Added support for Tuples #374

FetBoba · 2026-01-01T15:13:06Z

No description provided.

…ples

…uples

LPTK

Neat! This looks well done to me. WDYT @Derppening?

LPTK · 2026-01-01T16:29:10Z

hkmc2/shared/src/main/scala/hkmc2/codegen/wasm/text/WatBuilder.scala

+        label = N,
+        children = Seq(assignInstr, rstBlk),
+        resultTypes = rstBlk.resultTypes.map: ty =>
+          Result(if ty is UnreachableType then RefType.anyref else ty.asValType_!)


What is this UnreachableType special case for?

hkmc2/shared/src/main/scala/hkmc2/codegen/wasm/text/WatBuilder.scala

Derppening

Apart from some fields that appear to be better located in Ctx than WatBuilder, the rest looks good to me.

Please do add documentation where possible, including private members.

hkmc2/shared/src/main/scala/hkmc2/codegen/wasm/text/Wasm.scala

hkmc2/shared/src/main/scala/hkmc2/codegen/wasm/text/WatBuilder.scala

hkmc2/shared/src/main/scala/hkmc2/codegen/wasm/text/Instructions.scala

Derppening · 2026-01-02T10:31:18Z

hkmc2/shared/src/main/scala/hkmc2/codegen/wasm/text/WatBuilder.scala

+  private case class TupleArrayInfo(arrayType: TypeIdx, elemType: Type, mutable: Bool)
+  private var mutTupleArrayInfo: Opt[TupleArrayInfo] = N
+  private var tupleArrayInfo: Opt[TupleArrayInfo] = N


I don't really understand why this is necessary. Can't we just lookup/cache the Wasm type for tuples from the module?

And, since we assume that all tuples have anyrefs as their element type, why do we need to store elemType and mutable?

I think mutable is needed because we do differentiate mutable and immutable tuples (see the tests).

Also, it doesn't hurt to store elemType. We may later generalize support for more precisely-typed arrays.

I guess I wasn't being clear here - What I meant is why is mutTupleArrayInfo and tupleArrayInfo necessary and stored as separate fields? It appears to me that what we may want is a lookup table for intrinsic types (e.g. Unit, Tuple and mutable Tuple) to lookup its corresponding Wasm type.

While mutTupleArrayInfo and tupleArrayInfo is probably good enough when tuples are stored as anyref arrays as they are now, if (and when) we generalize support for more precisely-typed arrays, these fields probably wouldn't suffice.

Ok now I look up intrinsic types from MutMap declared in Ctx

Derppening · 2026-01-02T10:32:52Z

hkmc2/shared/src/main/scala/hkmc2/codegen/wasm/text/WatBuilder.scala

+    val mutInfo = tupleArray(true)
+    val immInfo = tupleArray(false)
+    val tupleIsMutable = ref.test(tupleExpr, RefType(mutInfo.arrayType, nullable = true))
+    val mutableBranch =
+      val tupleRef = ref.cast(tupleExpr, RefType(mutInfo.arrayType, nullable = false))
+      array.get(mutInfo.arrayType, tupleRef, idxBuilder(tupleRef), mutInfo.elemType)
+    val immutableBranch =
+      val tupleRef = ref.cast(tupleExpr, RefType(immInfo.arrayType, nullable = false))
+      array.get(immInfo.arrayType, tupleRef, idxBuilder(tupleRef), immInfo.elemType)


I thought we established that array.get can be used on both mutable and immutable arrays. Is there any reason why this is necessary?

I need to first cast to use array.get. I think that we can't cast immutable tuple to the mutable tuple type or vice versa.

But can't we just use array.get on both mutable and immutable arrays in Wasm?

Yes but before using array.get we need to cast expressions

Derppening · 2026-01-02T10:33:47Z

hkmc2/shared/src/main/scala/hkmc2/codegen/wasm/text/WatBuilder.scala

+      resultTypes = Seq(Result(mutInfo.elemType.asValType_!))
+    )
+
+  private def tupleIndexBuilder(


Nit: buildTupleIndex - This is an operation, not a class/factory.

hkmc2/shared/src/main/scala/hkmc2/codegen/wasm/text/WatBuilder.scala

…uples

FetBoba added 7 commits December 4, 2025 16:36

Basic tuple literal support

72d510b

Added selection for tuples

03851e1

Merge branch 'hkmc2' of https://github.com/b-temirov/mlscript into tu…

608e71c

…ples

Added negative indices and field assignment for tuples

d823a60

Added tupple pattern matching

39c2dde

Fixed distinction between mutable and immutable tuples

ac7b89b

Merge branch 'hkmc2' of https://github.com/Derppening/mlscript into t…

c40db27

…uples

LPTK reviewed Jan 1, 2026

View reviewed changes

LPTK requested a review from Derppening January 1, 2026 16:40

LPTK and others added 2 commits January 2, 2026 00:40

Merge branch 'hkmc2' into tuples

35aa152

Removed special-case of array access and toString in error expr

329be14

Derppening reviewed Jan 2, 2026

View reviewed changes

FetBoba and others added 3 commits January 6, 2026 13:30

Merge branch 'hkmc2' of https://github.com/Derppening/mlscript into t…

f9d0f18

…uples

Merge branch 'hkmc2' into tuples

1b49bc9

Moved Tupple info to Ctx and deleted unused parts

5aafd8f

Added support for Tuples #374

Are you sure you want to change the base?

Added support for Tuples #374

Uh oh!

Conversation

FetBoba commented Jan 1, 2026

Uh oh!

LPTK left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Derppening left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Derppening Jan 2, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

Derppening Jan 2, 2026 •

edited

Loading