Also added datadog application to commom django apps and set it to use
a local statsd server by default.

Necessary for RFC compliance with oauthlib

Conflicts:
	cms/djangoapps/contentstore/views/course.py
	common/lib/xmodule/xmodule/tests/test_combined_open_ended.py
	lms/envs/common.py

Affects:
Choicegroups, option inputs, textline, and formula equation inputs

Comments (and processing instructions!) are handled oddly in lxml.
This change will keep them from causing failures.  They will be omitted
from the HTML generated, which is fine, since they aren't needed there.

Comments (and processing instructions!) are handled oddly in lxml.
This change will keep them from causing failures.  They will be omitted
from the HTML generated, which is fine, since they aren't needed there.

Patched Snuggletex server calls

Fixed coffeescript event logging error.

Fixed crowdsourced hinter dependence on self.field.append() not working.

Fixed a css bug in the hinter.

Fixed broken tests.

Got rid of answer signatures - these don't work with approximate answers at all.  Instead, implemented comparison-based hint matching.

Tests are broken, hint manager is probably broken.

In preparation for adding hints to formula response.

Rather than just calling `complex(...)` on the staff answers, also call
`calc.evaluator`. This way, simple things like `1/3` can be entered, as well
as more complicated explicit equations (e.g. `sqrt((12^2 + 26^2)/2)` etc, etc)

Simplify and update the tests to work with this.

Named `FormulaEquationInput` (name up for debate)

- Based off ChemEqnIn
- Add FormulaEquationInput in inputtypes.py
- Add a call to a skeleton method for a preview

javascript:

- Queue up some MathJax
- Put some ordering on the AJAX requests: add a parameter when the request was started, when it returns check that it isn't outdated before displaying the preview
- Tests

Note: we moved the `jsinput` tests and DISABLED them, because they were causing the tests to fail.

- Create a method called `parse_algebra`. It takes a string of math and returns with a `pyparsing.ParseResults` object representing it.
- `evaluator` takes this tree and applies the old "parse actions" to it to get the same number as it used to.
- Change calc's API: `evaluator` to use `case_sensitive` rather than `cs`
- Add most of the capability for latex rendering

Deleting the module object isn't needed to replace it, and deleting
a module object causes all of its attributes to be set to None.

Refactor a bit modx_dispatch