Re-land fixes format always saying the folder was dirty.
The --full option would skip setting the return value of the format
command when used with --dry-run. This in turn would cause the
presubmit check to always pass when --full was enabled in a repo by
default.
Was discovered in the ANGLE repository.
Bug: angleproject:4241
Change-Id: I986646336d1857af2826863ef4becbf2b9ee3049
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/tools/depot_tools/+/1993910
Commit-Queue: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Edward Lesmes <ehmaldonado@chromium.org>
This reverts commit fc132e61db.
Reason for revert: Causes all presubmit checks to return dirty
format because of a non-zero diff output.
Original change's description:
> git-cl: Fix format --dry-run not working with --full.
>
> The --full option would skip setting the return value of the format
> command when used with --dry-run. This in turn would cause the
> presubmit check to always pass when --full was enabled in a repo by
> default.
>
> Was discovered in the ANGLE repository.
>
> Bug: angleproject:4241
> Change-Id: Ie6cb423a6818c1e26781d77938a0dd22c02b4c16
> Reviewed-on: https://chromium-review.googlesource.com/c/chromium/tools/depot_tools/+/1987835
> Commit-Queue: Jamie Madill <jmadill@chromium.org>
> Reviewed-by: Edward Lesmes <ehmaldonado@chromium.org>
> Auto-Submit: Jamie Madill <jmadill@chromium.org>
TBR=ehmaldonado@chromium.org,jmadill@chromium.org
Change-Id: I0a4c51117ab20606f6dbb1f7a3ba40f87da1e939
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: angleproject:4241
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/tools/depot_tools/+/1993907
Reviewed-by: Jamie Madill <jmadill@chromium.org>
Commit-Queue: Jamie Madill <jmadill@chromium.org>
The --full option would skip setting the return value of the format
command when used with --dry-run. This in turn would cause the
presubmit check to always pass when --full was enabled in a repo by
default.
Was discovered in the ANGLE repository.
Bug: angleproject:4241
Change-Id: Ie6cb423a6818c1e26781d77938a0dd22c02b4c16
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/tools/depot_tools/+/1987835
Commit-Queue: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Edward Lesmes <ehmaldonado@chromium.org>
Auto-Submit: Jamie Madill <jmadill@chromium.org>
This matches the recent change Chromium's Python style guide.
This also changes the .style.yapf in depot_tools to be explicit
about the style used in this repo.
Bug: 846432
Change-Id: I4947a90aec1739a9a86ffc9bfc5eacc1182dc186
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/tools/depot_tools/+/1987006
Reviewed-by: Robbie Iannucci <iannucci@chromium.org>
Commit-Queue: Andrew Grieve <agrieve@chromium.org>
In repos that are checked out separately from chromium, invoking
"git cl format" on a change that has files that would be formatted using
clang-format causes an error because the depot_tools version of
clang-format looks for a clang-format inside the repo.
This prevents using proto-based properties (the preferred mechanism
going forward) in recipe repos that enforce formatting. The
--no-clang-format flag provides a workaround that can be used to
prevent trying to format any files where clang-format doesn't work.
Bug: 979330
Change-Id: Ice8561d88b29623deb953465253f92c88aa7fc2e
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/tools/depot_tools/+/1986111
Commit-Queue: Garrett Beaty <gbeaty@chromium.org>
Reviewed-by: Edward Lesmes <ehmaldonado@chromium.org>
The current behavior of the --no-python flag doesn't disable formatting
for python files. Instead, it has the same effect as the --python flag.
Change-Id: I26024d3030b9c873180b488d00351676c37c9492
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/tools/depot_tools/+/1985785
Commit-Queue: Garrett Beaty <gbeaty@chromium.org>
Reviewed-by: Edward Lesmes <ehmaldonado@chromium.org>
Adds error handling logic for pre-existing tags (which can occur
if "archive" is CTRL-C aborted midway through) and for undeletable
branches (which can happen if they are currently checked out in a
working dir).
Change-Id: I27b6da9f5860c307f49cbeabb1b0ccf9cfb28eb6
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/tools/depot_tools/+/1930023
Commit-Queue: Edward Lesmes <ehmaldonado@chromium.org>
Reviewed-by: Edward Lesmes <ehmaldonado@chromium.org>
Auto-Submit: Kevin Marshall <kmarshall@chromium.org>
Switches the yapfignore implementation in git_cl to use fnmatch instead
of glob. This ends up bringing the .yapfignore file parsing in line with
yapf's actual implementation, namely allowing uses such as '*pb2.py'
instead of having to list out each directory manually.
Bug: 1027953
Change-Id: Ibb1cb4252c546de6f1b1af720c4c29ffd0f7be0c
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/tools/depot_tools/+/1938026
Commit-Queue: Brian Sheedy <bsheedy@chromium.org>
Reviewed-by: Edward Lesmes <ehmaldonado@chromium.org>
Reviewed-by: Quinten Yearsley <qyearsley@chromium.org>
It was previously stored in each repository working dir, making it
neccessary to modify .gitignore files in many places. Save it in
depot_tools working directory, where there's already an entry in its
.gitignore file and no additional work is needed to hide this file.
Bug: None
Change-Id: I58102180e37358a206c8ecd1044aa772a684e066
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/tools/depot_tools/+/1944167
Commit-Queue: Edward Lesmes <ehmaldonado@chromium.org>
Reviewed-by: Edward Lesmes <ehmaldonado@chromium.org>
Avoid cluttering the home directory by saving the CL description in
the project's working directory.
Bug: None
Change-Id: I42c78d03cd239e7c0453867f21fe4be57086722e
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/tools/depot_tools/+/1925627
Commit-Queue: Edward Lesmes <ehmaldonado@chromium.org>
Reviewed-by: Edward Lesmes <ehmaldonado@chromium.org>
This reverts commit c87ed60623.
Reason for revert:
Causes a regressions for CL descriptions on Windows
Original change's description:
> git-cl: Fix some python3 compatibility errors.
>
> Also, fix bug in git cl status where the 'updated' field was used to compare messages, even though
> it doesn't exist (see [1]). This CL modifies it to use 'date', which does exist.
>
> [1] https://gerrit-review.googlesource.com/Documentation/rest-api-changes.html#change-message-info
>
> Bug: 1002209
> Change-Id: I5a5e1193b8502c3ad35d94808ea178cad7f44ac6
> Reviewed-on: https://chromium-review.googlesource.com/c/chromium/tools/depot_tools/+/1891259
> Commit-Queue: Edward Lesmes <ehmaldonado@chromium.org>
> Reviewed-by: Anthony Polito <apolito@google.com>
TBR=ehmaldonado@chromium.org,apolito@google.com
# Not skipping CQ checks because original CL landed > 1 day ago.
Bug: 1002209
Change-Id: I004f202b12c6b99cb6b24cb12a14fba7898569e2
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/tools/depot_tools/+/1898547
Reviewed-by: Edward Lesmes <ehmaldonado@chromium.org>
Commit-Queue: Edward Lesmes <ehmaldonado@chromium.org>
The problems on crbug.com/1017367 should have been fixed by crrev.com/c/1880013
Bug: 1009814
Change-Id: Ia98304ca983b4e7e293fedb9df1fee589de58ba9
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/tools/depot_tools/+/1887714
Reviewed-by: Anthony Polito <apolito@google.com>
Commit-Queue: Edward Lesmes <ehmaldonado@chromium.org>
Lets the client repository make 'git cl format' use the
'--full' option by default. This solves issues when using less
common clang-format options that don't behave well with small
diffs. For example, AlignConsecutiveAssignments.
This is a replacement for the 'diff-lines-of-context' setting.
Bug: angleproject:4003
Change-Id: I81dc3b4992a7420e7235da88ec78e51ec4c0d24f
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/tools/depot_tools/+/1879148
Reviewed-by: Edward Lesmes <ehmaldonado@chromium.org>
Commit-Queue: Jamie Madill <jmadill@chromium.org>
This reverts commit 812ac228e3.
Reason for revert: The fix was not sufficient to solve the git cl
format difference between 'git cl format' and 'git cl format --full'.
Looks like we'll need a different solution.
Original change's description:
> git-cl: Add diff lines of context setting.
>
> Clients using some Clang-Format options can run into bugs when using
> 0 lines of context. AlignConsecutiveAssignments in particular is picky
> and behaves differently with 0 lines of context.
>
> This behaviour lead to 'git cl format' behaving differently from
> 'git cl format --full'. That could also break presubmit format checks.
> Using >0 lines of context for 'git diff' fixes the inconsistency.
>
> Add a setting 'DIFF_LINES_OF_CONTEXT' that can be controlled via
> codereview.settings. Defaults to zero to preserve the old behaviour.
> The setting allows the client to control the number of lines of context
> to use when running 'git cl format'.
>
> Bug: angleproject:4003
> Change-Id: Ied2ebf23df4c41ba19bfbd5b8ddf526b56a20b31
> Reviewed-on: https://chromium-review.googlesource.com/c/chromium/tools/depot_tools/+/1864309
> Reviewed-by: Edward Lesmes <ehmaldonado@chromium.org>
> Commit-Queue: Jamie Madill <jmadill@chromium.org>
TBR=agable@chromium.org,ehmaldonado@chromium.org,jmadill@chromium.org
# Not skipping CQ checks because original CL landed > 1 day ago.
Bug: angleproject:4003
Change-Id: Ib6b70969509a1e5446922d488f11d13a0c59c536
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/tools/depot_tools/+/1877003
Reviewed-by: Aaron Gable <agable@chromium.org>
Commit-Queue: Jamie Madill <jmadill@chromium.org>
Clients using some Clang-Format options can run into bugs when using
0 lines of context. AlignConsecutiveAssignments in particular is picky
and behaves differently with 0 lines of context.
This behaviour lead to 'git cl format' behaving differently from
'git cl format --full'. That could also break presubmit format checks.
Using >0 lines of context for 'git diff' fixes the inconsistency.
Add a setting 'DIFF_LINES_OF_CONTEXT' that can be controlled via
codereview.settings. Defaults to zero to preserve the old behaviour.
The setting allows the client to control the number of lines of context
to use when running 'git cl format'.
Bug: angleproject:4003
Change-Id: Ied2ebf23df4c41ba19bfbd5b8ddf526b56a20b31
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/tools/depot_tools/+/1864309
Reviewed-by: Edward Lesmes <ehmaldonado@chromium.org>
Commit-Queue: Jamie Madill <jmadill@chromium.org>
Adds support for .yapfignore files to "git cl format" when formatting
Python files. yapf is supposed to parse the .yapfignore file in the
current working directory, but this appears to not work when files
are explicitly passed to yapf for formatting like they are in git cl
format. Instead, parse the .yapfignore file ourselves and skip over any
matching files.
Bug: angleproject:3985
Change-Id: I5e8469470fb8ddbaa914005b012ac1f39dfdd223
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/tools/depot_tools/+/1849807
Reviewed-by: Aaron Gable <agable@chromium.org>
Commit-Queue: Brian Sheedy <bsheedy@chromium.org>
Ran:
vi $(git grep --name-only iteritems | grep -v third_party)
vi $(git grep --name-only itervalues | grep -v third_party)
vi $(git grep --name-only 'print ' | grep -v third_party)
and edited the files quickly with adhoc macros. Then ran in recipes/:
./recipes.py test train
There was only a small subset of files that had been updated to use
six.iteritems() and six.itervalues(). Since the dataset size that is
being used in gclient is small (pretty much always below 200 items),
it's better to just switch to .items() right away and take the temporary
performance hit, so that we don't need to come back to rewrite the code.
Recipe-Nontrivial-Roll: build
Bug: 984182
Change-Id: I5faf11486b66b0d73c9098ab0f2ce1b15a45c53e
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/tools/depot_tools/+/1854900
Commit-Queue: Edward Lesmes <ehmaldonado@chromium.org>
Reviewed-by: Edward Lesmes <ehmaldonado@chromium.org>
Auto-Submit: Marc-Antoine Ruel <maruel@chromium.org>
This option take the same format as -b / --bug but
spits out "Fixed: #,#" instead of "Bug: #,#".
This CL also looks for fix-, fixed-, and fixes-
branch prefixes and uses those as cue to add "Fixed:"
to the CL description when editing. This is a slight
behavioral change for fix- specifically, as it used
to produce "Bug: " instead.
Bug: monorail:4470
Change-Id: Ib24a1ff33ca3674e53fc5437f459ea5708988290
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/tools/depot_tools/+/1846785
Commit-Queue: Dan Beam <dbeam@chromium.org>
Reviewed-by: Edward Lesmes <ehmaldonado@chromium.org>
Auto-Submit: Dan Beam <dbeam@chromium.org>
Tag is the same as used by Gerrit plugin for the same functionality:
retry_failed:1
R=ehmaldonado
Change-Id: I14ef435b676088c86ae7cc0536c6a68fa58bd89f
Bug: 445502
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/tools/depot_tools/+/1848258
Commit-Queue: Andrii Shyshkalov <tandrii@google.com>
Auto-Submit: Andrii Shyshkalov <tandrii@google.com>
Reviewed-by: Edward Lesmes <ehmaldonado@chromium.org>
This reverts commit f0faf48ddb.
Reason for revert:
Legacy properties are no longer used by Skia.
Original change's description:
> git-cl: Set legacy properties.
>
> Will be removed back again on Oct 9, 2019.
>
> Bug: 1004447
> Change-Id: I7003fac9ae8b184e2a7073b516535e8fd599f4f5
> Reviewed-on: https://chromium-review.googlesource.com/c/chromium/tools/depot_tools/+/1825446
> Commit-Queue: Edward Lesmes <ehmaldonado@chromium.org>
> Reviewed-by: Andrii Shyshkalov <tandrii@google.com>
TBR=tandrii@google.com,ehmaldonado@chromium.org
# Not skipping CQ checks because original CL landed > 1 day ago.
Bug: 1004447
Change-Id: I4e1fb31b6179bf10c3d9554cd313f7040aaf1b1f
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/tools/depot_tools/+/1847959
Reviewed-by: Edward Lesmes <ehmaldonado@chromium.org>
Reviewed-by: Andrii Shyshkalov <tandrii@google.com>
Commit-Queue: Edward Lesmes <ehmaldonado@chromium.org>
Commit-Queue: Andrii Shyshkalov <tandrii@google.com>
Auto-Submit: Edward Lesmes <ehmaldonado@chromium.org>
Make it work the same way as "git cl try --retry-failed".
R=ehmaldonado
Change-Id: Id3a709dc40b6b6d068acbba0c610da254ced8ff3
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/tools/depot_tools/+/1845478
Commit-Queue: Andrii Shyshkalov <tandrii@google.com>
Auto-Submit: Andrii Shyshkalov <tandrii@google.com>
Reviewed-by: Vadim Shtayura <vadimsh@chromium.org>
Get rid of _process_codereview_select_options and detected_codereview_from_url and simplify issue parsing.
Change-Id: I4200fd83ee868587c8627d6771c64f886b34a88b
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/tools/depot_tools/+/1838384
Reviewed-by: Anthony Polito <apolito@google.com>
Commit-Queue: Edward Lesmes <ehmaldonado@chromium.org>
Auto-Submit: Edward Lesmes <ehmaldonado@chromium.org>
This will prevent moving reviewers to cc when uploading patches.
Bug: 1010499
Change-Id: Ida16a4686f46350ba009bdaec1542895208c3f2d
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/tools/depot_tools/+/1835016
Commit-Queue: Edward Lesmes <ehmaldonado@chromium.org>
Reviewed-by: Anthony Polito <apolito@google.com>
The expected behavior is that for CLs that were already uploaded
before, where some tryjobs were run and failed, git cl upload
--retry-failed will be kind of like git cl upload --cq-dry-run
except it will only trigger tryjobs that failed.
Bug: 985887
Change-Id: I6371bca3ba501b1ea2cd7160e2f933530d7e633f
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/tools/depot_tools/+/1828322
Auto-Submit: Quinten Yearsley <qyearsley@chromium.org>
Commit-Queue: Andrii Shyshkalov <tandrii@google.com>
Reviewed-by: Edward Lesmes <ehmaldonado@chromium.org>
There are many functions in git_cl that appear unused, although
it's possible that some of them are used in other places, where
git_cl may be imported as a library.
This CL just removes some of the private ones, which at least
were not supposed to be used from other places.
Things removed here:
- _get_commiter_timestamp
- _git_amend_head
- GitCookiesChecker._get_usernames_of_domain
- constants MERGE_BRANCH, CHERRY_PICK_BRANCH;
the comment says that these are supposed to
be used in CMDland, but apparently they aren't.
contextlib which is only used by this, is also
removed
- CodereviewSettingsFileMock in git_cl_test;
it is not used anywhere in the test, and I
assume that nobody's importing and using
git_cl_test...
git_cl_test still passes, git cl seems to still work
Change-Id: Ib8eee0dfb181c8cda04c0894d2daea52b3ad42ac
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/tools/depot_tools/+/1818621
Reviewed-by: Andrii Shyshkalov <tandrii@google.com>
Commit-Queue: Quinten Yearsley <qyearsley@chromium.org>
With certain git remote problems an error message is emitted that
has the place of the url and the place of the branch wrong.
Example:
Remote "opera" for branch "git@github.com:operasoftware/gn-opera.git"
points to "bratell/no_jumbo_for_s", but it doesn't exist.
Should be:
Remote "opera" for branch "bratell/no_jumbo_for_s" points to
"git@github.com:operasoftware/gn-opera.git", but it doesn't exist.
R=dpranke@chromium.org
Change-Id: I484ac8af226903a49df1f62e837d01c345fb5f02
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/tools/depot_tools/+/1806595
Commit-Queue: Daniel Bratell <bratell@opera.com>
Reviewed-by: Dirk Pranke <dpranke@chromium.org>
Jamie Madill
Andrew Grieve
Garrett Beaty
Kevin Marshall
Anthony Polito
Brian Sheedy
Henrique Ferreiro
Edward Lemur
Andrii Shyshkalov
Marc-Antoine Ruel
Dan Beam
Andrii Shyshkalov
Quinten Yearsley
Daniel Bratell