[23] JEP 467: Render markdown doc comments

[stephan-herrmann]

• add dependency org.commonmark incl. extension for tables
• use it for rendering
  • markdown variant is detected by inspecting the raw javadoc text, might be replaced by new API in DOM
• minimal tests incl. examples from the specification in https://openjdk.org/jeps/467

fixes #1472

[stephan-herrmann]

@merks This PR boldly requires org.commonmark (actual client is org.eclipse.jdt.core.manipulation), and expectedly fails with

12:00:54  [ERROR] Cannot resolve project dependencies:
12:00:54  [ERROR]   Software being installed: org.eclipse.jdt.core.manipulation 1.21.200.qualifier
12:00:54  [ERROR]   Missing requirement: org.eclipse.jdt.core.manipulation 1.21.200.qualifier requires 'osgi.bundle; org.commonmark 0.22.0' but it could not be found

How do you suggest to fix this?

• Can you add org.commonmark to eclipse-sdk-prereqs.target?
  • Apparently, that repo doesn't have a BETA_JAVA23 branch, so would it be OK, to add the dep in master?
• Should I override the configuration for tycho's target-platform-configuration ?
  • org.eclipse.jdt.core.manipulation is pom-less, so I guess the change would go into eclipse.jdt.ui root pom.

[merks]

Please add it to the target in https://github.com/eclipse-platform/eclipse.platform.releng.aggregator

I’m away from the office but can review a PR if you add the to the orbit location in the target.

[stephan-herrmann]

This PR still shows the 4 failures from #1526 which are fixed in master by eclipse-jdt/eclipse.jdt.core#2747 so JDT/Core needs its beta branch updated.

@jarthana @mpalat are you doing merges master->BETA_JAVA23 by some schedule or just as needed?

[jarthana]

This PR still shows the 4 failures from #1526 which are fixed in master by eclipse-jdt/eclipse.jdt.core#2747 so JDT/Core needs its beta branch updated.

@jarthana @mpalat are you doing merges master->BETA_JAVA23 by some schedule or just as needed?

No fixed scheduled per se. Will do it today.

[jjohnstn]

Hi @stephan-herrmann I seemed to have caused a git accident. I was trying to merge all the PRs for markdown in my local branch using gh pr merge and it actually merged it here. I will have to revert. My apologies. I do think the patch is valid but I couldn't test it without the partitioning logic.

[jjohnstn]

Ok, I reverted the merged changes and then created a revert of the revert to add back the original PR commits. See #1586

[stephan-herrmann]

OMG, bye bye clean history ;-P

FYI, I have one more commit in my local branch, but all this waits for necessary changes in and around eclipse-jdt/eclipse.jdt.core#2807.

Locally, all markdown hovers look nice already :)

I might as well start a fresh new PR, once the core part is ready.

As we are doing incremental work in the face of a deadline approaching, perhaps you could leave the merging in BETA_JAVA23 to me, while I coordinate with Jay?

[stephan-herrmann]

BTW, @jjohnstn when I edit the new class CoreMarkdownAccessImpl saving consistently removes a required import. Do you observe the same? My build is 20240813-0159

[jjohnstn]

@stephan-herrmann I had no intention of merging anything in BETA_JAVA23. There was a stack overflow posting that mentioned how to get multiple PR patches using merge and I tried using gh which was the wrong thing to do. I'll take a look at the removed import.

[jjohnstn]

@stephan-herrmann Did you mean 20240813-0100 from https://download.eclipse.org/eclipse/downloads/ ? I tried the Linux x86_64 version to edit the BETAJAVA_23 checked out branches of jdt.ui, jdt.core, and jdt.debug. The project settings for jdt.core.manipulation has organize imports on save, but I see no changes after some minor edits and it properly adds any new collection classes I refer to. In my Target Platform I point to 4.33-I-builds and https://download.eclipse.org/tools/orbit/simrel/orbit-aggregation/milestone/S202407231258/ for the commonmark plug-ins.