TNL-4102

FEDX-121

The previous approach for handling NPM assets was
to symlink them into the static directory. This appeared
to cause trouble with the asset pipeline where the files
in question were not installed and then old versions were
picked up instead.

This change instead copies NPM libraries to a new
static directory so that the pipeline can consume them
as with any other file. This new directory is added to
.gitignore so that the files don't get accidentally
checked in.

This now correctly migrates and dumps files for both databases.

Run lettuce migrations for both databases

By default, migrations are applied as they always have been. Exporting DISABLE_MIGRATIONS=1 or passing --disable-migrations to Paver commands will create tables directly from apps' models.

Testsonly will not work currently, because the database flush will
remove some data fixtures. Even when they are loaded at the beginning
of the next test run, there are database corruptions. Instead of
flushing, do not do anything with the database at the end of the
testsonly test run. That way, tests that use dynamic data can be
run many times in a row.

- Rename escape_json_dumps to dump_js_escaped_json
- Rename escape_js_string to js_escaped_string
- Update js_escaped_string to output empty string for None
- Introduce dump_html_escaped_json
- Move dump_js_escaped_json after the pipe as new best practice
- Introduce additional uses of helpers
- Introduce new djangolib directory and move js_utils

ECOM-3384

revert shell script change

Extends the Programs ConfigurationModel, cleans up Programs-related utilities and corresponding tests, and corrects caching. Uses the Programs API to list programs within Studio. ECOM-2769.

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

This includes:
* Ability to specify number of processes to run bok-choy tests in
* A forked nose commit to get the multiprocess plugin's logging to work
* A different plugin (xunitmp) must be used for pulling together xunit results

This works by:
* Starting the various servers that are needed for the acceptance test environment
* Running the tests themselves in multiprocess mode

* Fix paver violations to stablize edx-lint update
* Parens, Line2Long
* Fix missing docstrings
* Fix PEP8 issues
* Address PR feedback (thanks @nedbat!)

On my devstack, the old line takes about 5.5 seconds, the new line takes
about 1.3 seconds.

For background:
http://nedbatchelder.com/blog/201505/be_careful_deleting_files_around_git.html

I want to make clear that these tests rely on XBlocks in other repos,
and that ideally in time we would move these tests into the XBlock repo.

The old line:

    from path import path

produced pylint errors because of the baroque way that path.py defined
"path".  We tried to get them to change how they defined it, but they
deleted the name instead: https://github.com/jaraco/path.py/issues/102

(Jason then changed his mind, but this is a better way to use path.py,
it avoids the pylint error at least.)

TNL-2465