Add deriveEq for Plinth similar to deriving stock Eq #7433

bezirg · 2025-11-19T11:15:58Z

Pre-submit checklist:

Branch
- Tests are provided (if possible)
- Commit sequence broadly makes sense
- Key commits have useful messages
- Changelog fragments have been written (if appropriate)
- Relevant tickets are mentioned in commit messages
- Formatting, PNG optimization, etc. are updated
PR
- (For external contributions) Corresponding issue exists and is linked in the description
- Targeting master unless this is a cherry-pick backport
- Self-reviewed the diff
- Useful pull request description
- Reviewer requested

github-actions · 2025-11-24T10:06:13Z

Execution Budget Golden Diff

f96714c (master) vs 97a7b4b

output

plutus-benchmark/cardano-loans/test/9.6/main.golden.eval

Metric	Old	New	Δ%
Flat Size	`8_666`	`8_673`	+0.08%

plutus-benchmark/coop/test/9.6/certMpBurning.golden.eval

Metric	Old	New	Δ%
Flat Size	`8_031`	`8_035`	+0.05%

plutus-benchmark/coop/test/9.6/certMpMinting.golden.eval

Metric	Old	New	Δ%
Flat Size	`8_554`	`8_558`	+0.05%

plutus-benchmark/coop/test/9.6/fsMpBurning.golden.eval

Metric	Old	New	Δ%
Flat Size	`7_395`	`7_399`	+0.05%

plutus-benchmark/coop/test/9.6/fsMpMinting.golden.eval

Metric	Old	New	Δ%
Flat Size	`9_213`	`9_217`	+0.04%

plutus-benchmark/linear-vesting/test/9.6/main.golden.eval

Metric	Old	New	Δ%
Flat Size	`2_860`	`2_864`	+0.14%

plutus-benchmark/nofib/test/9.6/clausify-F5.golden.eval

Metric	Old	New	Δ%
Flat Size	`1_477`	`1_484`	+0.47%

plutus-benchmark/nofib/test/9.6/knights10-4x4.golden.eval

Metric	Old	New	Δ%
CPU	`1_023_470_754`	`1_025_262_754`	+0.18%
Memory	`5_395_058`	`5_406_258`	+0.21%
Flat Size	`1_674`	`1_682`	+0.48%

plutus-benchmark/nofib/test/9.6/queens4-bt.golden.eval

Metric	Old	New	Δ%
Flat Size	`1_764`	`1_767`	+0.17%

plutus-benchmark/nofib/test/9.6/queens5-fc.golden.eval

Metric	Old	New	Δ%
Flat Size	`1_764`	`1_767`	+0.17%

plutus-benchmark/script-contexts/test/V3/Data/9.6/purposeIsWellFormed-4.golden.eval

Metric	Old	New	Δ%
Flat Size	`1_839`	`1_843`	+0.22%

This comment will get updated when changes are made.

SeungheonOh

I'm not too confident in my ability to review TH code. So perhaps a look from @Unisay would be nice. But everything looks right from what I can tell.

SeungheonOh · 2025-11-26T15:26:58Z

plutus-tx/src/PlutusTx/Eq.hs

+deriveEq ''()
+deriveEq ''(,)
+deriveEq ''(,,)
+deriveEq ''(,,,)
+deriveEq ''(,,,,)
+deriveEq ''(,,,,,)
+deriveEq ''(,,,,,,)
+deriveEq ''(,,,,,,,)
+deriveEq ''(,,,,,,,,)
+deriveEq ''(,,,,,,,,,)
+deriveEq ''(,,,,,,,,,,)
+deriveEq ''(,,,,,,,,,,,)
+deriveEq ''(,,,,,,,,,,,,)
+deriveEq ''(,,,,,,,,,,,,,)
+deriveEq ''(,,,,,,,,,,,,,,)
+deriveEq ''(,,,,,,,,,,,,,,,)
+deriveEq ''(,,,,,,,,,,,,,,,,)
+deriveEq ''(,,,,,,,,,,,,,,,,,)
+deriveEq ''(,,,,,,,,,,,,,,,,,,)
+deriveEq ''(,,,,,,,,,,,,,,,,,,,)
+deriveEq ''(,,,,,,,,,,,,,,,,,,,,)
+deriveEq ''(,,,,,,,,,,,,,,,,,,,,,)
+deriveEq ''(,,,,,,,,,,,,,,,,,,,,,,)
+deriveEq ''(,,,,,,,,,,,,,,,,,,,,,,,)
+deriveEq ''(,,,,,,,,,,,,,,,,,,,,,,,,)
+deriveEq ''(,,,,,,,,,,,,,,,,,,,,,,,,,)
+deriveEq ''(,,,,,,,,,,,,,,,,,,,,,,,,,,)


The great pyramid

plutus-tx/test/Eq/Spec.hs

SeungheonOh · 2025-11-26T15:38:54Z

Also, can you check if this works with polymorphic phantom types? I'm curious as per #4537

plutus-tx/src/PlutusTx/Eq/Class.hs

plutus-tx/src/PlutusTx/Eq/TH.hs

bezirg · 2025-12-01T10:58:06Z

Also, can you check if this works with polymorphic phantom types? I'm curious as per #4537

I will test it out

bezirg · 2025-12-01T11:46:18Z

Also, can you check if this works with polymorphic phantom types? I'm curious as per #4537

Indeed, it does not work with polymorphic phantom types.

I don't know yet how to make it work.
One idea I had is to look at the datatype's kind using https://hackage.haskell.org/package/th-abstraction-0.7.1.0/docs/Language-Haskell-TH-Datatype.html#t:DatatypeInfo .

Unfortunately, it returns me kind the wrong kind *, whereas I was expecting kind forall k. k -> *. If I had the latter kind then I could infer that k is phantom. I think it is too much work , for little benefit. What do you think?

bezirg · 2025-12-15T14:18:34Z

Also, can you check if this works with polymorphic phantom types? I'm curious as per #4537

Phantom types should work now

zliu41

Why not use deriveEq on all the plutus-ledger-api Types?

plutus-tx/src/PlutusTx/Functor.hs

bezirg · 2025-12-15T18:08:13Z

Why not use deriveEq on all the plutus-ledger-api Types?

Good idea

Add some derived Eq instances

bezirg · 2025-12-16T17:00:32Z

plutus-ledger-api/src/PlutusLedgerApi/V1/Data/Value.hs


-instance Haskell.Semigroup Value where
-  (<>) = unionWith (+)
+instance Haskell.Eq Value where


@zliu41 Some cleanup along the way, and use a single implementation between PlutusTx && Haskell

bezirg · 2025-12-16T17:01:14Z

plutus-ledger-api/src/PlutusLedgerApi/V1/Interval.hs

+-- MAYBE: get rid of these and switch to deriving stock, when deriveOrd is merged
+instance Eq a => Haskell.Eq (Extended a) where
+  (==) = (PlutusTx.==)
+
+-- MAYBE: get rid of these and switch to deriving stock, when deriveOrd is merged


This can be changed to stock deriving when deriveOrd PR is merged

bezirg · 2025-12-16T17:01:35Z

plutus-ledger-api/src/PlutusLedgerApi/V1/Value.hs

+instance Haskell.Eq Value where
+  (==) = (PlutusTx.==)

-instance Semigroup Value where
+instance PlutusTx.Semigroup Value where
  {-# INLINEABLE (<>) #-}
  (<>) = unionWith (+)

-instance Haskell.Monoid Value where
-  mempty = Value Map.empty
+instance Haskell.Semigroup Value where
+  (<>) = (PlutusTx.<>)

-instance Monoid Value where
+instance PlutusTx.Monoid Value where
  {-# INLINEABLE mempty #-}
  mempty = Value Map.empty

+instance Haskell.Monoid Value where
+  mempty = PlutusTx.mempty
+


Same cleanup and sharing between implementations

bezirg · 2025-12-16T17:02:39Z

plutus-ledger-api/src/PlutusLedgerApi/V2/Contexts.hs

+-- FIXME: need deriveOrd to be merged to add this
+-- deriveEq ''TxInfo
+


~~This is needed because we have inside the datatype Maps, and keys to Maps need an Ord instance. I think it would be easier if we have the ord pr apply these changes~~

The problem is different, we don't have Eq AssocMap instances because of problem with duplicated entries

bezirg · 2025-12-16T17:03:41Z

@zliu41 Can you take a look at the last commit only?

Unisay · 2025-12-23T15:32:19Z

plutus-tx/src/PlutusTx/Eq/TH.hs

+
+{-| derive a PlutusTx.Eq instance for a datatype/newtype, similar to Haskell's `deriving stock Eq`.
+
+One shortcoming compared to Haskell's deriving is that you cannot `PlutusTx.deriveEq` for polymorphic phantom types. -}


The documentation claims you cannot derive for polymorphic phantom types, but PhantomADT test demonstrates it does work?

Since this is a public function, it would make sense to better document limitations, etc:

{-| Derive a PlutusTx 'Eq' instance for a datatype or newtype. Similar to Haskell's @deriving stock Eq@, this generates structural equality with short-circuit evaluation and INLINEABLE pragmas for optimal on-chain performance. __Usage:__ @ data MyType = Con1 Integer | Con2 Bool deriveEq ''MyType @ __Generated code:__ * Pattern-matching clauses for each constructor * Short-circuit evaluation (stops at first inequality) * @INLINEABLE@ pragma for cross-module optimization * Proper handling of phantom type parameters __Supported types:__ * Regular datatypes with any number of constructors * Newtypes * Types with phantom type parameters * Types with strict or lazy fields __Unsupported:__ * GADTs (may produce type errors) * Type families (not tested) See 'PlutusTx.Eq.Class.Eq' for the class definition. -} deriveEq :: Name -> Q [Dec]

The documentation claims you cannot derive for polymorphic phantom types, but PhantomADT test demonstrates it does work?

Forgot to clean the comment. Phantom types do work now.

Unisay · 2025-12-23T15:36:57Z

plutus-tx/src/PlutusTx/Eq/TH.hs

+    ( clause
+        [conP name (fmap varP argsL), conP name (fmap varP argsR)]
+        ( normalB $
+            case fields of
+              [] -> conE 'True
+              _ ->
+                foldr1 (\e acc -> infixE (pure e) (varE '(&&)) (pure acc)) $
+                  zipWith
+                    ( \argL argR ->
+                        infixE (pure $ varE argL) (varE '(==)) (pure $ varE argR)
+                    )
+                    argsL
+                    argsR
+        )
+        []
+      )


Redundant brackets:

Suggested change

( clause

[conP name (fmap varP argsL), conP name (fmap varP argsR)]

( normalB $

case fields of

[] -> conE 'True

_ ->

foldr1 (\e acc -> infixE (pure e) (varE '(&&)) (pure acc)) $

zipWith

( \argL argR ->

infixE (pure $ varE argL) (varE '(==)) (pure $ varE argR)

)

argsL

argsR

)

[]

)

clause

[conP name (fmap varP argsL), conP name (fmap varP argsR)]

( normalB $

case fields of

[] -> conE 'True

_ ->

foldr1 (\e acc -> infixE (pure e) (varE '(&&)) (pure acc)) $

zipWith

( \argL argR ->

infixE (pure $ varE argL) (varE '(==)) (pure $ varE argR)

)

argsL

argsR

)

[]

Unisay · 2025-12-23T15:44:31Z

plutus-tx/changelog.d/20251125_110335_bezirg_derive_eq.md

+### Added
+
+- A `deriveEq` command to derive PlutusTx.Eq instances for datatypes/newtypes, similar to Haskell's
+  `deriving stock Eq`


Suggested change

### Added

- A `deriveEq` command to derive PlutusTx.Eq instances for datatypes/newtypes, similar to Haskell's

`deriving stock Eq`

### Added

- A `deriveEq` Template Haskell function to automatically derive PlutusTx.Eq instances

for datatypes and newtypes, similar to Haskell's `deriving stock Eq`.

Usage: `PlutusTx.deriveEq ''MyType`

This generates short-circuiting equality checks with proper INLINEABLE pragmas for

optimal on-chain performance.

Unisay · 2025-12-23T15:46:59Z

plutus-tx/test/Eq/Spec.hs

+  runTestNested
+    ["test", "Eq", "Golden"]
+    [ $(goldenCodeGen "SomeLargeADT" (deriveEq ''SomeLargeADT))
+    , $(goldenCodeGen "PhantomADT" (deriveEq ''PhantomADT))


Do you mind adding tests for newtypes and records with fields?

Also, do you know if (self/manually)-recursive types are supported? Would be nice to have a comment (if not) or tests (if yes)

Unisay · 2025-12-23T15:51:05Z

plutus-tx/src/PlutusTx/Ratio.hs

-
 instance P.Ord Rational where
-  {-# INLINEABLE compare #-}
-  compare (Rational n d) (Rational n' d') = P.compare (n P.* d') (n' P.* d)


I wonder why is compare the only function removed here?

I will take a look at this

bezirg changed the title ~~Bezirg/derive eq~~ Add deriveEq for Plinth similar to deriving stock Eq Nov 19, 2025

bezirg force-pushed the bezirg/derive_eq branch from 1bdeaab to 956de09 Compare November 24, 2025 10:05

bezirg marked this pull request as ready for review November 24, 2025 10:07

bezirg marked this pull request as draft November 24, 2025 10:07

bezirg force-pushed the bezirg/derive_eq branch from 956de09 to d666dea Compare November 24, 2025 10:08

bezirg self-assigned this Nov 24, 2025

bezirg force-pushed the bezirg/derive_eq branch 2 times, most recently from 5b30be7 to 5d931a0 Compare November 25, 2025 10:05

bezirg marked this pull request as ready for review November 25, 2025 10:05

bezirg requested review from SeungheonOh and Unisay and removed request for SeungheonOh November 25, 2025 10:05

bezirg force-pushed the bezirg/derive_eq branch from 5d931a0 to 396e4e9 Compare November 25, 2025 13:39

bezirg marked this pull request as draft November 25, 2025 13:42

bezirg force-pushed the bezirg/derive_eq branch from 396e4e9 to 97a7b4b Compare November 25, 2025 13:50

bezirg marked this pull request as ready for review November 25, 2025 13:50

bezirg force-pushed the bezirg/derive_eq branch 5 times, most recently from 7593c53 to 3ca2a2e Compare November 26, 2025 11:18

SeungheonOh reviewed Nov 26, 2025

View reviewed changes

Unisay reviewed Dec 1, 2025

View reviewed changes

plutus-tx/src/PlutusTx/Eq/Class.hs Outdated Show resolved Hide resolved

Unisay reviewed Dec 1, 2025

View reviewed changes

plutus-tx/src/PlutusTx/Eq/Class.hs Outdated Show resolved Hide resolved

Unisay reviewed Dec 1, 2025

View reviewed changes

plutus-tx/src/PlutusTx/Eq/TH.hs Show resolved Hide resolved

bezirg force-pushed the bezirg/derive_eq branch from 770689f to 5e72678 Compare December 4, 2025 16:51

bezirg requested a review from zliu41 December 4, 2025 16:52

bezirg force-pushed the bezirg/derive_eq branch 3 times, most recently from 673b24b to f6a08c0 Compare December 15, 2025 14:00

zliu41 reviewed Dec 15, 2025

View reviewed changes

plutus-tx/src/PlutusTx/Functor.hs Show resolved Hide resolved

bezirg added 8 commits December 16, 2025 15:55

Add deriveEq for Plinth similar to deriving stock Eq

2748136

Add some derived Eq instances

Address come comments

0bf9bd7

fourmolu

b35dd05

Added golden th

0b4fc7d

Fixed deriveEq ''Void , removed redundant TH generated clause

2cf1d99

Added deriveEq for Rational

49a7e61

cabal-fmt

d1f0a1c

Make deriving eq phantom types work

ce324dd

bezirg force-pushed the bezirg/derive_eq branch 2 times, most recently from d9a2637 to d5851d7 Compare December 16, 2025 16:58

bezirg requested a review from zliu41 December 16, 2025 16:59

bezirg commented Dec 16, 2025

View reviewed changes

bezirg force-pushed the bezirg/derive_eq branch 2 times, most recently from dd788a1 to ce324dd Compare December 17, 2025 11:24

Add deriveEq for plutus-ledger-api

a00cb11

bezirg force-pushed the bezirg/derive_eq branch from dd788a1 to a00cb11 Compare December 18, 2025 09:17

bezirg requested a review from a team December 20, 2025 13:52

Unisay reviewed Dec 23, 2025

View reviewed changes

		-- FIXME: need deriveOrd to be merged to add this
		-- deriveEq ''TxInfo


		{-\| derive a PlutusTx.Eq instance for a datatype/newtype, similar to Haskell's `deriving stock Eq`.

		One shortcoming compared to Haskell's deriving is that you cannot `PlutusTx.deriveEq` for polymorphic phantom types. -}

-### Added
-- A `deriveEq` command to derive PlutusTx.Eq instances for datatypes/newtypes, similar to Haskell's
-  `deriving stock Eq`
+### Added
+- A `deriveEq` Template Haskell function to automatically derive PlutusTx.Eq instances
+  for datatypes and newtypes, similar to Haskell's `deriving stock Eq`.
+  Usage: `PlutusTx.deriveEq ''MyType`
+  This generates short-circuiting equality checks with proper INLINEABLE pragmas for
+  optimal on-chain performance.

Add deriveEq for Plinth similar to deriving stock Eq #7433

Are you sure you want to change the base?

Add deriveEq for Plinth similar to deriving stock Eq #7433

Conversation

bezirg commented Nov 19, 2025

Uh oh!

github-actions bot commented Nov 24, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Execution Budget Golden Diff

plutus-benchmark/cardano-loans/test/9.6/main.golden.eval

plutus-benchmark/coop/test/9.6/certMpBurning.golden.eval

plutus-benchmark/coop/test/9.6/certMpMinting.golden.eval

plutus-benchmark/coop/test/9.6/fsMpBurning.golden.eval

plutus-benchmark/coop/test/9.6/fsMpMinting.golden.eval

plutus-benchmark/linear-vesting/test/9.6/main.golden.eval

plutus-benchmark/nofib/test/9.6/clausify-F5.golden.eval

plutus-benchmark/nofib/test/9.6/knights10-4x4.golden.eval

plutus-benchmark/nofib/test/9.6/queens4-bt.golden.eval

plutus-benchmark/nofib/test/9.6/queens5-fc.golden.eval

plutus-benchmark/script-contexts/test/V3/Data/9.6/purposeIsWellFormed-4.golden.eval

Uh oh!

SeungheonOh left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

SeungheonOh commented Nov 26, 2025

Uh oh!

Uh oh!

Uh oh!

Uh oh!

bezirg commented Dec 1, 2025

Uh oh!

bezirg commented Dec 1, 2025

Uh oh!

bezirg commented Dec 15, 2025

Uh oh!

zliu41 left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

bezirg commented Dec 15, 2025

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!

bezirg Dec 16, 2025 • 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!

bezirg commented Dec 16, 2025

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

bezirg Dec 23, 2025 • 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!

bezirg Dec 23, 2025 • 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!

github-actions bot commented Nov 24, 2025 •

edited

Loading

bezirg Dec 16, 2025 •

edited

Loading

bezirg Dec 23, 2025 •

edited

Loading

bezirg Dec 23, 2025 •

edited

Loading