Rietveld doesn't post the message when it is too large.
R=petermayo@chromium.org
BUG=
TEST=Now the CQ should stop having weird case where it doesn't post a message on
CQ failure. It was because the message was simply too long!
Review URL: http://codereview.chromium.org/10383004
git-svn-id: svn://svn.chromium.org/chrome/trunk/tools/depot_tools@135326 0039d316-1c4b-4281-b951-d872f2087c98
Rietveld self-corrupts its status all the time, there is not point in bailing
just because of that. So just assume it's an 'M' and it'll work fine. In any
case, the diff is properly parsed to detect what was the real operation.
Add unit test that correctly process a deleted file with status:null.
TBR=nsylvain@chromium.org
BUG=
TEST=A corrupted patchset can still be committed with the Commit Queue.
Review URL: http://codereview.chromium.org/10272024
git-svn-id: svn://svn.chromium.org/chrome/trunk/tools/depot_tools@134733 0039d316-1c4b-4281-b951-d872f2087c98
r123775 contained copy-paste left-over that caused the code to throw a NameError
instead of a HTTPError.
R=petermayo@chromium.org
BUG=
TEST=
Review URL: http://codereview.chromium.org/9536014
git-svn-id: svn://svn.chromium.org/chrome/trunk/tools/depot_tools@124184 0039d316-1c4b-4281-b951-d872f2087c98
Occasionally, AppEngine throws HTTP 500. In practice it's not a big deal but
upload.py converts them into sys.exit() call which is problematic. This breaks
the commit queue since this behaviors forces the CQ to quit.
R=dpranke@chromium.org
BUG=
TEST=
Review URL: http://codereview.chromium.org/9480005
git-svn-id: svn://svn.chromium.org/chrome/trunk/tools/depot_tools@123775 0039d316-1c4b-4281-b951-d872f2087c98
It's annoying to have it crash much later in a member function call when a
trailing backslash is provided.
R=dpranke@chromium.org
BUG=
TEST=
Review URL: http://codereview.chromium.org/8056016
git-svn-id: svn://svn.chromium.org/chrome/trunk/tools/depot_tools@103058 0039d316-1c4b-4281-b951-d872f2087c98
Add support for a few svn:* properties.
Silently ignore svn:mergeinfo. It's useless once we're switched to git-land anyway.
More testing.
R=dpranke@chromium.org
BUG=
TEST=
Review URL: http://codereview.chromium.org/7840003
git-svn-id: svn://svn.chromium.org/chrome/trunk/tools/depot_tools@99952 0039d316-1c4b-4281-b951-d872f2087c98
That means some properties will being to be applied.
R=dpranke@chromium.org
BUG=
TEST=
Review URL: http://codereview.chromium.org/7776015
git-svn-id: svn://svn.chromium.org/chrome/trunk/tools/depot_tools@99182 0039d316-1c4b-4281-b951-d872f2087c98
Use p.is_new signal and set it manually with rietveld patches.
Add regression test for +x.
R=dpranke@chromium.org
BUG=
TEST=
Review URL: http://codereview.chromium.org/7054057
git-svn-id: svn://svn.chromium.org/chrome/trunk/tools/depot_tools@87858 0039d316-1c4b-4281-b951-d872f2087c98
presubmit_support.py can now creates a Rietveld object on its own.
This is necessary since the object needs to be recreated out of process. The
commit queue runs the presubmit check out of process for sanity reasons.
Added TODO to push the cookie instead of the password.
R=dpranke@chromium.org
BUG=
TEST=
Review URL: http://codereview.chromium.org/7058054
git-svn-id: svn://svn.chromium.org/chrome/trunk/tools/depot_tools@87842 0039d316-1c4b-4281-b951-d872f2087c98
Add is_new to be used in a later patch by checkout classes.
R=dpranke@chromium.org
BUG=
TEST=
Review URL: http://codereview.chromium.org/7054048
git-svn-id: svn://svn.chromium.org/chrome/trunk/tools/depot_tools@87838 0039d316-1c4b-4281-b951-d872f2087c98
It is because upload.py doesn't use sane default parameter.
TBR=dpranke@chromium.org
BUG=
TEST=
Review URL: http://codereview.chromium.org/7075029
git-svn-id: svn://svn.chromium.org/chrome/trunk/tools/depot_tools@87012 0039d316-1c4b-4281-b951-d872f2087c98
It's needed to start a mapreduce on GAE with this class.
R=dpranke@chromium.org
BUG=
TEST=
Review URL: http://codereview.chromium.org/7076016
git-svn-id: svn://svn.chromium.org/chrome/trunk/tools/depot_tools@87010 0039d316-1c4b-4281-b951-d872f2087c98
I wonder how it worked at all. That explains some weird behavior when it was
ignoring issues recently.
Also enabled limit=1000 in case there's a huge bottleneck.
R=dpranke@chromium.org
BUG=
TEST=
Review URL: http://codereview.chromium.org/7016038
git-svn-id: svn://svn.chromium.org/chrome/trunk/tools/depot_tools@85516 0039d316-1c4b-4281-b951-d872f2087c98
At the moment it just fetches the patch. It's still useful to debug rietveld
issues.
Improve resiliency to invalid patches.
R=dpranke@chromium.org
BUG=
TEST=
Review URL: http://codereview.chromium.org/6825085
git-svn-id: svn://svn.chromium.org/chrome/trunk/tools/depot_tools@81305 0039d316-1c4b-4281-b951-d872f2087c98
Exposes InputApi.rietveld and deprecate host_url.
This is useful since it places authentication at a single place.
BUG=
TEST=
Review URL: http://codereview.chromium.org/6825008
git-svn-id: svn://svn.chromium.org/chrome/trunk/tools/depot_tools@81019 0039d316-1c4b-4281-b951-d872f2087c98
The reason of this change (beside deleting 30 lines) is to fix pylint on Mac.
For an unknown reason, it was dying here trying to process the minidom import.
R=dpranke@chromium.org
BUG=
TEST=
Review URL: http://codereview.chromium.org/6799021
git-svn-id: svn://svn.chromium.org/chrome/trunk/tools/depot_tools@80770 0039d316-1c4b-4281-b951-d872f2087c98
It will be used for:
- git cl patch so binary files can be patched
- try jobs instead of doing a curl <url> | patch
BUG=
TEST=
Review URL: http://codereview.chromium.org/6792028
git-svn-id: svn://svn.chromium.org/chrome/trunk/tools/depot_tools@80355 0039d316-1c4b-4281-b951-d872f2087c98
maruel@chromium.org