fdc6d91588
This commit "undoes"a previous hotfix, and allows a cms course_publish
signal to trigger a block_structure update_course_in_cache task, which
is run on an lms worker queue.
Changes:
-exposes ALTERNATE_QUEUE_ENVS
-adds routing layer in celery.py
-moves prior dev_with_worker settings file to devstack_with_worker
-moves course_block api functionality into openedx/core/djangoapps/content/block_structure
25 lines
1.2 KiB
Python
25 lines
1.2 KiB
Python
"""
|
|
The Course Blocks app, built upon the Block Cache framework in
|
|
openedx.core.lib.block_structure, is a higher layer django app in LMS that
|
|
provides additional context of Courses and Users (via usage_info.py) with
|
|
implementations for Block Structure Transformers that are related to
|
|
block structure course access.
|
|
|
|
As described in the Block Cache framework's __init__ module, this
|
|
framework provides faster access to course blocks for performance
|
|
sensitive features, by caching all transformer-required data so no
|
|
modulestore access is necessary during block access.
|
|
|
|
It is expected that only Block Access related transformers reside in
|
|
this django app, as they are cross-cutting authorization transformers
|
|
required across other features. Other higher-level and feature-specific
|
|
transformers should be implemented in their own separate apps.
|
|
|
|
Note: Currently, some of the implementation is redundant with the
|
|
has_access code in courseware/access.py. However, we do have short-term
|
|
plans for refactoring the current has_access code to use Course Blocks
|
|
instead (https://openedx.atlassian.net/browse/MA-1019). We have
|
|
introduced this redundancy in the short-term as an incremental
|
|
implementation approach, reducing risk with initial release of this app.
|
|
"""
|