Refactor all S3 code to use the django storage API and add swift
settings for openstack deployments

Moves ENABLE_INSTRUCTOR_EMAIL and REQUIRE_COURSE_EMAIL_AUTH from settings files
to admin-accessible configuration models. This allows for the bulk email settings
to be modified without a new AMI deploy. See TNL-4504.

Also updates tests:
    -python tests mock out the new configurations in place of the old settings
    -lettuce test has been moved to bokchoy
        (note that there was some loss of coverage here - the lettuce tests had
        been doing some voodoo to allow for cross-process inspection of emails
        messages being "sent" by the server, from the client! In discussion with
        testeng, this seems outside the realm of a visual acceptance test. So,
        the bokchoy test simply confirm the successful queueing of the message,
        and leaves the validation of sending messages to the relevant unit tests.)
    -bok choy fixture has been added, to replace the settings in acceptance.py
    -lettuce and bok choy databases have been updated to reflect the backend changes

The new default is to have bulk_email disabled, we'll need to call this out in the
next OpenEdx release to ensure administrators enable this feature if needed.

TNL-4195

This is a clone (copy) of CSMH's declaration and methods with an added
id of UnsignedBigInAutoField

We should be able to delete the save_history code, but needs testing.

Add error logging when capa failures happen

Put StudentModuleHistory into its own database

Bump out the primary key on CSMHE

This gives us a gap to backfill as needed.
Since the new table's pk is an unsigned bigint, even for people who don't
consolidate CSMH into CSMHE, the lost rows are unlikely to matter.

Remove StudentModuleHistory cleaner

This feature check was commented out in 8ee9a967 more than 2 years ago,
we might as well remove the setting.

Deletes unused code from an old implementation of microsites.

Also remove its corresponding ACCESS_REQUIRE_STAFF_FOR_COURSE feature
flag.

Please note that this is a squshed commit and the work of:
Symbolist, macdiesel, nedbat, doctoryes, muzaffaryousaf and muhammad-ammar

This is primarily to reduce load on MongoDB, where we've lately
had performance problems that we suspect are caused by very
large course structures being evicted from MongoDB's cache. This
may potentially give us a path to better performance as well,
but that's not the goal of this commit.

Surprisingly, LZ4 seemed to actually run more slowly than zlib
for this. Possibly because of some overhead in the Python
bindings? GZip was also surprisingly slow given that it uses
zlib underneath (something like 5x slower).

Use separate cache backend for caching structures.

Abstract out course structure cache.

add datadog metrics for compressed course structure sizes

Since we're using a different cache background, we don't need to have a cache prefix

Use dummy cache backend for tests.

Fallback to default cache if course_structure_cache doesn't exist.

This is primarily to reduce load on MongoDB, where we've lately
had performance problems that we suspect are caused by very
large course structures being evicted from MongoDB's cache. This
may potentially give us a path to better performance as well,
but that's not the goal of this commit.

Surprisingly, LZ4 seemed to actually run more slowly than zlib
for this. Possibly because of some overhead in the Python
bindings? GZip was also surprisingly slow given that it uses
zlib underneath (something like 5x slower).

Use separate cache backend for caching structures.

Abstract out course structure cache.

add datadog metrics for compressed course structure sizes

Since we're using a different cache background, we don't need to have a cache prefix

Use dummy cache backend for tests.

Fallback to default cache if course_structure_cache doesn't exist.

This makes these files importable, and puts them into the new best guess
as to where files should live.

This is compatible with Django > 1.5, and allows us to re-enable the ProfilingDebugPanel

path.py objects report themselves as class objects, which confuses the heck out of
pylint. It tries to match variable names using the class-rgx regular expression
instead of the constant-rgx regular expression, and it doesn't match, so it
throws an error. Not sure how to fix pylint, so we'll just ignore these errors.

Instructors should use Insights instead.

Note that the features in this release are opt-in, and course and video
behavior will remain the same unless a course explicitly opts in.

Major pieces of functionality with this commit:

Allows the listing of a user's enrollments, course videos, and updates. In
order to make a course available for mobile use, course staff must explicitly
set the Course Advanced Setting "Mobile Course Available" to true. Course staff
will always see their own courses through the Mobile API regardless of this
setting, but students will only be allowed to see a course through the Mobile
API if this setting is set to "true". By default, a Course will *not* be
available for mobile use.

This is a Django app for video resource management. It is completely optional,
and is intended to allow video and operations teams to create new encodings of
videos (e.g. low res for mobile) and change CDNs without having to edit course
data directly. Course teams can now use a "EdX Video ID" setting for Videos,
which will leverage VAL. Video units that do not fill in an "EdX Video ID" will
behave exactly as they always have.

* The Mobile API is enabled with the ENABLE_MOBILE_REST_API feature flag.
* VAL is enabled with the ENABLE_VIDEO_ABSTRACTION_LAYER_API feature flag.
* VAL and the Mobile API both require ENABLE_OAUTH2_PROVIDER).
* The Mobile API is a read-only API, but VAL requires database migrations.
* Applications that make use of either the Mobile API or VAL must be registered
  with the OAuth2 provider app in Django Admin.

Also, removed the client-side analytics for logging in.
Ensures that analytics are collected for third-party-auth logins.
Fixed failing tests related to third-party-auth.

Clean up shopping cart processor API.
Fix UUID JSON serialization bug in CyberSource2 implementation.
Update test suite to use new CyberSource2 implementation.
Fix i18n messages in CyberSource2
Enable CyberSource2 implementation by default.