This reverts commit c4f56620.

This reverts commit ab437b94, reversing
changes made to 7a8e87c7.

- Create new cohorts.py file
- moved code there from models.py
- implemented necessary functions.

Next: testing :)

fix cohorts in create, and read, and update regular expressions to fix courses with periods not working in General commentable

If there is no announcement date (in the policy metadata) then use a
heuristic to sort them.

If there is an announcement date, the use it in the heuristic to
determine if the course is new or not.

The property can be set in the policy metadata. If it is not specified
then it is set to true if the course has not started yet.

Also adds a property to check how many days are left until the course starts.

This cuts the number of queries in 6.002 courseware loads from ~650 to
~5-40. Still to do: cache CustomTag templates so that we only load them
once per request.

initial commit. Add 'import_path' field to the GridFS so that when we export, we know where it originated from

make sure when defining the course namespace in the replace url methods that we explicitly state the parameter name. There are some optional parameters and if we don't define the parameter name, it'll end put being interpreted as a different parameter

be sure to always pass in a course namespace - which can be just the location of the module itself since it has the same org/course pair - when rewriting links. Also, allow for an option parameter in get_module() to disable the wrapping of the module's HTML with the xmodule_display.html. This is needed for ancillary course content such as 'about', 'course info', which are now stored as xmodules, but should not be wrapped in xmodule_display.html as it breaks some styling

add to existing test cases to exercise the 'course extras as modules' work in the CMS import. Also add to the existing 'full' test data collection to include policy, tabs, etc.

make course about view methods render the about content as a module, so we get all the url rewriting goodness. Also, since we're now handling the url re-writing via the module get_html pipelines, we can remove the link rewriting inside the xmodules itself - which is good because there's a wierd bug in lxml.html rewriting

implement static tabs. rejigger .tar.gz import to do the re-namespacing as a post-processing step as we need to retain the original url_name during the import. Also, migrate the presentation tier to use module_render.py::get_module() to better unify HTML'ifying stuff

format should be
<ul>
<li>link</li>
<li>other link</li>
</ul>

* looks in about/{course.url_name}/ first, then in about/
* same for info/