Still without any intention to use it, we add the scale field in
the database just so we can already write the right validation for
the scale ID which is pretty important as our code will start relying
on it. E.g. we only support scaleid null, or greater than 0. Adding
the validation method clearly documents what our expectations are.
The reason why we are doing that is to leave the door open for
competency scales later on. Which means that we can write code
that will rely on the competency to provide its scale even though
the default behaviour will be to use the framework one.
* Unit tests for base exporter
* Use the course_summary_exporter for list_courses_using_competency
* Rename competency_with_linked_courses_exporter to competency_summary_exporter
* Split properties_definition in 2 (added read_properties_definition)
* Improve phpdocs
* Make some methods final
* Throw an error when other_properties overlap with properties
Making the persistent_exporter implement templatable, implies that the
exporters are designed for output - but they are not - they are designed
for any data returned through an external function. So lets not make them templatable,
and make the default function name "export".
- Use to_record instead of custom getters
- Remove custom formatting for taxonomies
- Fix bugs in exporter (PARAM_FORMAT was wrong)
- use $comp->get_context() instead of $comp->get_framework()->get_context()
Fixes include format_text/format_string
New persistent_exporter class to export persistent objects for use in external apis.
Always validate_context
Always use validated params - not raw params
Do not use "xxformatted" properties.
Previously I was attaching the drag / drop handlers to the list items,
which were nested which caused many wierd behaviours. Now I attach them only
to the span in the lists which are never nested.
* Added missing jsdocs
* Disable multi-select for the main competencies tree (only allow it in the popups).
* Disallow selecting the top node in the competencies picker
* When multi-select is disabled, the selectionchanged event now only passes the one node, not a list with one item.
* Removed the flash when loading the picker (by hiding the tree while it's being rendered)
When the validation already happened we did not call before_validate().
It must always be called as before_validate() can be used as if it was
a before_all() callback.
This includes:
* Copy user competencies to user_competency_plan when updating plan
with status complete.
* User competencies in user_competency_plan are deleted when the plan
is set back to draft or active.
* the api::list_plan_competencies method is now returning
user_competency_plan objects if plan status is complete.
* external functions and renderer have been adapted to the new api
methods.
* Remove fields status and reviewerid from user_competency_plan
1. The competencies tree should open with all nodes initially collapsed
(unless there is less than 20 results, in which case they should all be expanded)
2. When the tree is refreshed, all previously selected nodes should be remembered
and the tree expanded to that point.
3. While the tree is reloading - there should be a "spinner"
Add custom .on methods instead of expecting code outside the module
to know which node to bind to to listen for events.
Also keep the selectionChanged handler private to the competencyactions module.
Finally - fix the broken move competency action.
