Adding support for evaluating watches containing binding variables.

[jlahoda]

When a watch contains a binding variable, it fails to show the value. This patch is trying to fix that - that's the case BINDING_VARIABLE: change.

In addition to this, this patch is trying to:

• add a test for the change, that will only execute under JDK 17.
• run debugger.jpda tests on JDK 11 and JDK 17
• disabling tests that failed for me locally - I used a fairly coarse way to disable them, so that e.g. in EvaluationTest, only one subtest fails. This could, presumably, be improved in further patches. IMO better to run some tests, than none.
• the info.getCompilationUnit() -> treePath.getCompilationUnit() change is a bit tricky: if scanning is running, the TreePath may originate in a different javac instance than is the one held in info. Which then leads to end position -1 (because the end position is held into CompilationUnitTree, but only for trees under that top-level tree), which will inject code at incorrect place. Using treePath.getCompilationUnit() is not 100% correct, I think, but it workarounds the issue at this specific place, at least

---

^Add meaningful description above

By opening a pull request you confirm that, unless explicitly stated otherwise, the changes -

• are all your own work, and you have the right to contribute them.
• are contributed solely under the terms and conditions of the Apache License 2.0 (see section 5 of the license for more information).

Please make sure (eg. git log) that all commits have a valid name and email address for you in the Author field.

If you're a first time contributor, see the Contributing guidelines for more information.

If you're a committer, please label the PR before pressing "Create pull request" so that the right test jobs can run.

[mbien]

hi @jlahoda, it is always nice to see more tests getting hooked into CI. Thanks!

the java-modules job has a section which runs a some debugger tests, this job is already very long (~50mins). Assuming the tests work on JDK 17 too, we could move them to the new job you created if you want:

netbeans/.github/workflows/main.yml

Lines 1289 to 1305 in 18db5e3

	# - name: debugger.jpda
	# run: ant $OPTS -f java/debugger.jpda test
	- name: debugger.jpda.js
	run: ant $OPTS -f java/debugger.jpda.js test
	- name: debugger.jpda.projects
	run: ant $OPTS -f java/debugger.jpda.projects test
	- name: debugger.jpda.projectsui
	run: ant $OPTS -f java/debugger.jpda.projectsui test
	# - name: debugger.jpda.truffle
	# run: ant $OPTS -f java/debugger.jpda.truffle test
	# - name: debugger.jpda.ui
	# run: ant $OPTS -f java/debugger.jpda.ui test

and

netbeans/.github/workflows/main.yml

Lines 1274 to 1275 in 18db5e3

	- name: api.debugger.jpda
	run: ant $OPTS -f java/api.debugger.jpda test

-> to new java-debugger job

This would have all debugging tests in one job and also help balancing job length a bit.

[mbien]

tests are now failing. The reason it didn't fail the first time it ran is because -Dtest.config= was set to an empty String, so nothing did actually run.

[mbien]

I am sorry, I forgot something. The job handle would have to be added to the list at:

netbeans/.github/workflows/main.yml

Lines 2393 to 2398 in 27ed405

	# last job depends on everything so that it is forced to run last even if a long job fails early
	cleanup:
	name: Cleanup Workflow Artifacts
	needs:
	- base-build
	- linux-commit-validation

so that the pipeline artifact is not cleaned up when the new job fails (allows restarts).

[mbien]

thanks. Looks good to me.

[mbien]

@jlahoda do you think this is good to merge or should we wait for another review? Don't forget to squash before merge.

[mbien]

@jlahoda features freeze for NB18 is on the 18th. Ideally all PRs should be merged till Monday night or they will be bumped to NB19.

[mbien]

thank you :)
all green -> merging