The problem was that after a failure to create the patch base the
resuming tg create did reset HEAD to the next tree to merge in which is
the right thing for a pristine create.
This addresses Debian Bug #513018 and is based on a patch by the bug
reporter Gábor Braun. A recipe triggering the bug is part of the report
that can be found at http://bugs.debian.org/513018.
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Uwe Kleine-König [Thu, 15 Jan 2009 21:45:57 +0000 (22:45 +0100)]
[TOPGIT] implement linearize export method
This is a draft that seems to work for my test case. The error handling
is to be improved though.
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Uwe Kleine-König [Wed, 14 Jan 2009 20:26:18 +0000 (21:26 +0100)]
[TOPGIT] make creating a commit from a topgit branch a function
This helps avoiding code duplication for the next commit.
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Uwe Kleine-König [Wed, 14 Jan 2009 20:23:51 +0000 (21:23 +0100)]
[TOPGIT] make tg remote idempotent
Before this patch each call to tg remote added three config entries
no matter if they already existed. After some time my .git/config was
crowded.
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Uwe Kleine-König [Wed, 25 Feb 2009 19:44:11 +0000 (20:44 +0100)]
[TOPGIT] allow working with annihilated branches
If you decide that you want to drop a patch, you can just merge in its
base with strategy "theirs". Then you have base=topic and so no .top*
files any more. This patch fixes tg summary and the helper function
recurse_deps() to handle these annihilated branches as if they don't
exist and don't show up in .topdeps files.
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Uwe Kleine-König [Wed, 25 Feb 2009 10:05:00 +0000 (11:05 +0100)]
[TOPGIT] limit rev-list in branch_contains to a single rev
This eases reading of debug output using sh -x and probably helps
performance, too.
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
martin f. krafft [Fri, 23 Jan 2009 05:33:43 +0000 (16:33 +1100)]
bump version number to 0.6
martin f. krafft [Fri, 23 Jan 2009 05:33:29 +0000 (16:33 +1100)]
put tg version into a variable at the top
Kirill Smelkov [Wed, 21 Jan 2009 20:18:42 +0000 (23:18 +0300)]
tg-patch: add support for generating patches against worktree and index
This implements `tg patch -i` and `tg patch -w` to see current patch as
generated against not-yet-committed index and worktree.
NOTE: unfortunately `git cat-file blob <file>` does not provide an option
to cat file from worktree (only from an object or from index), so I had to
unroll my own `cat file topic:file` with special support for '(i)' and
'(w)' topics.
Signed-off-by: Kirill Smelkov <kirr@landau.phys.spbu.ru>
Kirill Smelkov [Wed, 21 Jan 2009 20:18:42 +0000 (23:18 +0300)]
tg-patch: fix pagination
Previously, when I was invoking `tg patch` the following used to happen:
1. .topmsg content was sent directly to _terminal_
2. for each file in the patch, its diff was generated with `git diff`
and sent to *PAGER*
3. trailing 'tg: ...' was sent to terminal again
So the problem is that while `tg patch >file` works as expected, plain
`tg patch` does not -- in pager there is only a part of the whole patch
(first file diff) and header and trailer are ommitted.
I've finally decided to fix this inconvenience, and the way it works is
like in git -- we just hook `setup_pager` function in commands which
need to be paginated.
Signed-off-by: Kirill Smelkov <kirr@landau.phys.spbu.ru>
Kirill Smelkov [Wed, 21 Jan 2009 20:18:42 +0000 (23:18 +0300)]
Implement setup_pager just like in git
setup_pager() spawns a pager process and redirect the rest of our output
to it.
This will be needed to fix `tg patch` output in the next commit.
Signed-off-by: Kirill Smelkov <kirr@landau.phys.spbu.ru>
Uwe Kleine-König [Wed, 14 Jan 2009 20:23:51 +0000 (21:23 +0100)]
make tg remote idempotent
Before this patch each call to tg remote added three config entries
no matter if they already existed. After some time my .git/config was
crowded.
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
martin f. krafft [Wed, 21 Jan 2009 03:36:45 +0000 (14:36 +1100)]
Check for git-send-email and die if not found
Signed-off-by: martin f. krafft <madduck@debian.org>
Kirill Smelkov [Mon, 5 Jan 2009 15:08:20 +0000 (18:08 +0300)]
tg-completion: complete options for `tg remote`
Signed-off-by: Kirill Smelkov <kirr@landau.phys.spbu.ru>
Signed-off-by: martin f. krafft <madduck@debian.org>
Kirill Smelkov [Mon, 5 Jan 2009 15:08:19 +0000 (18:08 +0300)]
tg-completion: complete options for `tg summary`
Signed-off-by: Kirill Smelkov <kirr@landau.phys.spbu.ru>
Signed-off-by: martin f. krafft <madduck@debian.org>
Uwe Kleine-König [Fri, 19 Dec 2008 22:21:48 +0000 (23:21 +0100)]
tg export (quilt): Implement numbering the patches
To ease sending patches, with --numbered each patch gets a number prefix
similar to the output of git format-patch.
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Signed-off-by: martin f. krafft <madduck@debian.org>
Uwe Kleine-König [Fri, 19 Dec 2008 21:40:33 +0000 (22:40 +0100)]
tg export (quilt): Implement flattening patch paths
The result of providing the new flag --flatten is that the exported
patches are all placed directly in the output directory, not in
subdirectories below it.
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Signed-off-by: martin f. krafft <madduck@debian.org>
Uwe Kleine-König [Fri, 19 Dec 2008 21:18:37 +0000 (22:18 +0100)]
tg export (collapse): implement skipping empty patches
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Signed-off-by: martin f. krafft <madduck@debian.org>
Jonas Fonseca [Tue, 2 Dec 2008 07:45:46 +0000 (08:45 +0100)]
README: spelling fixes
Signed-off-by: Jonas Fonseca <fonseca@diku.dk>
Signed-off-by: martin f. krafft <madduck@debian.org>
Bert Wesarg [Wed, 26 Nov 2008 18:39:45 +0000 (19:39 +0100)]
tg-summary: -t and --graphviz are mutual exclusive
The -t and --graphviz don't work together, the output is not what one expect.
Signed-off-by: Bert Wesarg <bert.wesarg@googlemail.com>
Signed-off-by: martin f. krafft <madduck@debian.org>
Uwe Kleine-König [Wed, 26 Nov 2008 12:13:00 +0000 (13:13 +0100)]
tg-export: implement skipping empty patches for quilt mode
addionally fix the README item for skipping the export of empty patches
not to need an option (-n) as this should be the default.
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Signed-off-by: martin f. krafft <madduck@debian.org>
Marc Weber [Fri, 21 Nov 2008 09:05:30 +0000 (10:05 +0100)]
Pass -- to rev-list for branch/filename disambiguation
While playing around it happened that I called the topic branch and
a file "foo"... Here is a small incomplete patch: I've not tried to
catch all of these errors..
Signed-off-by: martin f. krafft <madduck@debian.org>
martin f. krafft [Thu, 20 Nov 2008 14:46:05 +0000 (15:46 +0100)]
Add Vim modelines for consistent spacing
TopGit shell scripts use tabs and this commit instructs Vim via modeline not
to expand tabs.
Signed-off-by: martin f. krafft <madduck@debian.org>
Acked-by: Petr Baudis <pasky@ucw.cz>
martin f. krafft [Thu, 20 Nov 2008 14:24:16 +0000 (15:24 +0100)]
Note that do_help is used when short messages might be wanted
From http://marc.info/?l=git&m=
122718711327376&w=2, by Petr:
Note that I think these changes are reasonable only as long as do_help
is just a short two-line usage help; in the event we also implement a
longer help, it certainly shouldn't be printed in these cases.
This commit adds a comment to the do_help function to remind us.
Signed-off-by: martin f. krafft <madduck@debian.org>
martin f. krafft [Thu, 20 Nov 2008 14:06:19 +0000 (15:06 +0100)]
Print help message when command is not proper
When invoked with an unknown subcommand, or when --help is sought for
a command that does not exist, TopGit prints the help message.
Signed-off-by: martin f. krafft <madduck@debian.org>
Acked-by: Petr Baudis <pasky@suse.cz>
martin f. krafft [Thu, 20 Nov 2008 14:04:48 +0000 (15:04 +0100)]
Require an argument to tg -r
Right now, if -r is passed to tg, an ugly shift error occurs. This patch
checks for that case and outputs the help message if it occurs.
Signed-off-by: martin f. krafft <madduck@debian.org>
Acked-by: Petr Baudis <pasky@suse.cz>
martin f. krafft [Thu, 20 Nov 2008 14:04:18 +0000 (15:04 +0100)]
fix indentation
Signed-off-by: martin f. krafft <madduck@debian.org>
martin f. krafft [Thu, 20 Nov 2008 12:07:53 +0000 (13:07 +0100)]
amend patch subject
Signed-off-by: martin f. krafft <madduck@debian.org>
martin f. krafft [Thu, 20 Nov 2008 12:03:29 +0000 (13:03 +0100)]
finish
martin f. krafft [Thu, 20 Nov 2008 12:01:38 +0000 (13:01 +0100)]
initial
martin f. krafft [Wed, 19 Nov 2008 15:10:23 +0000 (16:10 +0100)]
Print help output when no command is given
As much as I love your message (although you really ought not be
throwing around ducks!), this is more user-friendly.
Signed-off-by: martin f. krafft <madduck@debian.org>
Acked-by: Petr Baudis <pasky@suse.cz>
martin f. krafft [Wed, 19 Nov 2008 15:08:02 +0000 (16:08 +0100)]
Check for cmddir earlier
Without cmddir, tg is basically useless, even do_help() needs it, so
check it first and die hard if not found
Signed-off-by: martin f. krafft <madduck@debian.org>
Acked-by: Petr Baudis <pasky@suse.cz>
martin f. krafft [Wed, 19 Nov 2008 15:07:02 +0000 (16:07 +0100)]
Change tg help exit code to 0
Printing --help is not an error, but a successful operation, if the help
output could be printed.
Signed-off-by: martin f. krafft <madduck@debian.org>
Acked-by: Petr Baudis <pasky@suse.cz>
martin f. krafft [Wed, 19 Nov 2008 14:44:15 +0000 (15:44 +0100)]
add ignore patterns for quilt and debian build
Signed-off-by: martin f. krafft <madduck@debian.org>
martin f. krafft [Wed, 19 Nov 2008 14:42:47 +0000 (15:42 +0100)]
Make sure gitignore patterns are not recursive
martin f. krafft [Wed, 19 Nov 2008 14:20:16 +0000 (15:20 +0100)]
remove +x bit from tg-depend
martin f. krafft [Mon, 17 Nov 2008 21:57:35 +0000 (22:57 +0100)]
ignore tg-depend build files
Signed-off-by: martin f. krafft <madduck@debian.org>
martin f. krafft [Mon, 17 Nov 2008 21:34:13 +0000 (22:34 +0100)]
Update version in tg script to 0.5
Signed-off-by: martin f. krafft <madduck@debian.org>
Matt McDonald [Fri, 14 Nov 2008 00:15:37 +0000 (19:15 -0500)]
Make sure $root_dir does not have a trailing slash
This adds support for using 'tg create' from within a subdirectory.
Apparently, git doesn't seem a do a good job of normalizing paths.
When topgit tries to add a new .topdeps file in a newly created
branch, git bails out with:
$ tg create new-branch
tg: Automatically marking dependency on old-branch
tg: Creating new-branch base from old-branch...
Switched to a new branch "new-branch"
(here, tg tries to do: git add ../../..//.topdeps)
fatal: pathspec '/.topdeps' did not match any files
Since $root_dir is used everywhere like: "$root_dir/.topdeps", this
patch tries to make sure that $root_dir never has the trailing slash
so it always produces safe paths.
Signed-off-by: martin f. krafft <madduck@madduck.net>
Matt McDonald [Sat, 8 Nov 2008 19:01:02 +0000 (14:01 -0500)]
tg depend: Allow adding deps from a subdir inside the repo.
martin f. krafft [Fri, 26 Sep 2008 19:17:12 +0000 (21:17 +0200)]
tg-export: no current branch check with -b
When --quilt .. -b is given, TopGit dies with
piper:../pkg/topgit/topgit|build|% tg export --quilt -b debian/locations debian/patches
tg: fatal: not on a TopGit-controlled branch
which is a little too restrictive as we don't have to be on a tg branch with
-b. This patch simply removes this check if -b has been passed.
Signed-off-by: martin f. krafft <madduck@madduck.net>
martin f. krafft [Fri, 26 Sep 2008 19:04:19 +0000 (21:04 +0200)]
tg-mail: do not use arrays, which are bashisms
tg-mail uses shell arrays for the people array, which are not supported by
POSIX. The script runs as /bin/sh, so it cannot assume non-POSIX features.
This patch reimplements the array population with a POSIX-compliant approach.
Signed-off-by: martin f. krafft <madduck@madduck.net>
Antonio Ospite [Tue, 28 Oct 2008 17:54:14 +0000 (18:54 +0100)]
TopGit: Force adding .topdeps on tg-depend
Force adding .topdeps, overcomes the case when .* is in .gitignore
Signed-off-by: Antonio Ospite <ospite@studenti.unina.it>
Antonio Ospite [Fri, 24 Oct 2008 10:54:24 +0000 (12:54 +0200)]
TopGit: small Makefile nitpichink
Make 'prefix' replacement more explicit using ?=
Use simple substitution assignment := when there are some expansions in
place. This is a good practice to speed up builds, maybe not particularly
useful with such a simple Makefile.
Signed-off-by: Antonio Ospite <ospite@studenti.unina.it>
Petr Baudis [Fri, 7 Nov 2008 21:16:37 +0000 (22:16 +0100)]
README: tg depend rm TODO (insp. by Jonas)
Petr Baudis [Thu, 16 Oct 2008 10:35:16 +0000 (12:35 +0200)]
tg summary --graphviz: Dot-suitable dependency graph
Inspired by Antonio Ospite <ospite@studenti.unina.it>'s tg-graphviz.sh.
Petr Baudis [Thu, 16 Oct 2008 10:10:29 +0000 (12:10 +0200)]
README: lever -> level (spotted by jikos)
Petr Baudis [Fri, 3 Oct 2008 15:23:08 +0000 (17:23 +0200)]
tg export: Clarify usage
Petr Baudis [Fri, 3 Oct 2008 15:22:56 +0000 (17:22 +0200)]
tg mail -r MSGID: Generate In-reply-to header
Petr Baudis [Thu, 2 Oct 2008 11:09:14 +0000 (13:09 +0200)]
tg update: Always checkout head, even if we didn't update base further
Petr Baudis [Thu, 25 Sep 2008 18:50:36 +0000 (20:50 +0200)]
tg update: Fix bogus advice on failed base update
Petr Baudis [Thu, 25 Sep 2008 18:50:25 +0000 (20:50 +0200)]
tg import: More graceful conflicts handling
Petr Baudis [Thu, 25 Sep 2008 18:14:51 +0000 (20:14 +0200)]
tg import: Construct actually proper .topmsg file
Petr Baudis [Thu, 25 Sep 2008 18:11:57 +0000 (20:11 +0200)]
tg import -s: Whitespace fix
Petr Baudis [Thu, 25 Sep 2008 18:11:38 +0000 (20:11 +0200)]
tg import -d: Explicitly specify base dependency of the created branches
Petr Baudis [Thu, 25 Sep 2008 18:06:57 +0000 (20:06 +0200)]
tg import -s: Import single commit using well-defined name
Petr Baudis [Thu, 25 Sep 2008 16:55:28 +0000 (18:55 +0200)]
tg mail: Fix $header string construction
Prevents e.g. misinterpreting From: line at the body start.
Petr Baudis [Thu, 25 Sep 2008 14:55:41 +0000 (16:55 +0200)]
tg-completion: Use tg summary -t for __tg_topics()
Petr Baudis [Thu, 25 Sep 2008 14:55:09 +0000 (16:55 +0200)]
tg summary -t: Very quick mode, only branch names
Jonas Fonseca [Thu, 25 Sep 2008 13:47:06 +0000 (15:47 +0200)]
Complete depend subcommand
Also, removes a debug line and fixes the script header.
Signed-off-by: Jonas Fonseca <fonseca@diku.dk>
Petr Baudis [Thu, 25 Sep 2008 13:26:05 +0000 (15:26 +0200)]
tg depend: Even more robust subcommand handling
Jonas Fonseca [Thu, 25 Sep 2008 10:50:40 +0000 (12:50 +0200)]
depend: Fix help listing
Without this `tg help depend` causes all kinds of trouble such as adding
empty lines to .topdeps.
Signed-off-by: Jonas Fonseca <fonseca@diku.dk>
Jonas Fonseca [Thu, 25 Sep 2008 11:06:56 +0000 (13:06 +0200)]
Add bash completion script for TopGit
The script is based on git's bash completion script.
Signed-off-by: Jonas Fonseca <fonseca@diku.dk>
Petr Baudis [Thu, 25 Sep 2008 13:16:57 +0000 (15:16 +0200)]
README: Add warning about editing .topdeps manually