hook to include javascript

please try to strip out the Drupal 5 stuff. I cannot review two different things at once.

I don't really like the way you combined two entirely different hooks/functions in one hook.
Can we not introduce a single new hook for JS, hook_quicktag_add_js()?

Bèr

do you have to use drupal_add_js('?q=quicktags') code? if you just print the javascript code inside script tags, then everyone will be able to use drupal_add_js and we will get rid of an extra drupal load caused by ?q=quicktags.

here is the patch on what i'm talking about.
neither an extra hook nor an extra page call(?q=quicktags)

Looks good. I need to test it, or someone else should.
Once tested and confirmed to work, I will commit it.

one should use drupal_add_js inside the other hooks(alter and insert). then it will be included only when needed. i think this solves including js issue.
Besides being against js hook, i suggest removing one of the 2 hooks. since the item arrays are merged it is possible to alter a previosuly added item by using insert hook, isnt it? Actually i'm talking about something like drupal menu hook. it allows both inserting new items and altering previously added items. Anyway, this is a new topic of course.

we can first initiate our default items(not by insert hook, no need for own insert hook). then run a single hook (lets say hook_quicktags_items).
Mostly the case for modules is adding new items and/or altering our default items. And this single hook is sufficient for this. Only drawback of this is; a module that wants to alter items of another module(i don't know if there is such a case) should set it's weight high if it comes before the other alphabetically. this is exactly the same situation in menu hook.
and i think we should discuss this in another topic.

I am not a big fan of adding a new -private- way to include JS, when we already have one in Drupal.
I see the problem that the Drupal Add_js is not powerfull enough, so we should address that too in an issue against core, surely quicktags is not the only project which misses this power/flexibility in core.

That said, see several solutions, we should look for the best one:
* a hook in quicktags. Downside: extra layer on top of a core layer. Pro: simple and flexible
* a function quicktags_included(), returns TRUE wen we have QTs and FALSE when not. Allows any module to include its own JS with Drupal APis when needed by QT.
* a quicktags JS compile function: It will collect structured arrays with a hook, then turn that into a JS to be inclduded. This could be a starter for a Drupal core JS builder API; Downside: complex.

with the provided patch i can write for mymodule;

function mymodule_quicktags_insert() {
  drupal_add_js(drupal_get_path('module', 'mymodule').'/mymodule_qt_integrate.js');
  $path = base_path() . drupal_get_path('module','quicktags') .'/';
  $items = array(
    'mymodule-qt-image' => array(
      'icon' => $path .'ed_image.png',
      'location' => 'mymoduleQuickTags',
      'name' => 'Add an image',
      'accesskey' => 'm',
    ),
  );
  return $items;
}

this hook only runs when QT is in action.
so the javascript file is included only when QT is in action.
my button will fire mymoduleQuickTags function that is defined in this js file.
i know my button ID and can change anything about on pae load if needed.

what else may i need?

I like it. The explanation in #14 makes it celar to me, what you had in mind.

Therefore I would love it when people can reviw this a little closer. I have put it on y list of things to do after a 5.x release of quicktags (hin: test 5., and we can release it :) )

It is certainly worth it. Its just that I pointed out many times before, that I have a problem with patches that "do many things at once". I hope you understand that maintaining a project requires a lot of work and is a big responsibility. I cannot blindly commit patches that fix bug and slip in features in the same time.

Therefore, I told that this has to wait untill tehre is a stable 5.x release first. Only after such a release will start committing new features.

The work you did for the 5.x upgrade was almost all lost, because the patches included too much changes. I had to do it all over by hand. That is a pity. Lets work on this together: Its not that I don't like your work, I love it. Its just that in the form you present it to me, I cannot really use it. Sorry for that.