Commit Graph

4111 Commits

Author SHA1 Message Date
Calen Pennington
ec06b36660 Merge pull request #1787 from MITx/fix/will/random_seed_not_saved
Fix/will/random seed not saved
2013-04-01 10:15:14 -07:00
Will Daly
fb97602442 Save randomization seed if it is not already set,
so that the same problem loads when the user
checks/saves/resets
2013-04-01 12:08:53 -04:00
Will Daly
7f00487203 Reverted last commit due to concerns about database writes. 2013-04-01 11:48:10 -04:00
Calen Pennington
0e4369603e Merge pull request #1756 from MITx/fix/vik/oe-and-progress
Fix/vik/oe and progress
2013-04-01 08:31:58 -07:00
Will Daly
b76adb4b06 Updated tests to address on-going randomization bug.
Capa problem state was not being copied to module state
during module initialization.
This meant that when check_problem was called the first time,
the seed would be re-generated and saved.
2013-04-01 11:22:42 -04:00
Victor Shnayder
9c4a88b570 Merge pull request #1743 from MITx/fix/will/bug_288
Fix/will/bug 288
2013-04-01 07:22:20 -07:00
Will Daly
4322b83894 Added additional comments 2013-04-01 09:02:04 -04:00
Will Daly
f90e29ce96 Fixed a test error that occurred because the problem
was not set to done before resetting it.
To avoid to small chance that we'll get the same seed when
we reset, try to get a new seed multiple times.
2013-04-01 08:55:06 -04:00
David Ormsbee
24e64e3946 Merge pull request #1773 from MITx/hotfix/dave/mult_choice_display
HOTFIX: Accept either a list of possible values, or a string as a value for

I'm pulling this in myself because it's urgent, it's the weekend, and I don't want this to get forgotten.
2013-03-30 08:55:45 -07:00
David Ormsbee
dfd3a699b9 Accept either a list of possible values, or a string as a value for
comparison of correctness in multiple choice.

Multiple choice code is scattered and sometimes sends a list of
choices for the value, and sometimes a single string. We used to use
"in" which scarily handled both cases (list or substring search),
but that caused a bug when you had two choices like choice_1 and
choice10. Moving to == caused us to break when lists were sent to
us. So this ugly code is extra paranoid and checks both possibilities.

This really needs a better cleanup.
2013-03-30 11:09:44 -04:00
Don Mitchell
c655c814e8 Merge pull request #1721 from MITx/feature/cdodge/add-no-inheritable-metadata-on-verticals-to-xlint
add checking for metadata that we can't support editing for in Studio. T...
2013-03-29 12:38:42 -07:00
Don Mitchell
80f57db7c3 Merge pull request #1762 from MITx/feature/cdodge/course-overview-perf
Feature/cdodge/course overview perf
2013-03-29 12:35:41 -07:00
Chris Dodge
599ca4d429 oops. I'm not programming in C# any longer 2013-03-29 15:31:37 -04:00
Ned Batchelder
0cfcd183b2 No need to wrap comments that tightly. 2013-03-29 15:05:22 -04:00
Ned Batchelder
e8f8e9e197 Enough is enough. 2013-03-29 14:49:24 -04:00
Chris Dodge
3ce01882bb add an 'allowed' list of metadata (e.g. display_name, etc.) and also restrict metadata on sequentials 2013-03-29 13:59:59 -04:00
Diana Huang
17adc986bd Remove the default and prevent input_state from keeping around
unnecessary data.
2013-03-29 12:16:27 -04:00
Will Daly
4f9d18df8c Wrote unit tests to verify bug fix in
https://github.com/MITx/mitx/pull/1764
2013-03-29 12:00:17 -04:00
Vik Paruchuri
d5376e71ff Add in a test for the weight field 2013-03-29 11:46:44 -04:00
Vik Paruchuri
5391cefddc Add in tests to see if max score properly exposed and calculated in combinedopenended 2013-03-29 11:17:35 -04:00
Will Daly
f90dd49556 Fixed bug in parsing of urandom struct so that seed
is set to an integer (and correctly saved) instead of a tuple.
2013-03-29 10:18:11 -04:00
Vik Paruchuri
53620d7879 Merge remote-tracking branch 'origin/master' into fix/vik/oe-and-progress 2013-03-29 10:16:00 -04:00
Chris Dodge
c55d54b071 also, we don't support metadata on chapters 2013-03-28 13:03:34 -04:00
Jay Zoldak
c9a7be6b9b Merge pull request #1727 from MITx/feature/will/lettuce_test_helpers
Feature/will/lettuce test helpers
2013-03-28 07:26:14 -07:00
Don Mitchell
fece937669 Merge pull request #1758 from MITx/fix/cdodge/use-request-scroped-cache-for-metadata
Fix/cdodge/use request scoped cache for metadata inheritance
2013-03-28 06:51:20 -07:00
Chris Dodge
c7bafddace DRY things out a bit and share as much code between MongoModuleStore and DraftMongoModuleStore 2013-03-28 09:49:55 -04:00
Chris Dodge
3cdd973af4 get _cache_children to queyr both non-draft and draft versions of the children, then overwrite all non-drafts with the draft version, if available. This conforms with the semantics of the DraftMongoModuleStore 2013-03-28 09:28:19 -04:00
Chris Dodge
6c67ed1f57 Merge branch 'master' of github.com:MITx/mitx into feature/cdodge/course-overview-perf
Conflicts:
	common/lib/xmodule/xmodule/modulestore/mongo.py
2013-03-28 09:28:12 -04:00
Chris Dodge
3f52261b5b hmmm. actually, we should only write out to memcache if we've recomputed. Otherwise, a memcache hit will end up writing back to memcache... 2013-03-27 23:18:38 -04:00
Chris Dodge
d448aa1365 remove debug log messages 2013-03-27 23:13:33 -04:00
Chris Dodge
446397b23b remove unused thread.local() 2013-03-27 23:12:00 -04:00
Chris Dodge
b609a96902 ummm. forgot to commit stuff 2013-03-27 22:51:52 -04:00
Chris Dodge
03f9bb5d38 use a request-scoped cache to keep the metadata inheritence tree around for the whole request. This means we should only do one trip to Memcached/Mongo per course per request. This is expected to keep memory utilization down 2013-03-27 22:37:07 -04:00
Vik Paruchuri
0c218176d9 Run some code reformatting 2013-03-27 18:40:18 -04:00
Vik Paruchuri
df6d8fd2a3 Fix issues with progress page and open ended grading 2013-03-27 18:20:04 -04:00
Vik Paruchuri
4443afecaf Get rid of max score on open ended modules. Auto-calculate it from the rubric instead. 2013-03-27 18:05:00 -04:00
Will Daly
9c671163fd Added exc_info=True to log.warning
Changed log.exception to log.warning
2013-03-27 17:11:02 -04:00
Will Daly
f038237ee9 Changed log.exception to log.warning 2013-03-27 16:34:08 -04:00
Don Mitchell
6d63d13fa8 Merge pull request #1753 from MITx/bug/christina/studio
Converter cleanup.
2013-03-27 12:46:01 -07:00
Will Daly
df1be87739 * Changed 404 errors to 400 errors
* Removed duplicate traceback log message
* Now provide string, not Exception, as second tuple item to raise
2013-03-27 15:20:40 -04:00
David Ormsbee
6e3d9608b4 Merge pull request #1754 from MITx/fix/cale/enrollment-integrity-error
An integrity error while creating an enrollment just means that our work...
2013-03-27 12:03:26 -07:00
Calen Pennington
122c8567c5 An integrity error while creating an enrollment just means that our work has already been done. Fixes https://www.pivotaltracker.com/story/show/46915947 2013-03-27 15:00:08 -04:00
Will Daly
99cd3fafdb Added error handling of XModule processing errors to CMS
Added tests for SchematicResponse error handling
2013-03-27 14:48:44 -04:00
Will Daly
ac86687fa1 Added exception handling that solves SchematicResponse exceptions
causing a 500 error.

When XModule raises a ProcessingError during an AJAX request,
this module_render now returns a 404 to further reduce number of 500
responses.
2013-03-27 14:33:59 -04:00
Victor Shnayder
b0356e32b2 Merge pull request #1589 from MITx/peter/symbolic
Added support for superscripts in variables
2013-03-27 10:40:50 -07:00
Will Daly
cddc868656 Login URL resolves differently in LMS and CMS, which breaks login_test
when loaded by rake test_cms

I moved the test into lms/courseware/tests so they run correctly.
2013-03-27 13:01:10 -04:00
cahrens
227a5e8266 Delete converters, move unit tests to test_fields, add new additional test cases. 2013-03-27 12:56:06 -04:00
Will Daly
15ea32b095 Fixed bug 294, caused by unicode encoding error when creating logging
strings.

Added unit tests that verify the fix.
2013-03-27 12:53:58 -04:00
brianhw
a04881ceda Merge pull request #1740 from MITx/fix/vik/peer-image-grading
Fix/vik/peer image grading
2013-03-27 09:06:54 -07:00
Vik Paruchuri
7101c76016 comment on rewrite links change 2013-03-27 12:02:32 -04:00