summaryrefslogtreecommitdiff
path: root/tests
Commit message (Collapse)AuthorAgeFilesLines
* Merge "Extract test token data from auth_token middleware"0.3.2Jenkins2013-08-222-327/+342
|\
| * Extract test token data from auth_token middlewareJamie Lennox2013-08-192-327/+342
| | | | | | | | | | | | This data can be used elsewhere as well. Change-Id: I38b6c782b6d38e9e7d091f1c045a13ca10dfcc85
* | Merge "Initial Trusts support"Jenkins2013-08-226-14/+200
|\ \
| * | Initial Trusts supportSteven Hardy2013-08-206-14/+200
| |/ | | | | | | | | | | | | | | | | Implements client support for the basic trusts API operations, note this does not include support for the roles subpath operations, support for those can be added in a subsequent patch. Change-Id: I0c6ba12bad5cc8f3f10697d2a3dcf4f3be8c7ece blueprint: delegation-impersonation-support
* | Merge "Restore client.py for backward compatibility"Jenkins2013-08-221-0/+13
|\ \ | |/ |/|
| * Restore client.py for backward compatibilityJamie Lennox2013-08-211-0/+13
| | | | | | | | | | | | | | | | | | Simply pull in HTTPClient from its new location. This also restores the tests that were modified back to there original form to make sure nothing has changed from the original. Fixes: bug 1211998 Change-Id: Ia3241a4352eb6b37c6489cd4dd1bc373de107e12
* | Make auth_token middleware fetching respect prefixJamie Lennox2013-08-191-10/+96
|/ | | | | | | | | | In auth_token middleware auth_prefix is only added to requests that go through the _json_request method. This doesn't include the two certificate fetching functions. Manually add the auth_prefix to both those requests. Closes-Bug: #1211615 Change-Id: I25d1b401598c9a443ddef0fc3259ba859aee8c76
* Fix and enable Gating on H404Dirk Mueller2013-08-153-19/+11
| | | | | | | Enable gating on the Hacking H404 check - docstring should start with a summary. Change-Id: I80612a15bd11f689e9e9f4dc2ff812138630ddbd
* Merge "Adds support for passing extra tenant attributes to keystone"Jenkins2013-08-161-9/+28
|\
| * Adds support for passing extra tenant attributes to keystoneCharles V Bock2013-08-121-9/+28
| | | | | | | | | | | | | | | | | | | | | | | | Patch adds simple kwargs parsing to tenant update and create methods allowing users to exercise the extras functionality of keystone. As-per keystone data model docs: http://docs.openstack.org/developer/keystone/architecture.html#data-model Change-Id: I1c375c70a0ab4519626d2238d65fa0695f953dc5 Fixes: bug #1180317
* | Merge "Don't cache tokens as invalid on network errors"Jenkins2013-08-151-0/+21
|\ \
| * | Don't cache tokens as invalid on network errorsChmouel Boudjnah2013-08-091-0/+21
| | | | | | | | | | | | | | | Closes-Bug: 1172802 Change-Id: Ifc04f565b96d58a7a2f477a216b57af7f8de6a8e
* | | Merge "Refactor verify signing dir logic"Jenkins2013-08-141-0/+19
|\ \ \
| * | | Refactor verify signing dir logicWu Wenxiang2013-08-141-0/+19
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Add a function AuthProtocol::verify_signing_dir() in keystoneclient/middleware/auth_token.py. It will be called both in init stage and in fetching signing cert scenario. Fixes bug 1201577 Change-Id: Ice0e8a93ba9d85c20de84cf8da3e9b0f2274b740
* | | | Extract basic request callJamie Lennox2013-08-122-6/+101
|/ / / | | | | | | | | | | | | | | | | | | | | | Create a function out of the standard request call that handles common headers like user agent and logging. This makes future changes easier to digest. Change-Id: Ia25f997df64efdce27c8fb815e544922940145c3
* | | Rename client.py to httpclient.pyJamie Lennox2013-08-104-25/+25
| |/ |/| | | | | | | | | | | | | | | | | | | | | | | The discoverable entry point is to be client.Client however adding this functionality to the current client.py is impossible as we end up with circular dependencies. This patch simply renames the current client.py to httpclient.py to make future patches that will modify client.py more readable. Required for: blueprint api-version-discovery Change-Id: Ibcea03f6e1df0ae05329297166a8b8117fc3ce7b
* | Merge "Make TestResponse properly inherit Response."Jenkins2013-08-082-28/+11
|\ \
| * | Make TestResponse properly inherit Response.Jamie Lennox2013-08-062-28/+11
| | | | | | | | | | | | | | | | | | | | | There are a number of problems doing things like .json() if you don't inherit the correct parameters. Change-Id: I0dfc8ab6486090fbcd28792c18deae67e4769d53
* | | Fix test_request_no_token_dummy cms dependency.Clark Boylan2013-08-071-0/+6
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | * tests/test_auth_token_middleware.py: test_request_no_token_dummy has a dependency on the keystone.common.cms module importing subprocess prior to running. The cms._ensure_subprocess() function does this. Call this function in the base test class's setUp method to ensure subprocess is imported before symbols from that module are needed. Doing this in the base setUp will ensure that this works for any additional tests with the same dependency. This fix is needed because testr runs tests in non deterministic order which means that each test needs to be able to run on its own (eg in its own process without any other tests running first) and they need to run without interfering any other tests (each test must properly clean up after itself). Prior to this commit test_request_no_token_dummy would fail when executed on its own. Closes-Bug: #1209300 Change-Id: I3cf669b94ed77ad0cd729ead0b4162a3f1b8819f
* | | Pass the default_project_id when managing User.Lin Hua Cheng2013-08-021-1/+143
|/ / | | | | | | | | | | | | | | User default_project_id for setting the default project of the User. Fixes bug 1205520 Change-Id: I176d88bb6897c36404d8915385cc791adf10c586
* | Merge "Ec2Signer : Allow signature verification for older boto versions"Jenkins2013-08-021-0/+77
|\ \
| * | Ec2Signer : Allow signature verification for older boto versionsSteven Hardy2013-08-011-0/+77
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Since the fix for bug #1197553, verification for older clients (which strip the port when formatting the request) fails. This conditionally reverts to the original behavior, by detecting the boto version via the User-Agent header, the default behavior will be the new behavior (which doesn't strip the port), but this will allow a less painful transition for clients/distros to the new boto version. Fixes bug #1205281 Change-Id: I54ac9c5ba91e697004f1346a8f2d685da488992a
* | | Merge "flake8: fix alphabetical imports and enable H306"Jenkins2013-08-0210-11/+14
|\ \ \ | |_|/ |/| |
| * | flake8: fix alphabetical imports and enable H306Dolph Mathews2013-08-0110-11/+14
| |/ | | | | | | Change-Id: I0f4fcc9796e8529e7217dc24abe95660633cad33
* | Merge "Drop webob from auth_token.py"Jenkins2013-08-011-2/+45
|\ \ | |/ |/|
| * Drop webob from auth_token.pyPete Zaitcev2013-07-261-2/+45
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This came up as a packaging problem in Fedora. Now that Swift does not need python-webob RPM, some sites install a proxy node without. We add the proper webob dependency to the python-keystoneclient. Still, in the face of parallel installs, and we end patching setup.py and diverge from upstream, and we do not like it. It may be better just stick to WSGI protocol by small amount of hand-rolled code and drop WebOb altogether. Note that we keep WebOb in the testing code. The fix is purely about the stuff that runs inside a WSGI server. Therefore, requirements.txt continues having no WebOb (apparently, was forgotten before), but test-requirements.txt keeps WebOb in, and this patch changes nothing. Initial version of this patch produced an inexplicable drop in testing coverage and Chmouel put his foot down on it. We were unable to figure out why the drop happened. Most likely coverage simply lied about the previous percentages and getting rid of import webob permitted it a better analysis. So, we add some unrelated tests to bump coverage percentages up. Of course we add a targeted test for the new code too. For that one to work, we change how set_middleware() treats its fake_http argument. This version also adds comments rather than docstrings, because they need to source as context. Change-Id: Iba33b11e77b9910211983aa725c69d3886e1c7a7
* | Correct mis-spell in commentsZhang Xin2013-07-311-1/+1
|/ | | | | | | | In tests/v3/utils.py, parameterize() Fixes bug #1206895 Change-Id: I36e3d49209eb6896cb0e053bba9e4c966efea5db
* Merge "Fix and enable gating on H403"Jenkins2013-07-172-2/+4
|\
| * Fix and enable gating on H403Dirk Mueller2013-07-132-2/+4
| | | | | | | | | | | | | | Only a few occurrences of "multiline docstrings should end on a new line." Change-Id: I3fb4591259b1e7c8202365cbadc8967a1639254f
* | Merge "Ec2Signer : Modify v4 signer to match latest boto"Jenkins2013-07-151-0/+36
|\ \ | |/ |/|
| * Ec2Signer : Modify v4 signer to match latest botoSteven Hardy2013-07-061-0/+36
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Previously the port component of the host:port header was stripped to match what boto (2.6.0->2.9.2) did, however it seems that was a n error in boto as from boto commit cfaba39 (in 2.9.3) the port is now appended. This means that when this fix is used with keystone, APIs which use keystone to validate ec2 style v4 signatures (e.g the heat cfn API) will require python-boto >= 2.9.3 Fixes bug #1197553 Change-Id: I4c01e7aef7015a79e6e6263492c51caf3a08e9e4
* | Fix and enable gating on H402Dirk Mueller2013-07-124-7/+7
| | | | | | | | | | | | Docstring summaries need punctuation. Change-Id: I1b740c13d5fedf9a625ca0807c908f651ee08406
* | Add name arguments to keystone command.Ken'ichi Ohmichi2013-07-011-4/+39
|/ | | | | | | | | | | Fixes #Bug 1183655 This patch adds name arguments to the following subcommands: * user-list * user-create * endpoint-create Change-Id: I056e87de78f88f6fcf76baa4c820622c42f74f39
* Merge "Fix and enable H401"0.3.1Jenkins2013-06-283-17/+19
|\
| * Fix and enable H401Dirk Mueller2013-06-273-17/+19
| | | | | | | | | | | | Remove leading spaces from doc comments. Change-Id: I75b055c0d64dda478c63839d44158e301900107f
* | Merge "Unmock requests when testing complete."Jenkins2013-06-282-16/+34
|\ \
| * | Unmock requests when testing complete.Jamie Lennox2013-06-242-16/+34
| | | | | | | | | | | | | | | | | | | | | Any other tests running will unexpectedly have mocks installed for requests otherwise. Make the time mock behave the same way whilst there. Change-Id: Iec4f35c51ed4650a35ae2780637192c82042a980
* | | Merge "List groups by domain in keystoneclient."Jenkins2013-06-271-0/+24
|\ \ \ | |_|/ |/| |
| * | List groups by domain in keystoneclient.Nachiappan2013-06-241-0/+24
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Currently keystone support filtering of groups by domain. This fix is to add support in keystone-client for filtering groups by domain. Fixes Bug #1180581 Change-Id: I3906e2474d968426d06d498928ce2c7094c6fbe9
* | | Merge "Fix the cache interface to use time= by default."Jenkins2013-06-261-8/+14
|\ \ \ | |_|/ |/| |
| * | Fix the cache interface to use time= by default.Chmouel Boudjnah2013-06-211-8/+14
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Historically the swift cache conection used the argument timeout= for the cache timeout, but this has been unified with the official python memcache client with time= since grizzly, we still need to handle folsom for a while until this could get removed. Fixes bug 1193032 Change-Id: Ia9d544a0277beb197ed9824d7c1266d12968393d
* | | Merge "Python-2.6 compatibility for tests/test_keyring.py"0.3.0Jenkins2013-06-231-7/+2
|\ \ \
| * | | Python-2.6 compatibility for tests/test_keyring.pySascha Peilicke2013-06-171-7/+2
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | TestCase.setUpClass() and unittest.SkipTest are features added in 2.7. Calling super(KeyringTest, cls).setUpClass() could have never worked since tests.utils.TestCase doesn't implement that (class) method. Change-Id: Ifbb0465be24fd13ad68d52c0d7fa4c5090d0f694
* | | | Implements v3 auth client.lin-hua-cheng2013-06-2017-321/+1415
| |_|/ |/| | | | | | | | | | | | | | | | | | | | | | | | | | | | | Added support for domain scoping. Enhancement on AccessInfo to support reading v2/v3 token information. Enhancement on ServiceCatalog for reading/filtering v2/v3 service catalog information. Change-Id: Ibb678b9933d3673e37d0fba857a152a3c5d2b4f4
* | | Change memcache config entry name in Keystone to be consistent with Osloxingzhou2013-06-201-11/+11
| |/ |/| | | | | | | | | | | | | | | | | | | Currently, Keystone-Client is using 'memcache_servers' config option to store the memcache server info, while in OSLO project, it is using 'memcached_servers' to config the same option. It is better to change keystone-client's 'memcache_servers' to 'memcached_servers' to keep consistent between these two projects. Change-Id: I93ca0aa368f95a3ccf6de6984262057e61f75ffe Fixes: Bug 1172793
* | Fix memcache encryption middlewareBryan D. Payne2013-06-192-110/+75
|/ | | | | | | | | | This fixes lp1175367 and lp1175368 by redesigning the memcache crypt middleware to not do dangerous things. It is forward compatible, but will invalidate any existing ephemeral encrypted or signed memcache entries. Change-Id: Ice8724949a48bfad3b8b7c41b5f50a18a9ad9f42 Signed-off-by: Bryan D. Payne <bdpayne@acm.org>
* Merge "Use AuthRef for some client fields"Jenkins2013-06-151-2/+2
|\
| * Use AuthRef for some client fieldsJulien Danjou2013-06-111-2/+2
| | | | | | | | | | | | | | This tackles some TODO items left over. Change-Id: Ib062744acbf56f05d09857d244b78b35c0ef4d39 Signed-off-by: Julien Danjou <julien@danjou.info>
* | Merge "Cleanup shell's authentication check."Jenkins2013-06-141-1/+1
|\ \
| * | Cleanup shell's authentication check.Jamie Lennox2013-06-121-1/+1
| |/ | | | | | | | | | | | | | | The original logic for whether shell had enough information to authenticate a user was confusing and will be very difficult to extend to other forms of authentication. Change-Id: I88763a651e494e60070a30f0824505acb09310cb
View Lorry Controller Status