Inlining values inside debugger #8019

lahodaj · 2024-12-05T13:56:56Z

This is a prototype of a feature to show variable values directly in the editor, as permitted by the LSP protocol.

It looks like this inside VS Code:

and inside NetBeans:

mbien · 2024-12-20T00:22:12Z

really useful feature! We have to watch out for side effects of calling toString() though.

The debugger does have "Variable Formatters" which can be configured in the settings. Collections (which can be large) will only print their size by default for example. (but I imagine you are probably planning to do that anyway)

lahodaj · 2025-03-16T17:53:22Z

I did a lot of improvements lately. Lets see if this is heading in the right direction.

lahodaj · 2025-03-16T17:53:42Z

really useful feature! We have to watch out for side effects of calling toString() though.

The debugger does have "Variable Formatters" which can be configured in the settings. Collections (which can be large) will only print their size by default for example. (but I imagine you are probably planning to do that anyway)

Done.

lahodaj · 2025-03-16T17:54:46Z

FWIW, in VSCode, this needs to be enabled explicitly using the:

"debug.inlineValues": "on"

setting. I didn't find a way, yet, to make it work with auto.

mbien · 2025-03-20T18:12:54Z

gave it another try and it worked great. I got one exception once but wasn't able to reproduce it a second time:

java.lang.NullPointerException: Cannot invoke "com.sun.source.util.TreePath.getLeaf()" because "path" is null
	at com.sun.source.util.TreePathScanner.scan(TreePathScanner.java:66)
	at org.netbeans.modules.debugger.jpda.ui.values.ComputeInlineValues.computeVariables(ComputeInlineValues.java:115)
	at org.netbeans.modules.debugger.jpda.ui.models.InlineValueComputerImpl$ComputeInlineVariablesFactory$1.run(InlineValueComputerImpl.java:344)
	at org.netbeans.modules.debugger.jpda.ui.models.InlineValueComputerImpl$ComputeInlineVariablesFactory$1.run(InlineValueComputerImpl.java:321)
[catch] at org.netbeans.modules.java.source.JavaSourceAccessor$CancelableTaskWrapper.run(JavaSourceAccessor.java:273)
	at org.netbeans.modules.parsing.impl.TaskProcessor.callParserResultTask(TaskProcessor.java:561)
	at org.netbeans.modules.parsing.impl.TaskProcessor$RequestPerformer.run(TaskProcessor.java:786)
	at org.openide.util.lookup.Lookups.executeWith(Lookups.java:288)
	at org.netbeans.modules.parsing.impl.TaskProcessor$RequestPerformer.execute(TaskProcessor.java:702)
	at org.netbeans.modules.parsing.impl.TaskProcessor$CompilationJob.run(TaskProcessor.java:663)
	at java.base/java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:572)
	at java.base/java.util.concurrent.FutureTask.run(FutureTask.java:317)
	at org.openide.util.RequestProcessor$Task.run(RequestProcessor.java:1403)
	at org.netbeans.modules.openide.util.GlobalLookup.execute(GlobalLookup.java:45)
	at org.openide.util.lookup.Lookups.executeWith(Lookups.java:287)
	at org.openide.util.RequestProcessor$Processor.run(RequestProcessor.java:2018)

the inline hint should also listen to the main on/off toggle and be potentially also configurable like its other friends (see options window):

(I think listening to the main toggle would be sufficient, finer grained options could be added when needed in followup PRs - we don't have to solve everything at once)

lahodaj · 2025-03-21T14:24:32Z

Thanks for testing. I'll look at the exception.

For the toggle - I think I would prefer to have a different toggle (people may want the debugger values, but not parameter hints, and maybe even vice versa). I'll look into that.

neilcsmith-net · 2025-03-21T14:31:19Z

I like where this is going, and agree that this shouldn't be aligned with the inline hints toggle but something separately configured.

mbien · 2025-03-21T14:36:35Z

What I had in mind is that I think it would be nice to retain the master switch for all inline hints, but add more shortcuts for the sub-switches. It could be somewhere in the debugger toolbar too potentially. Completely decoupled from the editor setting could also work - not sure what is more intuitive. (I suppose i could double-bind my mouse button to achieve something like a master switch still no matter how its implemented)

neilcsmith-net · 2025-04-09T13:36:57Z

Aside from commit cleanup (cause of failing test) is this close to ready for merging, or are we retargetting for NB27?

lahodaj · 2025-04-15T12:36:06Z

Moving to NB 27, although assuming I can stabilize tests, I would like to integrate as early as possible (so that there's time to fix any breakages).

mbien

there seems to be a problem where the inline hints state and the displayed values in the UI are out of sync.

When I press alt+I the checkbox in the view menu (and options) don't update to the new value, even though the inline hints do toggle within the editor. The UI does still work, but it can become inverted with the editor state.

This is a bit of a dejavu since I think this happened before while working on a prior inline hint PR - but I can't remember what the problem was.

java/debugger.jpda.ui/src/org/netbeans/modules/debugger/jpda/ui/values/ComputeInlineValues.java

ide/api.lsp/src/org/netbeans/api/lsp/InlineValue.java

ide/spi.debugger.ui/apichanges.xml

...bugger.jpda.ui/src/org/netbeans/modules/debugger/jpda/ui/models/InlineValueComputerImpl.java

lahodaj · 2025-05-19T08:15:44Z

there seems to be a problem where the inline hints state and the displayed values in the UI are out of sync.

When I press alt+I the checkbox in the view menu (and options) don't update to the new value, even though the inline hints do toggle within the editor. The UI does still work, but it can become inverted with the editor state.

This is a bit of a dejavu since I think this happened before while working on a prior inline hint PR - but I can't remember what the problem was.

I tried to reproduce, but does not happen for me. Is there something specific you do? (I did a merge, so there's a possibility that fixed it, although I'd be surprised if it did.)

mbien · 2025-05-19T12:40:43Z

got the same NPE again while testing (#8019 (comment)):

java.lang.NullPointerException: Cannot invoke "com.sun.source.util.TreePath.getLeaf()" because "path" is null
	at com.sun.source.util.TreePathScanner.scan(TreePathScanner.java:66)
	at org.netbeans.modules.debugger.jpda.ui.values.ComputeInlineValues.computeVariables(ComputeInlineValues.java:113)
	at org.netbeans.modules.debugger.jpda.ui.models.InlineValueComputerImpl$ComputeInlineVariablesFactory$1.run(InlineValueComputerImpl.java:361)
	at org.netbeans.modules.debugger.jpda.ui.models.InlineValueComputerImpl$ComputeInlineVariablesFactory$1.run(InlineValueComputerImpl.java:338)
[catch] at org.netbeans.modules.java.source.JavaSourceAccessor$CancelableTaskWrapper.run(JavaSourceAccessor.java:273)
	at org.netbeans.modules.parsing.impl.TaskProcessor.callParserResultTask(TaskProcessor.java:561)
	at org.netbeans.modules.parsing.impl.TaskProcessor$RequestPerformer.run(TaskProcessor.java:786)
	at org.openide.util.lookup.Lookups.executeWith(Lookups.java:288)
	at org.netbeans.modules.parsing.impl.TaskProcessor$RequestPerformer.execute(TaskProcessor.java:702)
	at org.netbeans.modules.parsing.impl.TaskProcessor$CompilationJob.run(TaskProcessor.java:663)
	at java.base/java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:545)
	at java.base/java.util.concurrent.FutureTask.run(FutureTask.java:328)
	at org.openide.util.RequestProcessor$Task.run(RequestProcessor.java:1403)
	at org.netbeans.modules.openide.util.GlobalLookup.execute(GlobalLookup.java:45)
	at org.openide.util.lookup.Lookups.executeWith(Lookups.java:287)
	at org.openide.util.RequestProcessor$Processor.run(RequestProcessor.java:2018)

I tried to reproduce, but does not happen for me. Is there something specific you do? (I did a merge, so there's a possibility that fixed it, although I'd be surprised if it did.)

Oops, I just noticed that this problem is already in NB 26 - so we can probably ignore it here since it is not a regression:

I don't have a reliable reproducer yet unfortunately. At some point the View menu can get out-of-sync if the state is toggled by other means, e.g alt-I hotkey or the checkbox in options -> inline hints. The same happened while testing the new value hints - but it turns out its nothing new.

lahodaj · 2025-05-20T15:18:01Z

I've fixed the NPE (hopefully), and squashed (to hopefully get a clean GH Actions). I wonder what are the current impressions on this?

mbien

looks good overall. Left a few comments.

...bugger.jpda.ui/src/org/netbeans/modules/debugger/jpda/ui/models/InlineValueComputerImpl.java

lahodaj · 2025-05-22T08:32:09Z

what is the purpose of the empty loop?

It was there to initialize the computer. But turned out I can simply instantiate it, and it (hopefully) works as well, using Session instead of ContextProvider).

mbien · 2025-05-22T12:27:34Z

what is the purpose of the empty loop?

It was there to initialize the computer. But turned out I can simply instantiate it, and it (hopefully) works as well, using Session instead of ContextProvider).

I guessed that something like that might have been the reason for it. I was just thinking since lookup returns a List and no lazily initialized structure, it could have been already sufficient without the loop itself. But the new impl is more direct and less mysterious - thanks.

lahodaj · 2025-05-27T11:21:16Z

Unless there are objections, I would like to merge sometime soon - probably today or tomorrow.

mbien · 2025-05-27T13:49:10Z

great feature! I agree that it would be good to get it in. Don't forget to squash ;)

lahodaj added LSP [ci] enable Language Server Protocol tests VSCode Extension already fixed labels Dec 5, 2024

lahodaj added this to the NB25 milestone Dec 5, 2024

apache locked and limited conversation to collaborators Dec 5, 2024

apache unlocked this conversation Dec 5, 2024

mbien added Java [ci] enable extra Java tests (java.completion, java.source.base, java.hints, refactoring.java, form) ci:dev-build [ci] produce a dev-build zip artifact (7 days expiration, see link on workflow summary page) labels Dec 6, 2024

apache locked and limited conversation to collaborators Dec 6, 2024

apache unlocked this conversation Dec 6, 2024

mbien removed the already fixed label Dec 6, 2024

mbien modified the milestones: NB25, NB26 Jan 15, 2025

lahodaj marked this pull request as ready for review March 16, 2025 17:52

lahodaj modified the milestones: NB26, NB27 Apr 15, 2025

lahodaj force-pushed the inline-values branch from 7551692 to 4e79eac Compare April 15, 2025 16:32

lahodaj changed the title ~~Prototype for inlining values inside debugger~~ Inlining values inside debugger Apr 15, 2025

lahodaj force-pushed the inline-values branch 2 times, most recently from af6ff55 to 470ee74 Compare April 17, 2025 17:08

lahodaj force-pushed the inline-values branch 2 times, most recently from b6b14eb to 380837c Compare April 22, 2025 06:15

mbien reviewed May 8, 2025

View reviewed changes

lahodaj force-pushed the inline-values branch from ddd8982 to 2498b7f Compare May 19, 2025 08:20

lahodaj force-pushed the inline-values branch from 10b9335 to 1cb6da7 Compare May 20, 2025 15:16

mbien approved these changes May 21, 2025

View reviewed changes

Inlining values inside debugger

a7da46c

lahodaj force-pushed the inline-values branch from 7f6b822 to a7da46c Compare May 27, 2025 14:06

lahodaj merged commit 98c78a5 into apache:master May 29, 2025
37 checks passed

neilcsmith-net mentioned this pull request Aug 22, 2025

NPE when Evaulate Expression window is opening #8749

Closed

Inlining values inside debugger #8019

Inlining values inside debugger #8019

Uh oh!

Conversation

lahodaj commented Dec 5, 2024

Uh oh!

mbien commented Dec 20, 2024 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

lahodaj commented Mar 16, 2025

Uh oh!

lahodaj commented Mar 16, 2025

Uh oh!

lahodaj commented Mar 16, 2025

Uh oh!

mbien commented Mar 20, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

lahodaj commented Mar 21, 2025

Uh oh!

neilcsmith-net commented Mar 21, 2025

Uh oh!

mbien commented Mar 21, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

neilcsmith-net commented Apr 9, 2025

Uh oh!

lahodaj commented Apr 15, 2025

Uh oh!

mbien left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

lahodaj commented May 19, 2025

Uh oh!

mbien commented May 19, 2025

Uh oh!

lahodaj commented May 20, 2025

Uh oh!

mbien left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

lahodaj commented May 22, 2025

Uh oh!

mbien commented May 22, 2025

Uh oh!

lahodaj commented May 27, 2025

Uh oh!

mbien commented May 27, 2025

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

mbien commented Dec 20, 2024 •

edited

Loading

mbien commented Mar 20, 2025 •

edited

Loading

mbien commented Mar 21, 2025 •

edited

Loading