@ts-expect-error chipaway

[marlitas]

Remove unnecessary @ts-expect-errors, or add a comment explaining why it is necessary.

Results from chipAway:

• gravity-and-orbits: @samreid 9 errors in 4 files.
• center-and-variability: @samreid, @chrisklus 3 errors in 3 files.
• circuit-construction-kit-dc: @samreid 1 errors in 1 files.
• circuit-construction-kit-common: @samreid 28 14 errors in 11 files see 14 ts-expect-error occurrences circuit-construction-kit-common#918
• bamboo: @samreid 2 errors in 2 files blocked by Node subtypes mutate function not allowing subtype-specific options scenery#1428
• axon: @samreid 56 errors in 16 files (@samreid requests help for these)
• bending-light: @samreid 15 errors in 7 files.
• sound: @samreid 4 errors in 1 files.
• calculus-grapher: @pixelzoom @veillette 1 errors in 1 files.
• counting-common: @chrisklus 6 errors in 2 files.
• density-buoyancy-common: @jonathanolson 14 errors in 11 files.
• kite: @jonathanolson 7 errors in 2 files.
• dot: @jonathanolson 6 errors in 4 files.
• mobius: @jonathanolson 3 errors in 3 files.
• phet-core: @jonathanolson 4 errors in 3 files.
• wilder: @jonathanolson 2 errors in 2 files.
• scenery: @jonathanolson 137 errors in 35 files.
• joist: @zepumph 13 errors in 8 files.
• molecule-polarity: @pixelzoom 2 errors in 1 files.
• my-solar-system: @AgustinVallejo, @jonathanolson 2 errors in 2 files.
• natural-selection: @pixelzoom 3 errors in 2 files.
• number-play: @chrisklus 4 errors in 4 files.
• studio: @chrisklus 13 errors in 8 files.
• phet-io-wrappers: @zepumph 35 errors in 11 files.
• phet-io: @zepumph 4 errors in 1 files.
• quadrilateral: @jessegreenberg 18 errors in 4 files.
• ratio-and-proportion: @zepumph 4 errors in 1 files.
• scenery-phet: @pixelzoom 1 errors in 1 files.
• sun: @jbphet 7 errors in 5 files.
• tambo: @jbphet 2 errors in 1 file.
• tandem: @zepumph 29 errors in 10 files.
• tangible: @zepumph 5 errors in 1 files.
• tappi: @jessegreenberg 2 errors in 1 files.
• utterance-queue: @jessegreenberg 9 errors in 5 files.

[pixelzoom]

I've addressed the ts-expect-errors in molecule-polarity and scenery-phet. The 3 in natural-selection all have (blocked) issues noted, and I'm expecting to remove them as those issues are addressed. Unassigning.

[jessegreenberg]

I reviewed my repos and ended up leaving most of my @ts-expect-errors in quadrilateral, utterance-queue, and tappi. They all involved prototype code or experimental technology (Web Bluetooth API, Vibration API, that kind of thing).

[pixelzoom]

@jessegreenberg That's great. But looking at utterance-queue, the majority of those @ts-expect-errors have either no associated comment, or a comment that's not very useful (e.g. "Don't judge me TypeScript..."). So there's no way that the reader is going to know that those are prototype or experimental source files.

[jessegreenberg]

I agree! Those comments were fun at the time but not so informative. They were updated in previous commits (I should have mentioned in my comment). Feel free to pull master to verify.

[veillette]

The @ts-expect-error in Calculus-Grapher was addressed.

[zepumph]

There are still 10 usages of @ts-expect-error in joist. This is due to:

• scenery.js is still in javascript (and will be some work to convert since examples/doc/playground files load that directly).
• assigning globals to window.
• Using nonstandard browser globals like window.navigator.pointerEnabled
• Node subtypes mutate function not allowing subtype-specific options scenery#1428

[pixelzoom]

In #1366 (comment), @AgustinVallejo said:

The two in My Solar System are not avoidable... One is inherited logic from bamboo/BarPlot, and the other one has to do with multilink handling of mapped arrays.

The one in CommonModel.ts (Multilink) is very avoidable. Here’s the diff:

-    // @ts-expect-error (multilink doesn't like non-tuple arrays for TypeScript, but works perfectly fine)
-    Multilink.multilink( this.availableBodies.map( body => body.isActiveProperty ), () => {
+    Multilink.multilinkAny( this.availableBodies.map( body => body.isActiveProperty ), () => {

When you have a Multilink or DerivedProperty that has an unknown number of dependencies, use Multilink.multilinkAny and DerivedProperty.deriveAny respectively.

[samreid]

We have started using @ts-expect-error as a way of doing "type check" unit tests. These @ts-expect-error should not be removed or considered in metrics with the "bad" ones. Should we standardize something to indicate this, like:

// @ts-expect-error for testing

[zepumph]

How about // @ts-expect-error INTENTIONAL - doc here? @samreid and I feel like intentional is already what we use for IntentionalAny.

[pixelzoom]

That's fine too, as long the doc here part is not optional. There needs to be information about why it's intentional.

[samreid]

That's fine too, as long the doc here part is not optional. There needs to be information about why it's intentional.

Perhaps one day a bad-typescript-text rule will enforce this.

[zepumph]

With this patch:

Subject: [PATCH] NumberProperty.reset handles issues with setting range or Property first, https://github.com/phetsims/axon/issues/427
---
Index: eslint/rules/bad-typescript-text.js
IDEA additional info:
Subsystem: com.intellij.openapi.diff.impl.patch.CharsetEP
<+>UTF-8
===================================================================
diff --git a/eslint/rules/bad-typescript-text.js b/eslint/rules/bad-typescript-text.js
--- a/eslint/rules/bad-typescript-text.js	(revision f1bf431c9398154880ddd568dfd351931bf5633d)
+++ b/eslint/rules/bad-typescript-text.js	(date 1671229930872)
@@ -49,6 +49,10 @@
     {
       id: '@returns with type and/or without extra doc',
       regex: /(@returns \{)|(@returns *$)/
+    },
+    {
+      id: '@ts-expect-error needs documentation',
+      regex: /@ts-expect-error\s*$/
     }
   ];
 

We only have 88 problems:

✖ 88 problems (88 errors, 0 warnings)

Results from chipAway:

• dot: @jonathanolson 2 errors in 1 files.
• kite: @jonathanolson 2 errors in 1 files.
• phet-core: @jonathanolson 1 errors in 1 files.
• scenery: @jonathanolson 69 errors in 23 files.
• studio: @chrisklus 11 errors in 7 files.
• tambo: @jbphet 2 errors in 1 files.
• utterance-queue: @jessegreenberg 1 errors in 1 files.
  Fatal error: Lint failed�

[samreid]

Are there ts-expect-errors in axon that weren't discovered in the above comment?

[zepumph]

Probably!

[marlitas]

SR: I would like some help with tricky Axon ts-expect-errors
JO: I can help

Thanks JO!

[samreid]

I finished the Circuit Construction Kit ones as part of phetsims/circuit-construction-kit-common#918