Add git-reup and friends

Add a family of git-commands to assist with the management of multiple CLs/branches.

git-rebase-update - ensure all branches are up to date
git-new-branch - create branches
git-rename-branch - rename a branch while preserving parentage relationships
git-reparent-branch - change the parent of a branch, including rebasing it correctly onto that new parent.
git-squash-branch - collapse a branch into a single commit
git-upstream-diff - show the diff between the current branch and it's upstream branch
git-mark-merge-base - explicitly set what you want the above tools to consider the merge-base for the current branch.

R=agable@chromium.org, hinoka@chromium.org, stip@chromium.org, szager@chromium.org
BUG=261738
Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=259520

[agable, 2014-02-28 20:14:16]

https://codereview.chromium.org/184253003/diff/1/git-upstream-diff
File git-upstream-diff (right):

https://codereview.chromium.org/184253003/diff/1/git-upstream-diff#newcode6
git-upstream-diff:6: git diff $(git merge-base '@{u}' HEAD) --patience
--word-diff=color --word-diff-regex='\\w+|[^[:space:]]' -C -C
Word diff is problematic and it's not what people expect. Don't add it here, or
if you do, make it an option.

https://codereview.chromium.org/184253003/diff/1/git_common.py
File git_common.py (right):

https://codereview.chromium.org/184253003/diff/1/git_common.py#newcode30
git_common.py:30: MERGE_BASE_TAG_FMT = "gitscripts.merge_base_for_%s"
Ugly that some of these are using a gitscripts config namespace and some are
using a depot_tools namespace.

https://codereview.chromium.org/184253003/diff/1/git_common.py#newcode230
git_common.py:230: def clean_refs():
This is really only cleaning tags, not refs...

https://codereview.chromium.org/184253003/diff/1/git_common.py#newcode332
git_common.py:332: def nuke_merge_base_tag(branch):
s/nuke/remove/

https://codereview.chromium.org/184253003/diff/1/git_merge_base_tag.py
File git_merge_base_tag.py (right):

https://codereview.chromium.org/184253003/diff/1/git_merge_base_tag.py#newcode17
git_merge_base_tag.py:17: else:
"git merge-base-tag other-branch" creates a tag.
"git merge-base-tag" removes a tag.

That's semantically crazy. You need to enforce "-d" or something like that to
remove the tag, like removing branches.

In fact, I'd say "git merge-base-tag other-branch" should create a tag, "git
merge-base-tag -d other-branch" should remove the merge-base tag with that other
branch, and any merge-base-tag command without a branch name should assume HEAD.

https://codereview.chromium.org/184253003/diff/1/git_rebase_update.py
File git_rebase_update.py (right):

https://codereview.chromium.org/184253003/diff/1/git_rebase_update.py#newcode14
git_rebase_update.py:14: 
No newline here, and put git_squash_branch bellow git_common.

https://codereview.chromium.org/184253003/diff/1/git_rebase_update.py#newcode64
git_rebase_update.py:64: if (hash_one(parent) != hash_one(starting_ref)
Main line of code should be at top level. Reverse the test and return inside it.

https://codereview.chromium.org/184253003/diff/1/git_rebase_update.py#newcode86
git_rebase_update.py:86: assert final_rebase.success == squash_ret.success
Add error message here.

https://codereview.chromium.org/184253003/diff/1/git_rebase_update.py#newcode97
git_rebase_update.py:97: # TODO(iannucci): use argparse
Probably do this now (less important than the other use of argparse, since
you're only getting one argument right now).

https://codereview.chromium.org/184253003/diff/1/git_rename_branch.py
File git_rename_branch.py (right):

https://codereview.chromium.org/184253003/diff/1/git_rename_branch.py#newcode11
git_rename_branch.py:11: # TODO(iannucci): Use argparse
Do this now.

https://codereview.chromium.org/184253003/diff/1/git_rename_branch.py#newcode19
git_rename_branch.py:19: assert old_name in branches(), "<oldname> must exist"
Since it shells out, only make one call to branches()

https://codereview.chromium.org/184253003/diff/1/git_reparent_branch.py
File git_reparent_branch.py (right):

https://codereview.chromium.org/184253003/diff/1/git_reparent_branch.py#newco...
git_reparent_branch.py:14: assert branch != 'HEAD', 'Must be on the branch you
want to reparent'
Could let them specify the branch they want to reparent on the command line too.

[Ryan Tseng, 2014-02-28 21:22:05]

Usage/help strings will be helpful, I have no idea what most of these do based
on the name

https://codereview.chromium.org/184253003/diff/1/git_rebase_update.py
File git_rebase_update.py (right):

https://codereview.chromium.org/184253003/diff/1/git_rebase_update.py#newcode5
git_rebase_update.py:5: import logging
2 lines before line 5 + sort

https://codereview.chromium.org/184253003/diff/1/git_rebase_update.py#newcode22
git_rebase_update.py:22: def bclean():
What does this stand for?

https://codereview.chromium.org/184253003/diff/1/git_rebase_update.py#newcode91
git_rebase_update.py:91: print 'Your working copy is in mid-rebase. Please
completely resolve and'
Does this leave HEAD mid rebase?
I think we'll need some instructions about what they do next, (edit files,
remove >>>, ===, <<< lines, add files, git --rebase continue, or git  rebase
--abort to exit).  "Completely resolve" is a bit confusing for git noobies.

https://codereview.chromium.org/184253003/diff/1/git_rebase_update.py#newcode97
git_rebase_update.py:97: # TODO(iannucci): use argparse
We at least need a usage string behind --help.  "reup" is not self evident.

https://codereview.chromium.org/184253003/diff/1/git_rebase_update.py#newcode118
git_rebase_update.py:118: if 'origin' in run('remote').splitlines():
splitlines() is probably unnecessary.

https://codereview.chromium.org/184253003/diff/1/git_rebase_update.py#newcode119
git_rebase_update.py:119: run('fetch', 'origin', stderr=None)
This might take a while, can we stream the output?

https://codereview.chromium.org/184253003/diff/1/git_rebase_update.py#newcode137
git_rebase_update.py:137: # XXX There is a more efficient way to do this, but
for now...
XXX?

https://codereview.chromium.org/184253003/diff/1/git_rename_branch.py
File git_rename_branch.py (right):

https://codereview.chromium.org/184253003/diff/1/git_rename_branch.py#newcode11
git_rename_branch.py:11: # TODO(iannucci): Use argparse
How is this different than just "git branch -m"?

https://codereview.chromium.org/184253003/diff/1/git_rename_branch.py#newcode19
git_rename_branch.py:19: assert old_name in branches(), "<oldname> must exist"
"Error: Branch \"%s\" does not exist" % old_name

https://codereview.chromium.org/184253003/diff/1/git_rename_branch.py#newcode20
git_rename_branch.py:20: assert new_name not in branches(), "<newname> must not
exist"
"Error: A branch named \"%s\" already exists" % new_name

https://codereview.chromium.org/184253003/diff/1/git_reparent_branch.py
File git_reparent_branch.py (right):

https://codereview.chromium.org/184253003/diff/1/git_reparent_branch.py#newcode8
git_reparent_branch.py:8: # TODO(iannucci): Add --verbose
Add --help.  "reparent branch" also isn't self evident.  What does this do?  Is
this just an alias for "rebase --onto x y z"? Is it just an alias for
--set-upstream?

https://codereview.chromium.org/184253003/diff/1/git_squash_branch.py
File git_squash_branch.py (right):

https://codereview.chromium.org/184253003/diff/1/git_squash_branch.py#newcode19
git_squash_branch.py:19: def main():
Add usage and --help

[szager1, 2014-02-28 22:56:39]

https://codereview.chromium.org/184253003/diff/1/git-merge-base-tag
File git-merge-base-tag (right):

https://codereview.chromium.org/184253003/diff/1/git-merge-base-tag#newcode6
git-merge-base-tag:6: . ${0/%$(basename "$0")/python_git_runner.sh}
${0%/*}/python_git_runner.sh

And where is this "python_git_runner.sh" ?

https://codereview.chromium.org/184253003/diff/1/git-rebase-update
File git-rebase-update (right):

https://codereview.chromium.org/184253003/diff/1/git-rebase-update#newcode6
git-rebase-update:6: . ${0/%$(basename "$0")/python_git_runner.sh}
Same comment.

https://codereview.chromium.org/184253003/diff/1/git-rename-branch
File git-rename-branch (right):

https://codereview.chromium.org/184253003/diff/1/git-rename-branch#newcode6
git-rename-branch:6: . ${0/%$(basename "$0")/python_git_runner.sh}
Same comment.

https://codereview.chromium.org/184253003/diff/1/git-reparent-branch
File git-reparent-branch (right):

https://codereview.chromium.org/184253003/diff/1/git-reparent-branch#newcode6
git-reparent-branch:6: . ${0/%$(basename "$0")/python_git_runner.sh}
Same comment.

https://codereview.chromium.org/184253003/diff/1/git-squash-branch
File git-squash-branch (right):

https://codereview.chromium.org/184253003/diff/1/git-squash-branch#newcode6
git-squash-branch:6: . ${0/%$(basename "$0")/python_git_runner.sh}
Same comment.

https://codereview.chromium.org/184253003/diff/1/git-upstream-diff
File git-upstream-diff (right):

https://codereview.chromium.org/184253003/diff/1/git-upstream-diff#newcode6
git-upstream-diff:6: git diff $(git merge-base '@{u}' HEAD) --patience
--word-diff=color --word-diff-regex='\\w+|[^[:space:]]' -C -C
On 2014/02/28 20:14:16, agable wrote:
> Word diff is problematic and it's not what people expect. Don't add it here,
or
> if you do, make it an option.

Yeah, this is a personal preference.  Providing shorthand is OK, but it's kinda
weird as default behavior.

[iannucci, 2014-03-20 10:45:50]

OK, lots of fixes (basically 100% different), plus tests. PTAL

manpages coming soon.

[iannucci, 2014-03-20 10:59:37]

https://codereview.chromium.org/184253003/diff/1/git_rebase_update.py
File git_rebase_update.py (right):

https://codereview.chromium.org/184253003/diff/1/git_rebase_update.py#newcode91
git_rebase_update.py:91: print 'Your working copy is in mid-rebase. Please
completely resolve and'
On 2014/02/28 21:22:05, Ryan T. wrote:
> Does this leave HEAD mid rebase?
> I think we'll need some instructions about what they do next, (edit files,
> remove >>>, ===, <<< lines, add files, git --rebase continue, or git  rebase
> --abort to exit).  "Completely resolve" is a bit confusing for git noobies.

It also prints the standard git rebase boilerplate. I don't think adding more
output here is appropriate. I also added some more checks in case they try to
run re-up without resolving things.

https://codereview.chromium.org/184253003/diff/1/git_reparent_branch.py
File git_reparent_branch.py (right):

https://codereview.chromium.org/184253003/diff/1/git_reparent_branch.py#newco...
git_reparent_branch.py:14: assert branch != 'HEAD', 'Must be on the branch you
want to reparent'
On 2014/02/28 20:14:16, agable wrote:
> Could let them specify the branch they want to reparent on the command line
too.

added TODO

[agable, 2014-03-21 01:14:20]

Incomplete review (haven't looked at all of reup yet) but I have 9 more reviews
to do before I leave today so I'm leaving this here and will come back to it.

https://codereview.chromium.org/184253003/diff/60001/git-reparent-branch
File git-reparent-branch (right):

https://codereview.chromium.org/184253003/diff/60001/git-reparent-branch#newc...
git-reparent-branch:7: # branch. Afterwards, run a `git rebase-update` cycle to
ensure that all branches
80 chars

https://codereview.chromium.org/184253003/diff/60001/git_common.py
File git_common.py (right):

https://codereview.chromium.org/184253003/diff/60001/git_common.py#newcode253
git_common.py:253: skipped = []
set?

https://codereview.chromium.org/184253003/diff/60001/git_common.py#newcode289
git_common.py:289: def run_both(*cmd, **kwargs):
run_both? really? least descriptive name evar.

https://codereview.chromium.org/184253003/diff/60001/git_common.py#newcode318
git_common.py:318: def stream(*cmd, **kwargs):
Also poorly named. Since it's executing a command too, not just streaming
something somewhere. Also doesn't mention that it drops stderr on the floor.

https://codereview.chromium.org/184253003/diff/60001/git_common.py#newcode409
git_common.py:409: """Rebases |start|..|branch| onto the branch |parent|.
start..branch doesn't actually include start, in git syntax. This method does
include start in the set of commits that get rebased.

https://codereview.chromium.org/184253003/diff/60001/git_merge_base_tag.py
File git_merge_base_tag.py (right):

https://codereview.chromium.org/184253003/diff/60001/git_merge_base_tag.py#ne...
git_merge_base_tag.py:37: remove_merge_base(current_branch())
running this with no argument removes the tag? That's unintuitive.

https://codereview.chromium.org/184253003/diff/60001/git_new_branch.py
File git_new_branch.py (right):

https://codereview.chromium.org/184253003/diff/60001/git_new_branch.py#newcode24
git_new_branch.py:24: '--lkgr', action='store_const', const=('.',
'refs/tags/lkgr'),
indent +4

https://codereview.chromium.org/184253003/diff/60001/git_new_branch.py#newcode34
git_new_branch.py:34: if opts.basis_ref:
You implemented a general approach to tracking tags, but only allow lkgr to be
tracked. This is especially weird considering lkgr isn't even a tag yet, and
won't be until flag day.

https://codereview.chromium.org/184253003/diff/60001/git_rebase_update.py
File git_rebase_update.py (right):

https://codereview.chromium.org/184253003/diff/60001/git_rebase_update.py#new...
git_rebase_update.py:5: 
docstrong

https://codereview.chromium.org/184253003/diff/60001/git_rebase_update.py#new...
git_rebase_update.py:13: from git_common import run, root, tags, topo_iter,
once, hash_one, rebase
Honestly might be more reasonable to do
import git_common as git
at this point

https://codereview.chromium.org/184253003/diff/60001/git_rename_branch.py
File git_rename_branch.py (right):

https://codereview.chromium.org/184253003/diff/60001/git_rename_branch.py#new...
git_rename_branch.py:5: 
docstring

https://codereview.chromium.org/184253003/diff/60001/git_rename_branch.py#new...
git_rename_branch.py:12: # TODO(iannucci): Use argparse
Do this now

https://codereview.chromium.org/184253003/diff/60001/git_rename_branch.py#new...
git_rename_branch.py:28: key, value = line.split('=', 1)
does this properly strip? Lots of gitconfigs have tabs in them, and I don't want
to check for "value == 'refs/heads/\told_name'"

https://codereview.chromium.org/184253003/diff/60001/git_rename_branch.py#new...
git_rename_branch.py:34: if remote == '.':
Please comment on what the special value '.' means.

https://codereview.chromium.org/184253003/diff/60001/git_reparent_branch.py
File git_reparent_branch.py (right):

https://codereview.chromium.org/184253003/diff/60001/git_reparent_branch.py#n...
git_reparent_branch.py:5: 
docstring

https://codereview.chromium.org/184253003/diff/60001/git_reparent_branch.py#n...
git_reparent_branch.py:18: assert len(argv) == 2, "Must supply new parent"
use argparse

https://codereview.chromium.org/184253003/diff/60001/git_reparent_branch.py#n...
git_reparent_branch.py:32: print "Resetting parent back to %s" % (cur_parent)
Print failure message in between; otherwise not clear why it is setting back.

https://codereview.chromium.org/184253003/diff/60001/git_squash_branch.py
File git_squash_branch.py (right):

https://codereview.chromium.org/184253003/diff/60001/git_squash_branch.py#new...
git_squash_branch.py:5: 
docstring

https://codereview.chromium.org/184253003/diff/60001/git_squash_branch.py#new...
git_squash_branch.py:11: # TODO(iannucci): Use argparse
do this now

https://codereview.chromium.org/184253003/diff/60001/git_upstream_diff.py
File git_upstream_diff.py (right):

https://codereview.chromium.org/184253003/diff/60001/git_upstream_diff.py#new...
git_upstream_diff.py:21: 'Print a colorized wordwise diff '
should be at +4

https://codereview.chromium.org/184253003/diff/60001/tests/git_rebase_update_...
File tests/git_rebase_update_test.py (right):

https://codereview.chromium.org/184253003/diff/60001/tests/git_rebase_update_...
tests/git_rebase_update_test.py:178: def testTrackTag(self):
use git-new-branch --lkgr in this test?

[iannucci, 2014-03-22 04:17:35]

PTAL

https://codereview.chromium.org/184253003/diff/60001/git-reparent-branch
File git-reparent-branch (right):

https://codereview.chromium.org/184253003/diff/60001/git-reparent-branch#newc...
git-reparent-branch:7: # branch. Afterwards, run a `git rebase-update` cycle to
ensure that all branches
On 2014/03/21 01:14:21, agable wrote:
> 80 chars

Done.

https://codereview.chromium.org/184253003/diff/60001/git_common.py
File git_common.py (right):

https://codereview.chromium.org/184253003/diff/60001/git_common.py#newcode253
git_common.py:253: skipped = []
On 2014/03/21 01:14:21, agable wrote:
> set?

Done.

https://codereview.chromium.org/184253003/diff/60001/git_common.py#newcode289
git_common.py:289: def run_both(*cmd, **kwargs):
On 2014/03/21 01:14:21, agable wrote:
> run_both? really? least descriptive name evar.

Done.

https://codereview.chromium.org/184253003/diff/60001/git_common.py#newcode318
git_common.py:318: def stream(*cmd, **kwargs):
On 2014/03/21 01:14:21, agable wrote:
> Also poorly named. Since it's executing a command too, not just streaming
> something somewhere. Also doesn't mention that it drops stderr on the floor.

Done.

https://codereview.chromium.org/184253003/diff/60001/git_common.py#newcode409
git_common.py:409: """Rebases |start|..|branch| onto the branch |parent|.
On 2014/03/21 01:14:21, agable wrote:
> start..branch doesn't actually include start, in git syntax. This method does
> include start in the set of commits that get rebased.

Er... no it doesn't include start, which is why I used A..B :).

I updated git_common_test:....testRebase to more clearly illustrate this.

https://codereview.chromium.org/184253003/diff/60001/git_merge_base_tag.py
File git_merge_base_tag.py (right):

https://codereview.chromium.org/184253003/diff/60001/git_merge_base_tag.py#ne...
git_merge_base_tag.py:37: remove_merge_base(current_branch())
On 2014/03/21 01:14:21, agable wrote:
> running this with no argument removes the tag? That's unintuitive.

Well, the helptext does say that. I guess no args should be 'show' and '-d'
should be remove.

Done (and improved).

https://codereview.chromium.org/184253003/diff/60001/git_new_branch.py
File git_new_branch.py (right):

https://codereview.chromium.org/184253003/diff/60001/git_new_branch.py#newcode24
git_new_branch.py:24: '--lkgr', action='store_const', const=('.',
'refs/tags/lkgr'),
On 2014/03/21 01:14:21, agable wrote:
> indent +4

Done.

https://codereview.chromium.org/184253003/diff/60001/git_new_branch.py#newcode34
git_new_branch.py:34: if opts.basis_ref:
On 2014/03/21 01:14:21, agable wrote:
> You implemented a general approach to tracking tags, but only allow lkgr to be
> tracked. This is especially weird considering lkgr isn't even a tag yet, and
> won't be until flag day.

Done.

https://codereview.chromium.org/184253003/diff/60001/git_rename_branch.py
File git_rename_branch.py (right):

https://codereview.chromium.org/184253003/diff/60001/git_rename_branch.py#new...
git_rename_branch.py:34: if remote == '.':
On 2014/03/21 01:14:21, agable wrote:
> Please comment on what the special value '.' means.

Er... why here? This is defined by git-config...

https://codereview.chromium.org/184253003/diff/60001/git_reparent_branch.py
File git_reparent_branch.py (right):

https://codereview.chromium.org/184253003/diff/60001/git_reparent_branch.py#n...
git_reparent_branch.py:32: print "Resetting parent back to %s" % (cur_parent)
On 2014/03/21 01:14:21, agable wrote:
> Print failure message in between; otherwise not clear why it is setting back.

The rebase-update command already prints a lot of error material when it can't
do it's operation cleanly.

https://codereview.chromium.org/184253003/diff/60001/tests/git_rebase_update_...
File tests/git_rebase_update_test.py (right):

https://codereview.chromium.org/184253003/diff/60001/tests/git_rebase_update_...
tests/git_rebase_update_test.py:178: def testTrackTag(self):
On 2014/03/21 01:14:21, agable wrote:
> use git-new-branch --lkgr in this test?

Done.

[iannucci, 2014-03-23 09:46:39]

OK, man pages are up for everything now too.

[agable, 2014-03-25 19:37:23]

https://codereview.chromium.org/184253003/diff/180001/docs/html/git-freeze.html
File docs/html/git-freeze.html (right):

https://codereview.chromium.org/184253003/diff/180001/docs/html/git-freeze.ht...
docs/html/git-freeze.html:844: Last updated 2014-03-22 12:59:00 PDT
Is there any way these timestamps can be generated only if the rest of the text
actually differs from the previous version?

https://codereview.chromium.org/184253003/diff/180001/docs/src/git-new-branch...
File docs/src/git-new-branch.txt (right):

https://codereview.chromium.org/184253003/diff/180001/docs/src/git-new-branch...
docs/src/git-new-branch.txt:27: the default upstream (i.e. `git new-branch
<branch_name>`). If you have features
link to new-branch doc

https://codereview.chromium.org/184253003/diff/180001/docs/src/git-new-branch...
docs/src/git-new-branch.txt:38: Set the tracking (upstream) branch to <REF>.
<REF> may be a local branch,
80 chars

https://codereview.chromium.org/184253003/diff/180001/docs/src/git-rebase-upd...
File docs/src/git-rebase-update.txt (right):

https://codereview.chromium.org/184253003/diff/180001/docs/src/git-rebase-upd...
docs/src/git-rebase-update.txt:21: If you currently have a branch checked out,
any changes on that branch are
Lots of >80 character lines in this file.

https://codereview.chromium.org/184253003/diff/180001/docs/src/git-rebase-upd...
docs/src/git-rebase-update.txt:34: All branches are rebased in topological order
from roots (upstreams) to leafs.
leaves. We're not rebasing a hockey team.

https://codereview.chromium.org/184253003/diff/180001/docs/src/git-rebase-upd...
docs/src/git-rebase-update.txt:97: branch.<name>.base
I think most of this documentation should be in mark-merge-base. I find it
really annoying when most of the docs for a feature are in the docs for the
primary customer of that feature, not the feature itself.

https://codereview.chromium.org/184253003/diff/180001/docs/src/git-squash-bra...
File docs/src/git-squash-branch.txt (right):

https://codereview.chromium.org/184253003/diff/180001/docs/src/git-squash-bra...
docs/src/git-squash-branch.txt:12: 'git squash-branch' [<header message>]
Preferably 'git squash-branch -m <header message>'.

https://codereview.chromium.org/184253003/diff/180001/docs/src/git-upstream-d...
File docs/src/git-upstream-diff.txt (right):

https://codereview.chromium.org/184253003/diff/180001/docs/src/git-upstream-d...
docs/src/git-upstream-diff.txt:28: your branch (See
linkgit:git-rebase-update[1]). This means that if your upstream
linkgit:git-mark-merge-base

https://codereview.chromium.org/184253003/diff/180001/docs/src/git-upstream-d...
docs/src/git-upstream-diff.txt:45: Extra arguments are included in the
invocation of linkgit:git-diff[1]. These
80 chars

https://codereview.chromium.org/184253003/diff/180001/git_mark_merge_base.py
File git_mark_merge_base.py (right):

https://codereview.chromium.org/184253003/diff/180001/git_mark_merge_base.py#...
git_mark_merge_base.py:27: epilog=''.join(__doc__.strip().splitlines()[1:]))
Do ' '.join(blah) and get rid of the leading spaces in the docstring.

https://codereview.chromium.org/184253003/diff/180001/git_new_branch.py
File git_new_branch.py (right):

https://codereview.chromium.org/184253003/diff/180001/git_new_branch.py#newco...
git_new_branch.py:24: help='upstream branch to track.')
(or tag)

https://codereview.chromium.org/184253003/diff/180001/git_rebase_update.py
File git_rebase_update.py (right):

https://codereview.chromium.org/184253003/diff/180001/git_rebase_update.py#ne...
git_rebase_update.py:5: 
module docstring

https://codereview.chromium.org/184253003/diff/180001/git_rebase_update.py#ne...
git_rebase_update.py:15: from git_freezer import freeze, thaw
See similar comment on another file -- I think the only interface between these
scripts should be git_common and their command line interfaces. They shouldn't
import each other.

https://codereview.chromium.org/184253003/diff/180001/git_rebase_update.py#ne...
git_rebase_update.py:22: """Finds the branch which we should return to after
rebase-update completes
single line for first line of docstring

https://codereview.chromium.org/184253003/diff/180001/git_rebase_update.py#ne...
git_rebase_update.py:35: from git_common import in_rebase
remove import

https://codereview.chromium.org/184253003/diff/180001/git_rebase_update.py#ne...
git_rebase_update.py:37: if in_rebase():
git.in_rebase

https://codereview.chromium.org/184253003/diff/180001/git_rebase_update.py#ne...
git_rebase_update.py:40: print (
This print is out of place. This method should either just return True/False, or
it should print this message and exit entirely.

https://codereview.chromium.org/184253003/diff/180001/git_rebase_update.py#ne...
git_rebase_update.py:50: 
no newline

https://codereview.chromium.org/184253003/diff/180001/git_rebase_update.py#ne...
git_rebase_update.py:97: git.set_config('branch.%s.remote' % down, '.')
You now have the set_branch_config methods

https://codereview.chromium.org/184253003/diff/180001/git_rebase_update.py#ne...
git_rebase_update.py:149: print 'Squashing failed. You probably have a real
merge conflict.'
Consider printing a textwrap.dedented triple-quoted string.

https://codereview.chromium.org/184253003/diff/180001/git_rebase_update.py#ne...
git_rebase_update.py:185: return 1
See other comment; should print here.

https://codereview.chromium.org/184253003/diff/180001/git_rename_branch.py
File git_rename_branch.py (right):

https://codereview.chromium.org/184253003/diff/180001/git_rename_branch.py#ne...
git_rename_branch.py:5: 
still no docstring.

https://codereview.chromium.org/184253003/diff/180001/git_rename_branch.py#ne...
git_rename_branch.py:19: if current:
if current != 'HEAD':

remove lines 16,17.

https://codereview.chromium.org/184253003/diff/180001/git_rename_branch.py#ne...
git_rename_branch.py:44: sys.exit(main(sys.argv[1:]))
why? argparse parse_args() automatically grabs sys.argv[1:]. Why are you
handling it at all here? (This comment holds true for all of these python
files.)

https://codereview.chromium.org/184253003/diff/180001/git_reparent_branch.py
File git_reparent_branch.py (right):

https://codereview.chromium.org/184253003/diff/180001/git_reparent_branch.py#...
git_reparent_branch.py:5: 
still no docstring

https://codereview.chromium.org/184253003/diff/180001/git_reparent_branch.py#...
git_reparent_branch.py:55: return git_rebase_update.main(['--no_fetch'])
directly calling another script's main() method makes me a little sad. These
scripts should interface with each other in two ways: through git_common, and
through their respective command line interfaces.

[ghost stip (do not use), 2014-03-25 23:44:22]

some comments

https://chromiumcodereview.appspot.com/184253003/diff/180001/git_common.py
File git_common.py (left):

https://chromiumcodereview.appspot.com/184253003/diff/180001/git_common.py#ol...
git_common.py:204: NO_BRANCH = ('* (no branch)', '* (detached from ')
why remove the )?

https://chromiumcodereview.appspot.com/184253003/diff/180001/git_common.py
File git_common.py (right):

https://chromiumcodereview.appspot.com/184253003/diff/180001/git_common.py#ne...
git_common.py:207: """@Decorates |function| so that it only performs its action
once, no matter
nit: """One line.
"""

https://chromiumcodereview.appspot.com/184253003/diff/180001/git_common.py#ne...
git_common.py:213: return _inner_gen().next
can you add a comment how this works?

https://chromiumcodereview.appspot.com/184253003/diff/180001/git_common.py#ne...
git_common.py:310: """The same as run_with_stderr, except it only returns
stdout."""
'Calls run_with_stderr but only returns stdout.'

https://chromiumcodereview.appspot.com/184253003/diff/180001/git_common.py#ne...
git_common.py:442: RebaseRet = collections.namedtuple('RebaseRet', 'success
message')
wouldn't this be like DefaultRebaseRet? RebaseRetTemplate?

[ghost stip (do not use), 2014-03-25 23:44:24]

some comments

[iannucci, 2014-03-26 01:39:49]

done. I vote that any other changes go into another cl

https://codereview.chromium.org/184253003/diff/180001/docs/html/git-freeze.html
File docs/html/git-freeze.html (right):

https://codereview.chromium.org/184253003/diff/180001/docs/html/git-freeze.ht...
docs/html/git-freeze.html:844: Last updated 2014-03-22 12:59:00 PDT
On 2014/03/25 19:37:23, agable wrote:
> Is there any way these timestamps can be generated only if the rest of the
text
> actually differs from the previous version?

Probably. v2.

We'd also wan the version number on the man pages to somehow be useful. Right
now it's the short hash of the local parent commit, which is totally useless.

https://codereview.chromium.org/184253003/diff/180001/docs/src/git-new-branch...
File docs/src/git-new-branch.txt (right):

https://codereview.chromium.org/184253003/diff/180001/docs/src/git-new-branch...
docs/src/git-new-branch.txt:27: the default upstream (i.e. `git new-branch
<branch_name>`). If you have features
On 2014/03/25 19:37:23, agable wrote:
> link to new-branch doc

Er... I'm IN the new-branch doc :P

https://codereview.chromium.org/184253003/diff/180001/docs/src/git-new-branch...
docs/src/git-new-branch.txt:38: Set the tracking (upstream) branch to <REF>.
<REF> may be a local branch,
On 2014/03/25 19:37:23, agable wrote:
> 80 chars

It's tab expansion!!!!!! (and we need tabs here :/)

edit: it's our lucky day. Looks like space-indent also works!

https://codereview.chromium.org/184253003/diff/180001/docs/src/git-rebase-upd...
File docs/src/git-rebase-update.txt (right):

https://codereview.chromium.org/184253003/diff/180001/docs/src/git-rebase-upd...
docs/src/git-rebase-update.txt:34: All branches are rebased in topological order
from roots (upstreams) to leafs.
On 2014/03/25 19:37:23, agable wrote:
> leaves. We're not rebasing a hockey team.

Done.

https://codereview.chromium.org/184253003/diff/180001/docs/src/git-rebase-upd...
docs/src/git-rebase-update.txt:97: branch.<name>.base
On 2014/03/25 19:37:23, agable wrote:
> I think most of this documentation should be in mark-merge-base. I find it
> really annoying when most of the docs for a feature are in the docs for the
> primary customer of that feature, not the feature itself.

Yeah but mark-merge-base will very infrequently be used by humans... I would
like to get rid of it entirely, but I can't really do that :(

edit per offline: Going to keep it in rebase-update, because otherwise the docs
will all black-hole into mark-merge-base, which would be confusing for users of
the tool. The merge base pointer only exists for the purposes of
git-rebase-update.

https://codereview.chromium.org/184253003/diff/180001/docs/src/git-squash-bra...
File docs/src/git-squash-branch.txt (right):

https://codereview.chromium.org/184253003/diff/180001/docs/src/git-squash-bra...
docs/src/git-squash-branch.txt:12: 'git squash-branch' [<header message>]
On 2014/03/25 19:37:23, agable wrote:
> Preferably 'git squash-branch -m <header message>'.

Done.

https://codereview.chromium.org/184253003/diff/180001/docs/src/git-upstream-d...
File docs/src/git-upstream-diff.txt (right):

https://codereview.chromium.org/184253003/diff/180001/docs/src/git-upstream-d...
docs/src/git-upstream-diff.txt:45: Extra arguments are included in the
invocation of linkgit:git-diff[1]. These
On 2014/03/25 19:37:23, agable wrote:
> 80 chars

Done.

https://codereview.chromium.org/184253003/diff/180001/git_mark_merge_base.py
File git_mark_merge_base.py (right):

https://codereview.chromium.org/184253003/diff/180001/git_mark_merge_base.py#...
git_mark_merge_base.py:27: epilog=''.join(__doc__.strip().splitlines()[1:]))
On 2014/03/25 19:37:23, agable wrote:
> Do ' '.join(blah) and get rid of the leading spaces in the docstring.

Derpdone

https://codereview.chromium.org/184253003/diff/180001/git_new_branch.py
File git_new_branch.py (right):

https://codereview.chromium.org/184253003/diff/180001/git_new_branch.py#newco...
git_new_branch.py:24: help='upstream branch to track.')
On 2014/03/25 19:37:23, agable wrote:
> (or tag)

Done.

https://codereview.chromium.org/184253003/diff/180001/git_rebase_update.py
File git_rebase_update.py (right):

https://codereview.chromium.org/184253003/diff/180001/git_rebase_update.py#ne...
git_rebase_update.py:5: 
On 2014/03/25 19:37:23, agable wrote:
> module docstring

Done.

https://codereview.chromium.org/184253003/diff/180001/git_rebase_update.py#ne...
git_rebase_update.py:15: from git_freezer import freeze, thaw
On 2014/03/25 19:37:23, agable wrote:
> See similar comment on another file -- I think the only interface between
these
> scripts should be git_common and their command line interfaces. They shouldn't
> import each other.

Done.

https://codereview.chromium.org/184253003/diff/180001/git_rebase_update.py#ne...
git_rebase_update.py:22: """Finds the branch which we should return to after
rebase-update completes
On 2014/03/25 19:37:23, agable wrote:
> single line for first line of docstring

Done.

https://codereview.chromium.org/184253003/diff/180001/git_rebase_update.py#ne...
git_rebase_update.py:35: from git_common import in_rebase
On 2014/03/25 19:37:23, agable wrote:
> remove import

oops, done

https://codereview.chromium.org/184253003/diff/180001/git_rebase_update.py#ne...
git_rebase_update.py:37: if in_rebase():
On 2014/03/25 19:37:23, agable wrote:
> git.in_rebase

Done.

https://codereview.chromium.org/184253003/diff/180001/git_rebase_update.py#ne...
git_rebase_update.py:40: print (
On 2014/03/25 19:37:23, agable wrote:
> This print is out of place. This method should either just return True/False,
or
> it should print this message and exit entirely.

Function should not exist. Done.

https://codereview.chromium.org/184253003/diff/180001/git_rebase_update.py#ne...
git_rebase_update.py:50: 
On 2014/03/25 19:37:23, agable wrote:
> no newline

Done.

https://codereview.chromium.org/184253003/diff/180001/git_rebase_update.py#ne...
git_rebase_update.py:97: git.set_config('branch.%s.remote' % down, '.')
On 2014/03/25 19:37:23, agable wrote:
> You now have the set_branch_config methods

I knew I missed some!

https://codereview.chromium.org/184253003/diff/180001/git_rebase_update.py#ne...
git_rebase_update.py:149: print 'Squashing failed. You probably have a real
merge conflict.'
On 2014/03/25 19:37:23, agable wrote:
> Consider printing a textwrap.dedented triple-quoted string.

Done.

https://codereview.chromium.org/184253003/diff/180001/git_rebase_update.py#ne...
git_rebase_update.py:185: return 1
On 2014/03/25 19:37:23, agable wrote:
> See other comment; should print here.

Done.

https://codereview.chromium.org/184253003/diff/180001/git_rename_branch.py
File git_rename_branch.py (right):

https://codereview.chromium.org/184253003/diff/180001/git_rename_branch.py#ne...
git_rename_branch.py:5: 
On 2014/03/25 19:37:23, agable wrote:
> still no docstring.

Done.

https://codereview.chromium.org/184253003/diff/180001/git_rename_branch.py#ne...
git_rename_branch.py:19: if current:
On 2014/03/25 19:37:23, agable wrote:
> if current != 'HEAD':
> 
> remove lines 16,17.

Agree, but can't refactor :/

https://codereview.chromium.org/184253003/diff/180001/git_rename_branch.py#ne...
git_rename_branch.py:44: sys.exit(main(sys.argv[1:]))
On 2014/03/25 19:37:23, agable wrote:
> why? argparse parse_args() automatically grabs sys.argv[1:]. Why are you
> handling it at all here? (This comment holds true for all of these python
> files.)

b/c tests

https://codereview.chromium.org/184253003/diff/180001/git_reparent_branch.py
File git_reparent_branch.py (right):

https://codereview.chromium.org/184253003/diff/180001/git_reparent_branch.py#...
git_reparent_branch.py:5: 
On 2014/03/25 19:37:23, agable wrote:
> still no docstring

Done.

https://codereview.chromium.org/184253003/diff/180001/git_reparent_branch.py#...
git_reparent_branch.py:55: return git_rebase_update.main(['--no_fetch'])
On 2014/03/25 19:37:23, agable wrote:
> directly calling another script's main() method makes me a little sad. These
> scripts should interface with each other in two ways: through git_common, and
> through their respective command line interfaces.

done as much as I'm going to :p

[agable, 2014-03-26 02:13:43]

Sweeeeet. LGTM shipitshipit.

[iannucci, 2014-03-26 02:17:46]

The CQ bit was checked by iannucci@chromium.org

[commit-bot: I haz the power, 2014-03-26 02:17:52]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/iannucci@chromium.org/184253003/450002

[commit-bot: I haz the power, 2014-03-26 02:19:45]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2014-03-26 02:19:46]

Presubmit check for 184253003-450002 failed and returned exit status 1.

Running presubmit commit checks ...
Checking out rietveld...
Running save-description-on-failure.sh
Running push-basic.sh
Running upstream.sh
Running submit-from-new-dir.sh
Running abandon.sh
Running submodule-merge-test.sh
Running upload-local-tracking-branch.sh
Running hooks.sh
Running post-dcommit-hook-test.sh
Running upload-stale.sh
Running patch.sh
Running basic.sh

** Presubmit ERRORS **
Pylint (94 files) (23.45s) failed
Traceback (most recent call last):
  File "/b/depot_tools/third_party/pylint.py", line 31, in <module>
    lint.Run(args)
  File "/b/depot_tools/third_party/pylint/lint.py", line 879, in __init__
    linter.check(args)
  File "/b/depot_tools/third_party/pylint/lint.py", line 502, in check
    self.check_astng_module(astng, walker, rawcheckers)
  File "/b/depot_tools/third_party/pylint/lint.py", line 574, in
check_astng_module
    walker.walk(astng)
  File "/b/depot_tools/third_party/pylint/utils.py", line 528, in walk
    self.walk(child)
  File "/b/depot_tools/third_party/pylint/utils.py", line 528, in walk
    self.walk(child)
  File "/b/depot_tools/third_party/pylint/utils.py", line 528, in walk
    self.walk(child)
  File "/b/depot_tools/third_party/pylint/utils.py", line 525, in walk
    cb(astng)
  File "/b/depot_tools/third_party/pylint/checkers/base.py", line 161, in
visit_function
    if not redefined_by_decorator(node):
  File "/b/depot_tools/third_party/pylint/checkers/base.py", line 116, in
redefined_by_decorator
    decorator.expr.name == node.name):
AttributeError: 'Getattr' object has no attribute 'name'


tests/git_rebase_update_test.py (13.11s) failed
.FF
======================================================================
FAIL: testRebaseUpdate (__main__.GitRebaseUpdateTest)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "tests/git_rebase_update_test.py", line 258, in testRebaseUpdate
    '* A'
AssertionError: Lists differ: ['* FREEZE.unindexed', '* sub_... != ['*
old_file', '| * sub_K', '|...

First differing element 0:
* FREEZE.unindexed
* old_file

First list contains 2 additional elements.
First extra element 21:
|/

Diff is 792 characters long. Set self.maxDiff to None to see it.

======================================================================
FAIL: testTrackTag (__main__.GitRebaseUpdateTest)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "tests/git_rebase_update_test.py", line 414, in testTrackTag
    '* A'
AssertionError: Lists differ: ['* foobar2', '* foobar1', '| ... != ['* O', '*
N', '| * foobar2', ...

First differing element 0:
* foobar2
* O

Second list contains 3 additional elements.
First extra element 16:
* C

+ ['* O',
+  '* N',
- ['* foobar2',
? ^

+  '| * foobar2',
? ^ ++

-  '* foobar1',
+  '| * foobar1',
?   ++

+  '|/',
   '| * L',
   '| * K',
   '| * J',
   '| * I',
   '| * H',
   '|/',
   '* M',
   '* G',
   '* F',
   '* E',
   '* D',
   '* C',
   '* B',
   '* A']

----------------------------------------------------------------------
Ran 3 tests in 12.765s

FAILED (failures=2)
Name                  Stmts   Miss  Cover   Missing
---------------------------------------------------
git_new_branch           26      0   100%   
git_rebase_update       130      0   100%   
git_rename_branch        28     21    25%   17-46
git_reparent_branch      42      9    79%   35, 37, 51, 55-57, 60-63
---------------------------------------------------
TOTAL                   226     30    87%   
FATAL: not at 100% coverage.


Presubmit checks took 94.0s to calculate.

[iannucci, 2014-03-26 04:38:38]

The CQ bit was checked by iannucci@chromium.org

[commit-bot: I haz the power, 2014-03-26 04:38:47]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/iannucci@chromium.org/184253003/230005

[iannucci, 2014-03-26 04:39:18]

The CQ bit was unchecked by iannucci@chromium.org

[commit-bot: I haz the power, 2014-03-26 04:40:50]

The CQ bit was unchecked by commit-bot@chromium.org

[iannucci, 2014-03-26 04:40:50]

The CQ bit was checked by iannucci@chromium.org

[commit-bot: I haz the power, 2014-03-26 04:40:50]

Presubmit check for 184253003-230005 failed and returned exit status 1.

Running presubmit commit checks ...
Checking out rietveld...
Running save-description-on-failure.sh
Running push-basic.sh
Running upstream.sh
Running submit-from-new-dir.sh
Running abandon.sh
Running submodule-merge-test.sh
Running upload-local-tracking-branch.sh
Running hooks.sh
Running post-dcommit-hook-test.sh
Running upload-stale.sh
Running patch.sh
Running basic.sh

** Presubmit ERRORS **
Pylint (94 files) (21.84s) failed
Traceback (most recent call last):
  File "/b/depot_tools/third_party/pylint.py", line 31, in <module>
    lint.Run(args)
  File "/b/depot_tools/third_party/pylint/lint.py", line 879, in __init__
    linter.check(args)
  File "/b/depot_tools/third_party/pylint/lint.py", line 502, in check
    self.check_astng_module(astng, walker, rawcheckers)
  File "/b/depot_tools/third_party/pylint/lint.py", line 574, in
check_astng_module
    walker.walk(astng)
  File "/b/depot_tools/third_party/pylint/utils.py", line 528, in walk
    self.walk(child)
  File "/b/depot_tools/third_party/pylint/utils.py", line 528, in walk
    self.walk(child)
  File "/b/depot_tools/third_party/pylint/utils.py", line 528, in walk
    self.walk(child)
  File "/b/depot_tools/third_party/pylint/utils.py", line 525, in walk
    cb(astng)
  File "/b/depot_tools/third_party/pylint/checkers/base.py", line 161, in
visit_function
    if not redefined_by_decorator(node):
  File "/b/depot_tools/third_party/pylint/checkers/base.py", line 116, in
redefined_by_decorator
    decorator.expr.name == node.name):
AttributeError: 'Getattr' object has no attribute 'name'


tests/git_rebase_update_test.py (16.55s) failed
.FF
======================================================================
FAIL: testRebaseUpdate (__main__.GitRebaseUpdateTest)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "tests/git_rebase_update_test.py", line 178, in testRebaseUpdate
    """)
  File
"/b/commit-queue/workdir/tools/depot_tools/testing_support/git_test_utils.py",
line 505, in assertSchema
    self.repo.to_schema().simple_graph())
AssertionError: {'A': set([]), 'old_file': set(['A']), 'C': set(['B']), 'B':
set(['A']), 'E': se [truncated]... != {'A': set([]), 'old_file': set(['A']),
'C': set(['B']), 'B': set(['A']), 'E': se [truncated]...
  {'A': set(),
   'B': set(['A']),
   'C': set(['B']),
   'D': set(['C']),
   'E': set(['D']),
   'F': set(['E']),
+  'FREEZE.unindexed': set(['sub_K']),
   'G': set(['F']),
   'H': set(['foobar2']),
   'I': set(['H']),
   'J': set(['I']),
   'K': set(['J']),
   'L': set(['K']),
   'M': set(['G']),
   'N': set(['M']),
   'O': set(['N']),
   'foobar1': set(['O']),
   'foobar2': set(['foobar1']),
   'old_file': set(['A']),
   'sub_K': set(['K'])}

======================================================================
FAIL: testTrackTag (__main__.GitRebaseUpdateTest)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "tests/git_rebase_update_test.py", line 296, in testTrackTag
    """)
  File
"/b/commit-queue/workdir/tools/depot_tools/testing_support/git_test_utils.py",
line 505, in assertSchema
    self.repo.to_schema().simple_graph())
AssertionError: {'A': set([]), 'C': set(['B']), 'B': set(['A']), 'E':
set(['D']), 'D': set(['C'] [truncated]... != {'A': set([]), 'C': set(['B']),
'B': set(['A']), 'E': set(['D']), 'D': set(['C'] [truncated]...
  {'A': set(),
   'B': set(['A']),
   'C': set(['B']),
   'D': set(['C']),
   'E': set(['D']),
   'F': set(['E']),
   'G': set(['F']),
   'H': set(['M']),
   'I': set(['H']),
   'J': set(['I']),
   'K': set(['J']),
   'L': set(['K']),
   'M': set(['G']),
-  'N': set(['M']),
-  'O': set(['N']),
   'foobar1': set(['M']),
   'foobar2': set(['foobar1'])}

----------------------------------------------------------------------
Ran 3 tests in 16.331s

FAILED (failures=2)
Name                  Stmts   Miss  Cover   Missing
---------------------------------------------------
git_new_branch           26      0   100%   
git_rebase_update       130      0   100%   
git_rename_branch        28     21    25%   17-46
git_reparent_branch      42      9    79%   35, 37, 51, 55-57, 60-63
---------------------------------------------------
TOTAL                   226     30    87%   
FATAL: not at 100% coverage.


Presubmit checks took 104.7s to calculate.

[commit-bot: I haz the power, 2014-03-26 04:40:53]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/iannucci@chromium.org/184253003/230005

[commit-bot: I haz the power, 2014-03-26 04:44:02]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2014-03-26 04:44:03]

Presubmit check for 184253003-230005 failed and returned exit status 1.

Running presubmit commit checks ...
Checking out rietveld...
Running save-description-on-failure.sh
Running push-basic.sh
Running upstream.sh
Running submit-from-new-dir.sh
Running abandon.sh
Running submodule-merge-test.sh
Running upload-local-tracking-branch.sh
Running hooks.sh
Running post-dcommit-hook-test.sh
Running upload-stale.sh
Running patch.sh
Running basic.sh

** Presubmit ERRORS **
Pylint (94 files) (19.87s) failed
Traceback (most recent call last):
  File "/b/depot_tools/third_party/pylint.py", line 31, in <module>
    lint.Run(args)
  File "/b/depot_tools/third_party/pylint/lint.py", line 879, in __init__
    linter.check(args)
  File "/b/depot_tools/third_party/pylint/lint.py", line 502, in check
    self.check_astng_module(astng, walker, rawcheckers)
  File "/b/depot_tools/third_party/pylint/lint.py", line 574, in
check_astng_module
    walker.walk(astng)
  File "/b/depot_tools/third_party/pylint/utils.py", line 528, in walk
    self.walk(child)
  File "/b/depot_tools/third_party/pylint/utils.py", line 528, in walk
    self.walk(child)
  File "/b/depot_tools/third_party/pylint/utils.py", line 528, in walk
    self.walk(child)
  File "/b/depot_tools/third_party/pylint/utils.py", line 525, in walk
    cb(astng)
  File "/b/depot_tools/third_party/pylint/checkers/base.py", line 161, in
visit_function
    if not redefined_by_decorator(node):
  File "/b/depot_tools/third_party/pylint/checkers/base.py", line 116, in
redefined_by_decorator
    decorator.expr.name == node.name):
AttributeError: 'Getattr' object has no attribute 'name'


tests/git_rebase_update_test.py (10.17s) failed
.FF
======================================================================
FAIL: testRebaseUpdate (__main__.GitRebaseUpdateTest)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "tests/git_rebase_update_test.py", line 178, in testRebaseUpdate
    """)
  File
"/b/commit-queue/workdir/tools/depot_tools/testing_support/git_test_utils.py",
line 505, in assertSchema
    self.repo.to_schema().simple_graph())
AssertionError: {'A': set([]), 'old_file': set(['A']), 'C': set(['B']), 'B':
set(['A']), 'E': se [truncated]... != {'A': set([]), 'old_file': set(['A']),
'C': set(['B']), 'B': set(['A']), 'E': se [truncated]...
  {'A': set(),
   'B': set(['A']),
   'C': set(['B']),
   'D': set(['C']),
   'E': set(['D']),
   'F': set(['E']),
+  'FREEZE.unindexed': set(['sub_K']),
   'G': set(['F']),
   'H': set(['foobar2']),
   'I': set(['H']),
   'J': set(['I']),
   'K': set(['J']),
   'L': set(['K']),
   'M': set(['G']),
   'N': set(['M']),
   'O': set(['N']),
   'foobar1': set(['O']),
   'foobar2': set(['foobar1']),
   'old_file': set(['A']),
   'sub_K': set(['K'])}

======================================================================
FAIL: testTrackTag (__main__.GitRebaseUpdateTest)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "tests/git_rebase_update_test.py", line 296, in testTrackTag
    """)
  File
"/b/commit-queue/workdir/tools/depot_tools/testing_support/git_test_utils.py",
line 505, in assertSchema
    self.repo.to_schema().simple_graph())
AssertionError: {'A': set([]), 'C': set(['B']), 'B': set(['A']), 'E':
set(['D']), 'D': set(['C'] [truncated]... != {'A': set([]), 'C': set(['B']),
'B': set(['A']), 'E': set(['D']), 'D': set(['C'] [truncated]...
  {'A': set(),
   'B': set(['A']),
   'C': set(['B']),
   'D': set(['C']),
   'E': set(['D']),
   'F': set(['E']),
   'G': set(['F']),
   'H': set(['M']),
   'I': set(['H']),
   'J': set(['I']),
   'K': set(['J']),
   'L': set(['K']),
   'M': set(['G']),
-  'N': set(['M']),
-  'O': set(['N']),
   'foobar1': set(['M']),
   'foobar2': set(['foobar1'])}

----------------------------------------------------------------------
Ran 3 tests in 9.975s

FAILED (failures=2)
Name                  Stmts   Miss  Cover   Missing
---------------------------------------------------
git_new_branch           26      0   100%   
git_rebase_update       130      0   100%   
git_rename_branch        28     21    25%   17-46
git_reparent_branch      42      9    79%   35, 37, 51, 55-57, 60-63
---------------------------------------------------
TOTAL                   226     30    87%   
FATAL: not at 100% coverage.


Presubmit checks took 108.5s to calculate.

[iannucci, 2014-03-26 05:51:44]

The CQ bit was checked by iannucci@chromium.org

[commit-bot: I haz the power, 2014-03-26 05:51:49]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/iannucci@chromium.org/184253003/940001

[commit-bot: I haz the power, 2014-03-26 05:53:59]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2014-03-26 05:53:59]

Presubmit check for 184253003-940001 failed and returned exit status 1.

Running presubmit commit checks ...
Checking out rietveld...
Running save-description-on-failure.sh
Running push-basic.sh
Running upstream.sh
Running submit-from-new-dir.sh
Running abandon.sh
Running submodule-merge-test.sh
Running upload-local-tracking-branch.sh
Running hooks.sh
Running post-dcommit-hook-test.sh
Running upload-stale.sh
Running patch.sh
Running basic.sh

** Presubmit ERRORS **
Pylint (94 files) (19.06s) failed
Traceback (most recent call last):
  File "/b/depot_tools/third_party/pylint.py", line 31, in <module>
    lint.Run(args)
  File "/b/depot_tools/third_party/pylint/lint.py", line 879, in __init__
    linter.check(args)
  File "/b/depot_tools/third_party/pylint/lint.py", line 502, in check
    self.check_astng_module(astng, walker, rawcheckers)
  File "/b/depot_tools/third_party/pylint/lint.py", line 574, in
check_astng_module
    walker.walk(astng)
  File "/b/depot_tools/third_party/pylint/utils.py", line 528, in walk
    self.walk(child)
  File "/b/depot_tools/third_party/pylint/utils.py", line 528, in walk
    self.walk(child)
  File "/b/depot_tools/third_party/pylint/utils.py", line 528, in walk
    self.walk(child)
  File "/b/depot_tools/third_party/pylint/utils.py", line 525, in walk
    cb(astng)
  File "/b/depot_tools/third_party/pylint/checkers/base.py", line 161, in
visit_function
    if not redefined_by_decorator(node):
  File "/b/depot_tools/third_party/pylint/checkers/base.py", line 116, in
redefined_by_decorator
    decorator.expr.name == node.name):
AttributeError: 'Getattr' object has no attribute 'name'


Presubmit checks took 110.0s to calculate.

[iannucci, 2014-03-26 06:15:47]

The CQ bit was checked by iannucci@chromium.org

[commit-bot: I haz the power, 2014-03-26 06:15:53]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/iannucci@chromium.org/184253003/990001

[commit-bot: I haz the power, 2014-03-26 06:18:51]

Change committed as 259520