And don't load javascript assets if plugin is disabled.
* precompile auto generated plugin js assets
* SPEC: remove spec test functions
* remove plugin js from test_helper
Co-Authored-By: Régis Hanol <regis@hanol.fr>
* DEV: using equality is slightly easier to read than inequality
Co-Authored-By: Régis Hanol <regis@hanol.fr>
* DEV: use `select` method instead of `find_all` for readability
Co-Authored-By: Régis Hanol <regis@hanol.fr>
* Calling `Discourse.reset()` creates a new container
We should run our de-initializers only after acceptance tests,
since initializers are not run outside of acceptance tests anyway,
and the container at this point can be passed properly to the
`teardown()` method.
* Remove `Discourse.reset` from tests
This would cause a new container to be created which leaks many objects.
* `updateCurrentUser` is more accurate than `replaceCurrentUser`
* renames `select-box-kit` into `select-kit`
* introduces `single-select` and `multi-select` as base components
* introduces {{search-advanced-category-chooser}} as a better component for selecting category in advanced search
* improves events handling in select-kit
* recreates color selection inputs using {{multi-select}} and a custom {{selected-color}} component
* replaces category-selector by a component using select-kit and based on multi-select
* improves positioning of wrapper
* removes the need for offscreen, and instead use `select-kit-header` as a base focus point for all select-kit based components
* introduces a formal plugin api for select-kit based components
* introduces a formal pattern for loading and updating select-kit based components:
```
computeValue()
computeContent()
mutateValue()
```
* Allow running specific plugin tests using ENV variables
* Add a `rake plugin:qunit` task to match the existing `rake plugin:spec` task
* Improve docker.rake to allow running specific plugin qunit tests
* Purge cache before and after qunit tests
* Stop module auto-loader trying to auto-load tests
* Use URL query parameters to pass config into Qunit, avoiding caching issues
* Oops, searchParams doesn’t work in phantomJS. Parse the URL manually.
* Escape ampersands before passing URL to phantomJS, otherwise multiple parameters go wrong
This commit removes the old evilstreak markdownjs engine.
- Adds specs to WhiteLister and changes it to stop using globals
(Fixes large memory leak)
- Fixes edge cases around bbcode handling
- Removes mdtest which is no longer valid (to be replaced with
CommonMark)
- Updates MiniRacer to correct minor unmanaged memory leak
- Fixes plugin specs
This adds the markdown.it engine to Discourse.
https://github.com/markdown-it/markdown-it
As the migration is going to take a while the new engine is default
disabled. To enable it you must change the hidden site setting:
enable_experimental_markdown_it.
This commit is a squash of many other commits, it also includes some
improvements to autospec (ability to run plugins), and a dev dependency
on the og gem for html normalization.
FIX: client's translation overrides were not working when the current locale was missing a key
FIX: ExtraLocalesController.show was not properly handling multiple translations
FIX: JsLocaleHelper#output_locale was not properly handling multiple translations
FIX: ExtraLocalesController.show's spec which was randomly failing
FIX: JsLocaleHelper#output_locale was muting cached translations hashes
REFACTOR: move 'enableVerboseLocalization' to the 'localization' initializer
REFACTOR: remove unused I18n.js methods (getFallbacks, localize, parseDate, toTime, strftime, toCurrency, toPercentage)
REFACTOR: remove all I18n.pluralizationRules and instead use MessageFormat's pluralization rules
TEST: add tests for localization initializer
TEST: add tests for I18n.js
