Update lint.js now that we are using ESLint 9

[samreid]

From https://github.com/phetsims/chipper/issues/1451 there were many TODOs to upgrade the lint process. Most of this should be done before #1415.

[zepumph]

From https://github.com/phetsims/chipper/issues/1469 here are some checklist items:

• [x] Should lint-all do multiple in one batch? Or separate? Do we still want progress bar? If we combine repos and use same monolithic cache, will it augment or overwrite?
• [x] grunt lint-everything no longer works with --chip-away
• [x] lint-all is taking longer

[samreid]

I'll run performance tests of using the API or processes on cached or --clean, on mac and windows. That will help us know for sure whether we need both implementations.

[samreid]

mac
cache is fresh - API only => 9sec
cache is fresh - spawn only => 52sec
--clean - spawn only => 4m22s
--clean - API only => 54% then crash at 2m40s

windows
cache is fresh - API only => 16sec
cache is fresh - spawn only => 63s
--clean - spawn only => 4m23s
--clean - API only => 54% then crash at 4m14s

[zepumph]

Design goals:

1. No crashing possible
2. Cleaner implementation than one file with two implementations

Next steps:

1. Let's go all in on the Node CLI (API), and use sub processes to keep heaps for overloading. Lets start with 50 repos per sub process.
2. We can run multiple sub processes at once (multi threading)
3. We can remove the npx implementation
4. Keep the CLI running a single repo at a time, with a cache file per repo.

[zepumph]

• [ ] ~I think we should keep the ability to have debug logging, maybe turned on via the --verbose flag? What do you think?~

Subject: [PATCH] add authors, https://github.com/phetsims/chipper/issues/1502
---
Index: js/grunt/lint.ts
IDEA additional info:
Subsystem: com.intellij.openapi.diff.impl.patch.CharsetEP
<+>UTF-8
===================================================================
diff --git a/js/grunt/lint.ts b/js/grunt/lint.ts
--- a/js/grunt/lint.ts  (revision 9c6461e0c47a003490903950a5c25b4448071083)
+++ b/js/grunt/lint.ts  (date 1730747840002)
@@ -175,7 +175,8 @@
     cache: true,
     cacheLocation: path.resolve( getCacheLocation( repo ) ),
     fix: options.fix,
-    errorOnUnmatchedPattern: false
+    errorOnUnmatchedPattern: false,
+    flags: [ '--debug' ]
   };

   // Create ESLint instance

[samreid]

• [x] delete chip away
• [x] use number of processors to estimate batching

[samreid]

Ready for review in the chipper and perennial-alias branches lint-batches-chipper-1484. Note there is one remaining TODO that I could not figure out how to solve and need help on.

[zepumph]

• [x] Even before this patch, I was feeling like js/eslint/ was getting a bit unweildy. I think we should make a subdir called config/, and keep linting module and running code that the top level. SEE https://github.com/phetsims/perennial/issues/399
• [x] This morning I fixed the "reopen issues with TODOs" task and it will likely yield an annoying merge conflict, let's merge main into the branches together.
• I committed a change that removed cleanOnly and can just run a function. I'm unsure if cleaning tsc is needed from lint, but I like keeping it for now.
• I ran grunt lint --clean --all and it took 2 minutes before this commit https://github.com/phetsims/perennial/commit/8bc2e5238d298ba007ec367f2596e1b98a67cae9
• require( 'os').cpus().length -> 20 on MK windows, and 128 on sparky.

[zepumph]

 mjkauzmann ~/PHET/git/perennial-alias (lint-batches-chipper-1484)
 $ time grunt lint --all --clean
Running "lint" task
processes 4
Done.

real    2m18.486s
user    0m0.061s
sys     0m0.061s

 mjkauzmann ~/PHET/git/perennial-alias (lint-batches-chipper-1484)
 $ time grunt lint --all --clean
Running "lint" task
processes 15
Done.

real    1m56.170s
user    0m0.060s
sys     0m0.062s

[zepumph]

I added 6 TODOs as part of the review. @samreid and I merged to main, and the rest is in his court. Thanks SR!1

[samreid]

TODOs addressed, --processes added which sets the maximum number of allowable processes. We are merged to main. Feature branches deleted. Closing.

[samreid]

On second thought, I'm very confused about all the entry points that are allowed to pass through LintOptions. The default occurs in several places. Would be good for @zepumph to review and comment.

[zepumph]

@samreid and I agree it is awkward, and getLintOptions shouldn't always provide defaults that are duplicated with the module defaults. We will make a separate side issue.