openstack/nova - nova - OpenDev: Free Software Needs Free Tools

Apart of making pep8 version standard across all openstack
projects.
With this change we ignore E712 since it is normal to use
"column == True" in sqlalchemy.
Change-Id: I73a162847a79558cb158112878033edf18039805
Signed-off-by: Chuck Short <chuck.short@canonical.com>

Some ways of running run_tests.sh will not output a line for us to
parse the count from. Treat no "Ran X tests" as "Ran 0 tests".
Change-Id: I96ba13e797ceaf9b9fa2b60ccb558510b5e32a88

pep8 E502 and E712 were temporarily disabld when upgrading to pep8 1.2
Change-Id: If5158304cb8e73ef844dea9f726df6f8fe305269

Right now, run_tests seems to report test success if there is a
catastrophic failure early in the process. This can happen if you
have a python file that fails to parse, or if something else in
the test environment setup fails to initialize (such as creating
the empty database for the test run). Nosetests will report
success, but indicate that zero tests were run.
This patch snatches the output of nosetests and, if zero tests
ran, reports to the user that something is broken and offers a few
suggestions.
Change-Id: Ide6fd8b212f0650cb227f60b25d4b4e62ee150ac

Patch I32a8808c6c9cccfedcc4d2a26649333aca1cd713 changed pep8 to 1.2
but did not update the ignore list in run_tests.sh
Change-Id: I9d2cab79197ba3cdd7103d4532e28a999c9454f4

Our new lockfile code works fine if the locks already exist, but
it is annoying to pollute the source tree with lockfiles in the
tests.
This modifies the lockfile code slightly to use a tempdir if
lock_path is none. Note that a tempdir will not work in production
but it is fine for testing. We set the lock_path to None in the
testing flags.
Also removes the explicit cleanup in run_tests.sh. The new solution
works with tox and will work even if we add new locks.
Fixes bug 1035426
Change-Id: I613926e4288442e79df50ea5f5bd8f9eafdfe7e7

Messages translated into other language like Chinese cause some
unit tests to fail.
Fixes LP bug #1037946.
Change-Id: I01d04333d28bab52d6f424432524b206f3845fba

Remove a number of limitations from the external locks.
- They can be nested now
- They do not need external cleanup in case of failures
- They do not rely on lockfile or greenlet internal implementation
New implementation is based on fcntl locks and any crashing process
will drop the lock. It does not have to rely on any cleanup code or
handling exceptions. Because no cleanup is needed, a number of tests
have been removed.
This implementation is not portable outside of POSIX/BSD/SVR4 systems.
Fcntl locks should work correctly with NFS mounts.
Locks are cleaned up after the tests finish running via run_tests.sh,
even though it's not strictly needed.
This change requires eventlet >= 0.9.17.
bp improve-external-locking
Change-Id: Idf5424c04645f25097733848a007b150145b0b27

Updates tox.ini and run_tests.sh to exclude openstack-common code
from pep8 checks.
Fixes LP Bug #1020792
Change-Id: I6bb9747f86f7677f0e9c4169c04d6f3fbd4b3480

find . -executable is not available on Mac OS X. We switch back to
the old -perm syntax to find files to test.
Change-Id: Iea58a249f282c98ae6cffbe8e6359f70f8b1d63f

Fixes bug 1018641
Change-Id: Id61947252fe4d3f5b93279518128ba90100f3748

Shrink tox.ini to the new short version.
Fix the test cases to be able to be run in nosetets plus the
openstack.nose_plugin, which finally removes the need for
nova/testing/runner.py
Also, now we'll just output directly to stdout, which will
make nose collect the trace logging directly and either output
it at the end of the run, or inject it into the xunit output
appropriately.
Change-Id: I1456e18a11a840145492038108bdfe812c8230d1

And fix existing alphabetical ordering problems
Change-Id: I3ab887e118adc8022c84ff7230e7cc2d261804f6

Ignore nova/openstack/common code when calculating coverage.
We do this because the unit tests for the common code are in a
different repository.
Change-Id: I0103b8acfc4cdaace68a18b1d1723a13595debe2

* cleaner output
* fix bug 980009
* Fix N201
* N306: alphabetical order imports
* N401: docstring start
* N402: one line docstring start
* N403: multi line docstring end
* Until fixed, N40* will be disabled by default
Change-Id: I9addafdaa7a1f8fb950e14a5409f661dec6c7b87

Coverage report generation was not using --include filters.
This resulted in generating a report that included all external
dependencies. Some of these may include problematic characters for
utf8 decoding. In the odd case this was found, coverage would fail
with a UnicodeDecodeError.
Fixes bug #987077.
Change-Id: I276698e65ac1037972209a406058168433cd826b

Fixes LP Bug #987055.
Change-Id: I7132b88c5e0a90b16d0f8e35df45b82e7991280f

Change-Id: I9d4dc11083cb51040677eeda2d28d4b23539a304

Change-Id: I81db73588ece36dec4dba90b9841ed6d0fac6a8f

Fixes bug 977489
--repeat became the default in 0.7; since we're now requiring 1.0 or
above, this can go away.
Change-Id: I85ee8ab274594ce96ece6610247dfd8a6f082af5

Update test-requires to use pep8>=1.0. Removed PEP8 warning suppression
around 3-arg raises.
Change-Id: Ib4ed42adc167aa1e8078619a36b409b76b9f5d73

Fixes bug #945346
Change-Id: I07a303c2e503e50d7138585c683e0d1310339276

* Works around migrate issue 72 [1] by monkey patching migrate prior
 to import.
* Removes previous workaround
* Refactor test config to work under nosetests directly
* Fixes LP940407
1: https://code.google.com/p/sqlalchemy-migrate/issues/detail?id=72
Change-Id: I219e4cecf8bb2e34ae238ac270428f496378ee61

See Invalid bug 937463 for more details
Also change run_tests.sh to ignore more '*.pyc' files
Change-Id: I8c8ded902833fe4268966ff7a9630daaaad43a19

There is absolutely no reason to ignore E202 in the pep8 checks.
Change-Id: I4abf767639dd94e9e8b7b4a405b4a702a554b876

Fixes bug #854905
Use sqlalchemy reflection whenever possible:
 http://sqlalchemy-migrate.readthedocs.org/en/latest/versioning.html
Work around sqlalchemy-migrate sqlite 'bool column not deletable' issue:
 http://code.google.com/p/sqlalchemy-migrate/issues/detail?id=143
 Add following sql scripts for sqlite:
 002_sqlite_downgrade.sql
 015_sqlite_downgrade.sql
 033_sqlite_downgrade.sql
 050_sqlite_downgrade.sql
 068_sqlite_downgrade.sql
Work around sqlalchemy-migrate sqlite 'table with foreign key column not
deletable' issue:
 http://code.google.com/p/sqlalchemy-migrate/issues/detail?id=94
 Add following sql scripts for sqlite:
 003_sqlite_downgrade.sql
 006_sqlite_downgrade.sql
 007_sqlite_downgrade.sql
 012_sqlite_upgrade.sql
 013_sqlite_downgrade.sql
 020_sqlite_downgrade.sql
 030_sqlite_downgrade.sql
 038_sqlite_downgrade.sql
 042_sqlite_downgrade.sql
 053_sqlite_downgrade.sql
 067_sqlite_downgrade.sql
Work around sqlalchemy-migrate 'migrate drops engine reference' issue:
 http://code.google.com/p/sqlalchemy-migrate/issues/detail?id=72
Add following sql scripts for long primary key to work with utf-8
mysql table:
 072_mysql_upgrade.sql
Add following sql scripts for postgresql:
 002_postgresql_downgrade.sql
Add snake walk test cases for database migration based on glance migration
test.
Change-Id: Ib454ecb4662bbf47736c1b12d9a4f969f180ceb6

Change-Id: If7ad58f131d8b780125df7ecb5c4081014064f72

Fixes bug #929369
Fix "api-paste.ini* to "*api-paste.ini*" and
remove duplicate definitions of srcfiles.
Change-Id: I859783457d6c5df3f6e998e752cad648ca109b9c

check every python file under <nova_root>/plugins/xenserver/.
Ignore patch files.
Change-Id: Ib1170ddabe03de746aae570d30b133aaffb09c88

Fixes bug #928976
Avoid pep8 check on api-paste.ini
Change-Id: I14e84e3c704c4e90d87c73e484d8eb3b63803743

Removes copy of ajaxterm code, nova-ajax-console-proxy,
and support for get_ajax_console from Nova proper.
Implements blueprint remove-ajaxterm
Fixes bug 917963
Change-Id: I2c0ff427c53c0f63a18b10475d6b4cbe9a085d83

Tests so far:
N101 TODO format
N201 Except format
N301 One import per line
N302 import only modules
N303 Invalid Import
N304 Relative Import
Change-Id: I33c021b842e7199b1f5f1f699ea17f7fa5f8ca49

Fixes bug #904559
Change-Id: I078b060483890bf8bfc9b78d3174370dea4d6bbc

This simplifies a number of Jenkins jobs which currently, other
than directory names, could be the same for all OpenStack
projects. By renaming the virtualenv directory, the redundant
Jenkins virtualenv build and copy jobs can be eliminated.
Change-Id: Ieaf1dac3207ecb34b911c7edcd2086809abdf49e

This commit begins to implement blueprint consolidate-testing-infrastructure by
adding a 'testing' subpackage and moving some modules into it.
Change-Id: I04bf860bc386bd2016e7dbc5a6f6ef7379a855bb

Since moving to bzr, we no longer generate ChangeLog and vcsversion.py
and since commit deb31cb55 we no longer even have the bzr specific code
for generating them. So, let's just remove any references to them.
Change-Id: I4f96b9be48e289f9129ae8e3ad4cbc1b22db07d2

Added a flag to run_tests.sh to allow user to optionally install venv with --no-site-packages.
This fixes bug 880905
Change-Id: Ic645e0ec56c90b72fef526ebc9f55975d446e2ae

Fixes LP837687
Change-Id: I7c0ea900225024e69a0265c430724930ce7b892d

Also, add an option to run_tests.sh to skip running pep8. We have a separate
job in Jenkins which runs pep8, so there's no need to spin our wheels on it
during the test run.
Change-Id: I552330994e55a36fa96b63658493ed30e2720c85

Add set -eu to run_tests.sh. This will cause it to fail whenever anything
goes wrong, which is exactly what we want in a test script.
To do this, I had to remove the use of the "let" keyword, which has a bizarre
exit status in bash.
I also removed the "|| exit" after run_tests, which means that this script
will now exit with status 1, not status 0, if run_tests fails.