Compare commits

..

43 Commits

Author SHA1 Message Date
Leonardo de Moura
fba3d7443b checkpoint 2026-04-08 21:50:02 -07:00
Leonardo de Moura
fc4cf2ac1d checkpoint 2026-04-08 19:27:38 -07:00
Sebastian Graf
f9b2f6b597 fix: use getDecLevel/isLevelDefEq for for loop mut var universe constraints (#13332)
This PR fixes universe unification for `for` loops with `mut` variables
whose types span multiple implicit universes. The old approach used
`ensureHasType (mkSort mi.u.succ)` per variable, which generated
constraints like `max (?u+1) (?v+1) =?= ?u+1` that the universe solver
cannot decompose. The new approach uses `getDecLevel`/`isLevelDefEq` on
the decremented level, producing `max ?u ?v =?= ?u` which `solveSelfMax`
handles directly.

Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
2026-04-08 16:34:51 +00:00
Sebastian Ullrich
a3cc301de5 fix: wrapInstance should not reduce non-constructor instances (#13327)
This otherwise can break `Decidable` instances
2026-04-08 16:31:28 +00:00
Lean stage0 autoupdater
3a8db01ce8 chore: update stage0 2026-04-08 15:28:03 +00:00
Joachim Breitner
06fb4bec52 feat: require indentation in commands, allow empty tactic sequences (#13229)
This PR wraps the top-level command parser with `withPosition` to
enforce indentation in `by` blocks, combined with an empty-by fallback
for better error messages.

This subsumes #3215 (which introduced `withPosition commandParser` but
without the empty-by fallback). It is also related to #9524, which
explores elaboration with empty tactic sequences — this PR reuses that
idea for the empty-by fallback, so that a `by` not followed by an
indented tactic produces an elaboration error (unsolved goals) rather
than a parse error.

**Changes:**
- `topLevelCommandParserFn` now uses `(withPosition commandParser).fn`,
setting the saved position at the start of each top-level command
- `tacticSeqIndentGt` gains an empty tactic sequence fallback
(`pushNone`) so that missing indentation produces an elaboration error
(unsolved goals) instead of a parse error
- `isEmptyBy` in `goalsAt?` removed: with strict `by` indentation, empty
`by` blocks parse successfully via `pushNone` (producing empty nodes)
rather than producing `.missing` syntax, making the `isEmptyBy` check
dead code. The `isEmpty` helper in `isSyntheticTacticCompletion`
continues to work correctly because it handles both `.missing` and empty
nodes from `pushNone` (via the vacuously-true `args.all isEmpty` on
`#[]`)
- Test files updated to indent `by` blocks and expression continuations
that were previously at column 0

**Behavior:**
- Top-level `by` blocks now require indentation (column > 0 for commands
at column 0)
- Commands indented inside `section` require proofs to be indented past
the command's column
- `#guard_msgs in example : True := by` works because tactic indentation
is checked against the outermost command's column
- Expression continuations (not just `by`) must also be indented past
the command, which is slightly more strict but more consistent
- `have : True := by` followed by a dedent now correctly puts `this` in
scope in the outer tactic block (the `have` is structurally complete
with an unsolved-goal error, rather than a parse error)

**Code changes observed in practice (lean4 test suite + Mathlib):**

- `by` blocks: top-level `theorem ... := by` / `decreasing_by` followed
by tactics at column 0 must be indented
- `variable` continuations: `variable {A : Type*} [Foo A]\n{B : Type*}`
where the second line starts at column 0 must be indented (most common
category in Mathlib)
- Expression continuations: `def f : T :=\nexpr` or `#synth Foo\n[args]`
where the body/arguments start at column 0
- Structure literals: `.symm\n{ toFun := ...` where the struct literal
starts at column 0

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
2026-04-08 14:05:47 +00:00
Sebastian Ullrich
35b4c7dbfc feat: implicit public meta import Init in non-prelude files (#13323)
Ensure metaprograms have implicit access to `Init` like everyone else.
Closes #13310.
2026-04-08 11:46:46 +00:00
Joachim Breitner
2398d2cc66 feat: no [defeq] attribute on sizeOf_spec lemmas (#13320)
This PR changes the auto-generated `sizeOf` definitions to be not
exposed and the `sizeOf_spec` theorem to be not marked `[defeq]`.
2026-04-08 11:10:50 +00:00
Kim Morrison
8353964e55 feat: wire PowIdentity into grind ring solver (#13088)
This PR wires the `PowIdentity` typeclass (from
https://github.com/leanprover/lean4/pull/13086) into the `grind` ring
solver's Groebner basis engine.

When a ring has a `PowIdentity α p` instance, the solver pushes `x ^ p =
x` as a new fact for each variable `x`, which becomes `x^p - x = 0` in
the Groebner basis. Since `p` is an `outParam`, instance discovery is
decoupled from `IsCharP` — the solver synthesizes `PowIdentity α ?p`
with a fresh metavar and lets instance search find both the instance and
the exponent.

This correctly handles non-prime finite fields: for `F_4` (char 2, 4
elements), Mathlib would provide `PowIdentity F_4 4` and the solver
would discover `p = 4`, not `p = 2`.

Note: the original motivating example `(x + y)^2 = x^128 + y^2` from
https://github.com/leanprover/lean4/issues/12842 does not yet work
because the `ToInt` module lifts `Fin 2` expressions to integers and
expands `x^128` via the binomial theorem before the ring solver can
reduce it. Addressing that is a separate deeper change.

🤖 Prepared with Claude Code

---------

Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-04-08 10:14:10 +00:00
Kim Morrison
334d9bd4f3 feat: add markMeta parameter to addAndCompile (#13311)
This PR adds an optional `markMeta : Bool := false` parameter to
`addAndCompile`, so that callers can propagate the `meta` marking
without manually splitting into `addDecl` + `markMeta` + `compileDecl`.

Also updates `ParserCompiler` to use the new parameter.

🤖 Prepared with Claude Code

---------

Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-04-08 09:09:01 +00:00
Kim Morrison
f7f5fc5ecd feat: add PowIdentity typeclass for grind ring solver (#13086)
This PR adds a `Lean.Grind.PowIdentity` typeclass stating that `x ^ p =
x` for all elements of a commutative semiring, with `p` as an
`outParam`.

The primary source of instances is Fermat's little theorem: for a finite
field with `q` elements, `x ^ q = x`. Since `p` is an `outParam`,
instance synthesis discovers the correct exponent automatically — the
solver does not need to know the characteristic or cardinality in
advance.

A concrete instance for `Fin 2` is provided. Mathlib can provide
instances for general finite fields via `FiniteField.pow_card`.

🤖 Prepared with Claude Code

Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-04-08 09:05:12 +00:00
Joachim Breitner
659db85510 fix: suggest (rfl) not id rfl in linter (#13319)
This PR amends #13317 to suggest `:= (rfl)` as the recommended way to
avoid a theorem to be automatically marked `[defeq]`, for consistency
with existing documentation. Rationale: the special treatment of `:=
rfl` is based on syntax, not the proof term, so it’s appropriate to use
different syntax. And also I like the way it reads like a “muted whisper
of `rfl`”.
2026-04-08 08:21:23 +00:00
Wojciech Różowski
91dd99165a feat: add warning when applying global attribute using in (#13223)
This PR adds a warning preventing a user from applying global attribute
using `... in ...`, e.g.
```lean4
theorem a : True := trivial
attribute [simp] a in
def b : True := a
```
2026-04-08 06:20:34 +00:00
Lean stage0 autoupdater
e44351add9 chore: update stage0 2026-04-08 05:43:47 +00:00
Leonardo de Moura
fd2723d9c0 feat: add linter for rfl simp theorems at restricted transparency (#13317)
This PR adds an opt-in linter (`set_option simp.rfl.checkTransparency
true`) that warns when a `rfl` simp theorem's LHS and RHS are not
definitionally equal at `.instances` transparency. Bad rfl-simp theorems
— those that only hold at higher transparency — create problems
throughout the system because `simp` and `dsimp` operate at restricted
transparency. The linter suggests two fixes: use `id rfl` as the proof
(to remove the `rfl` status), or mark relevant constants as
`[implicit_reducible]`.

This is part of a broader effort to ensure `isDefEq` respects
transparency levels. The linter helps systematically identify
problematic rfl-simp theorems so they can be fixed incrementally.

Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
2026-04-08 04:49:07 +00:00
Kim Morrison
ad2105dc94 feat: add List.prod, Array.prod, and Vector.prod (#13200)
This PR adds `prod` (multiplicative fold) for `List`, `Array`, and
`Vector`, mirroring the existing `sum` API. Includes basic simp lemmas
(`prod_nil`, `prod_cons`, `prod_append`, `prod_singleton`,
`prod_reverse`, `prod_push`, `prod_eq_foldl`), Nat-specialized lemmas
(`prod_pos_iff_forall_pos_nat`, `prod_eq_zero_iff_exists_zero_nat`,
`prod_replicate_nat`), Int-specialized lemmas (`prod_replicate_int`),
cross-type lemmas (`prod_toArray`, `prod_toList`), and `Perm.prod_nat`
with grind patterns.

The min/max pigeonhole-style bounds from the `sum` Nat/Int files are
omitted as they don't have natural multiplicative analogues.

🤖 Prepared with Claude Code

Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-04-08 04:01:49 +00:00
Mac Malone
235aedfaf7 chore: use --wfail for core CI build (#13294)
This PR introduces the `WFAIL` CMake setting that uses `--wfail` for
Lake builds of core.

This also, by extension, fixes a mismatch in the trace of core CI builds
vs local core builds.
2026-04-08 02:17:39 +00:00
Kim Morrison
30dca7b545 fix: make delta-derived Prop-valued instances theorems (#13304)
This PR makes the delta-deriving handler create `theorem` declarations
instead of `def` declarations when the instance type is a `Prop`.
Previously, `deriving instance Nonempty for Foo` would always create a
`def`, which is inconsistent with the behavior of a handwritten
`instance` declaration.

For example, given:
```lean
def Foo (α : Type u) := List α
deriving instance Nonempty for Foo
```

Before: `@[implicit_reducible] def instNonemptyFoo ...`
After: `@[implicit_reducible] theorem instNonemptyFoo ...`

The implementation checks `isProp result.type` after constructing the
instance closure, and uses `mkThmOrUnsafeDef` for the Prop case (which
also handles the unsafe fallback correctly). The noncomputable check is
skipped for Prop-typed instances since theorems can freely reference
noncomputable constants.

Closes #13295

🤖 Prepared with Claude Code

---------

Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-04-08 01:19:48 +00:00
Kim Morrison
7e04970c58 fix: skip nightly-testing merge when branch does not exist (#13308)
This PR makes `release_steps.py` robust to release repos that have no
`nightly-testing` branch. Previously, `git merge origin/nightly-testing`
would fail with "not something we can merge" and the error handler
misinterpreted this as a merge conflict, then crashed trying to commit
with nothing to commit. Now we check for the branch with
`git ls-remote --heads` before attempting the merge.

🤖 Prepared with Claude Code

---------

Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-04-08 00:31:22 +00:00
Kim Morrison
0a6ee838df fix: update lean-toolchain in verso test-projects during release (#13309)
This PR updates `release_steps.py` to sync all `lean-toolchain` files in
verso's `test-projects/` subdirectories to match the root toolchain
during release bumps. The verso CI "SubVerso version consistency" check
requires these to match, and the script was only syncing
`lake-manifest.json` sub-manifests but not toolchain files.

🤖 Prepared with Claude Code

Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-04-08 00:29:38 +00:00
Sebastian Graf
ec72785927 chore: git ignore .claude/worktrees (#13299)
This PR ignores `.claude/worktrees` after I have had one instance too
many where I nuked all my worktrees due to a brain-dead `git clean
-fxd`.
2026-04-07 09:51:39 +00:00
Keith Seim
ba33c3daa4 fix: match stub signature of lean_uv_dns_get_info to real implementation (#13234)
This PR fixes a build issue when Lean is not linked against libuv.

## Problem

In `src/runtime/uv/dns.cpp`, the non-libuv stub of
`lean_uv_dns_get_info` (in the `#else` branch, compiled when building
without libuv) has a **4-parameter** signature:

```cpp
lean_uv_dns_get_info(b_obj_arg name, b_obj_arg service, uint8_t family, int8_t protocol)
```

But the real implementation above the `#else` has only **3 parameters**:

```cpp
lean_uv_dns_get_info(b_obj_arg name, b_obj_arg service, uint8_t family)
```

The Lean `@[extern]` declaration also expects 3 parameters. The stub has
an extra `int8_t protocol` parameter that the real function and the Lean
FFI caller do not use.

## Fix

Remove the extra `protocol` parameter from the stub so both branches
have the same signature.

## Evidence

Discovered while building Lean4 to WASM via Emscripten for a production
project ([specify-lean](https://github.com/kjsdesigns/specify)) since
v4.27.0. The stub branch is compiled in this configuration, and the
signature mismatch was caught at link time. The fix has been stable in
production across multiple Lean version bumps.

Related: [Zulip thread on WASM build
fixes](https://leanprover.zulipchat.com/#narrow/channel/270676-lean4/topic/WASM.20build.20fixes.3A.20libuv.20symbol.20leakage.20.28.236817.29.20and.20unique_lo/with/580836892)
(2026-03-21).

Co-authored-by: Keith Seim <keith@MacBook-Pro.local>
2026-04-07 09:39:53 +00:00
Keith Seim
db1e2ac34c fix: add missing release() and adopt_lock_t to single-threaded unique_lock stub (#13233)
This PR fixes runtime build issues when `LEAN_MULTI_THREAD` is not set.

## Problem

When building with `LEAN_MULTI_THREAD` undefined (required for
Emscripten/WASM targets), the stub `unique_lock<T>` in
`src/runtime/thread.h` is missing two members that the real
`std::unique_lock` provides:

1. **`release()`** — called by runtime code paths, causes a compile
error when the stub is active
2. **`unique_lock(T const &, std::adopt_lock_t)`** — required by code
that acquires a lock before constructing the `unique_lock`

The other stubs in this file (`mutex`, `lock_guard`,
`condition_variable`) are complete; only `unique_lock` is missing API
surface.

## Fix

Add the two missing members to the single-threaded `unique_lock` stub:

```cpp
unique_lock(T const &, std::adopt_lock_t) {}
T * release() { return nullptr; }
```

Both are no-ops, matching the semantics of a single-threaded
environment. `release()` returns `nullptr` (no mutex to release). The
`adopt_lock_t` constructor is a no-op (no lock to adopt).

## Evidence

I've been using this fix in a production project
([specify-lean](https://github.com/kjsdesigns/specify)) since v4.27.0 to
build the Lean4 runtime to WASM via Emscripten. The fix has been stable
across multiple Lean version bumps.

I posted about this on
[Zulip](https://leanprover.zulipchat.com/#narrow/channel/270676-lean4/topic/WASM.20build.20fixes.3A.20libuv.20symbol.20leakage.20.28.236817.29.20and.20unique_lo/with/580836892)
on 2026-03-21.

Co-authored-by: Keith Seim <keith@MacBook-Pro.local>
2026-04-07 09:30:13 +00:00
Jason Yuen
cb06946972 chore: fix typo in annotatedBorrows (#13276)
Fixes a public-facing typo from #13274.
2026-04-07 09:17:26 +00:00
Sebastian Ullrich
4f6bcc5ada chore: avoid segfault in stage2 (#13296)
Lake.Load must not define metaprograms under the current `--plugin`
setup
2026-04-07 09:08:04 +00:00
Jason Yuen
0650cbe0fa chore: fix typo in isInvalidContinuationByte (#13275)
Fixes a public-facing typo from #13274.
2026-04-07 09:01:19 +00:00
Kim Morrison
8bb07f336d chore: add leansqlite to release repos (#13293)
This PR adds `leansqlite` to `release_repos.yml` and lists it as a
dependency of `doc-gen4`, which requires it.

🤖 Prepared with Claude Code

Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-04-07 08:54:42 +00:00
Mac Malone
c16e88644c feat: Runtime.hold (#13270)
This PR adds `Runtime.hold`, which ensures its argument remains alive
until the callsite by holding a reference to it. This can be useful for
unsafe code (such as an FFI) that relies on a Lean object not being
freed until after some point in the program.

It is implemented via a `static inline` function in `lean.h` so that C
will optimize away the call.

---------

Co-authored-by: Henrik Böving <hargonix@gmail.com>
2026-04-07 08:44:35 +00:00
Mac Malone
96d502bd11 refactor: introduce LakefileConfig & well-formed workspaces (#13282)
This PR introduces `LakefileConfig`, which can be constructed from a
Lake configuration file without all the information required to
construct a full `Package`. Also, workspaces now have a well-formedness
property attached which ensures the workspace indices of its packages
match their index in the workspace. Finally, the facet configuration map
now has its own type: `FacetConfigMap`.

Created by splitting off the major self-contained (but overlapping)
refactors from #11662.
2026-04-06 23:34:48 +00:00
Joachim Breitner
d48863fc2b fix: add checkInterrupted to Core.withIncRecDepth (#13290)
This PR adds a `checkInterrupted` call to `Core.withIncRecDepth`, so
that all recursive CoreM-based operations (inferType, whnf, isDefEq,
simp, …) check for cancellation on each recursion step. Previously,
these operations could run for seconds without responding to IDE
cancellation requests.

This is the single highest-impact change for IDE cancellation
responsiveness. It covers all recursive MetaM/TacticM operations at
once, eliminating the vast majority of multi-second gaps identified by
the `LEAN_CHECK_SYSTEM_INTERVAL_MS` monitoring in #13212.

Gap measurements pending — will be added after CI benchmarks and
instrumented measurement.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
2026-04-06 13:45:30 +00:00
Joachim Breitner
c4a664eb5d fix: add checkSystem to LCNF compiler passes (#13231)
This PR adds checkSystem calls to the LCNF compiler to improve IDE
cancellation responsiveness during compilation of large declarations.

Two changes:
1. `Pass.mkPerDeclaration` now calls `checkSystem` before processing
each declaration, allowing interruption between declarations.
2. The LCNF `simp` pass calls `checkSystem` every 128 recursive visits,
allowing interruption within large single-declaration simplifications.

Note: LCNF simp has its own custom `withIncRecDepth` (in `SimpM.lean`)
that does **not** call `checkInterrupted`, unlike
`Core.withIncRecDepth`. So this `checkSystem` call is the only
cancellation check on this code path.

Performance: unconditional `checkSystem` added +0.44% instructions on
`big_do.lean` per CI benchmarks. Amortizing to every 128 visits brings
overhead to noise level while keeping the max simp gap under ~80M
instructions (~14ms at 6 Ginstr/s).

**Before** (measured with `LEAN_CHECK_SYSTEM_INTERVAL_INSN` on the
instrumentation branch):

| Test | Largest LCNF gap | Time at 6 Ginstr/s |
|------|------------------|-------------------|
| `big_do.lean` | 12,623M insn (simp) | 2.1s |
| `riscv-ast.lean` | 1,162M insn (simp) | 194ms |
| `riscv-ast.lean` | 37 gaps total | — |

**After:**

| Test | Largest LCNF gap | Time at 6 Ginstr/s |
|------|------------------|-------------------|
| `big_do.lean` | 869M insn (resetReuse) | 145ms |
| `riscv-ast.lean` | 232M insn (simp) | 39ms |
| `riscv-ast.lean` | reduced gap count | — |

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
2026-04-06 09:35:58 +00:00
Leonardo de Moura
0cd6dbaad2 feat: add Sym.Arith infrastructure for arithmetic normalization (#13289)
This PR adds the shared infrastructure for arithmetic normalization in
`Sym.Arith/`,
laying the groundwork for both `Sym.simp`'s arith pre-simproc and the
eventual
unification of grind's `CommRing` module.

The key components:

- **`Types.lean`**: Classification structures (`Semiring`, `Ring`,
`CommRing`,
`CommSemiring`) stored in a `SymExtension`. These are the
classification-only
subset of grind's ring types — no solver state. Includes
`withExpThreshold` for
  controlling exponent evaluation limits.

- **`EvalNum.lean`**: `evalNat?`/`evalInt?` for evaluating ground
Nat/Int
expressions in type classes (e.g., `IsCharP`), adapted from grind to
`SymM`.

- **`Classify.lean`**: Algebraic structure detection (CommRing > Ring >
CommSemiring > Semiring) with a single `typeClassify : PHashMap ExprPtr
ClassifyResult` cache.
  Detects `IsCharP`, `NoNatZeroDivisors`, and `Field` instances.

- **Type classes**: `MonadCanon`, `MonadRing`/`MonadCommRing`,
`MonadSemiring`/`MonadCommSemiring`, `MonadGetVar`/`MonadMkVar` —
abstract over
the monad so the same code works in both `SymM` and grind's
`RingM`/`SemiringM`.
Grind's `MonadCanon` is deleted; grind's monads inherit it from `SymM`
via
  `MonadLift`.

- **`Functions.lean`**: Cached function getters (`getAddFn`, `getMulFn`,
etc.)
generic over the type classes. Synthesizes instances, validates via
`isDefEqI`,
  canonicalizes via `canonExpr`.

- **`Reify.lean`**: Converts Lean expressions into
`RingExpr`/`SemiringExpr` for
reflection-based normalization. Variable creation abstracted via
`MonadMkVar`.

- **`DenoteExpr.lean`**: Converts reified expressions back to Lean
`Expr`s.
Roundtrip tests confirm reify→denote produces definitionally equal
results.

- **`ToExpr.lean`**, **`VarRename.lean`**, **`Poly.lean`**: Moved from
`Grind.Arith.CommRing/` — pure utilities on `Grind.CommRing` types with
no
  solver dependencies.

- **Tests**: Unit tests for classification (`Int` → commRing, `Nat` →
commSemiring,
`Rat` → commRing), `evalNat?`/`evalInt?`, exp threshold, and
reify-denote roundtrips.
  
  
**TODO**: use abstractions to implement `grind` ring module, and delete
code duplication.

---------

Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-04-06 05:21:09 +00:00
Joachim Breitner
34d00cb50d feat: add checkSystem calls to bv_decide for cancellation responsiveness (#13284)
This PR adds `checkSystem` calls to the `bv_decide` tactic's main loops
to improve IDE cancellation responsiveness. Previously, `bv_decide` had
zero `checkSystem` calls in its entire codebase, meaning long-running
invocations could not be cancelled promptly.

Insertion points:
- `fixpointPipeline` (normalization fixpoint loop)
- `ReifiedBVExpr.of.go` (recursive BitVec expression reification)
- `ReifiedBVLogical.of.go` (recursive boolean expression reification)
- `reflectBV` (hypothesis loop)

These cover the MetaM-based phases of `bv_decide`. The pure computation
phases (bitblasting via `IO.lazyPure`, AIG→CNF conversion) remain
without checks and would require restructuring to address.

Gap measurements (at 6.0 Ginstr/s, using
`LEAN_CHECK_SYSTEM_INTERVAL_INSN` from #13212):
- `bv_decide_large_aig.lean`: before ~3.3s max gap (19,899M insn),
after: pure bitblast/AIG gaps remain but reification and normalization
phases are now interruptible
- Other bv_decide bench files: normalization and reification loops are
now responsive to cancellation

Note: the largest remaining gaps in bv_decide benchmarks are in the
kernel type checker (proof term checking, up to 10.5s on
`bv_decide_mod.lean`) and in pure bitblasting — both out of scope for
this PR.

Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
2026-04-05 13:11:39 +00:00
Tom Levy
a73be70607 doc: fix typo in doc of Functor.mapConst (#13285)
This PR fixes a typo.

---

Additional comment:

I found the description hard to read (in part because the term "constant
function" is not one I encounter frequently enough, and because there is
no explicit constant function in the signature). Would you consider
changing the first sentence from "Mapping a constant function" to
"Replaces the value in a functor with a constant, retaining the
functor's structure" (based on Functor.discard)?

Also, I would write `(fun _ => a)` rather than `Function.const _ a`.

---------

Co-authored-by: Joachim Breitner <mail@joachim-breitner.de>
2026-04-05 09:16:49 +00:00
Lean stage0 autoupdater
3d49476058 chore: update stage0 2026-04-05 00:39:54 +00:00
Leonardo de Moura
adc45d7c7b feat: mark exposed match auxiliary declarations as implicit_reducible (#13281)
This PR marks any exposed (non-private) auxiliary match declaration as
`[implicit_reducible]`. This is essential when the outer declaration is
marked as `instance_reducible` — without it, reduction is blocked at the
match auxiliary. We do not inherit the attribute from the parent
declaration because match auxiliary declarations are reused across
definitions, and the reducibility setting of the parent can change
independently. This change prepares for implementing the TODO at
`ExprDefEq.lean:465`, which would otherwise cause too many failures
requiring manual `[implicit_reducible]` annotations on match
declarations whose names are not necessarily derived from the outer
function.

---------

Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
2026-04-04 23:55:47 +00:00
Jason Yuen
9efba691e7 chore: fix typo in checkArtifactsExist (#13277)
This PR fixes a public-facing typo in a function name:
`Module.checkArtifactsExsist` -> `Module.checkArtifactsExist`.
2026-04-04 18:56:10 +00:00
Lean stage0 autoupdater
681856324f chore: update stage0 2026-04-04 18:50:11 +00:00
Leonardo de Moura
9f49ea63e2 feat: add backward.isDefEq.respectTransparency.types option (#13280)
This PR adds a new option `backward.isDefEq.respectTransparency.types`
that controls the transparency used when checking whether the type of a
metavariable matches the type of the term being assigned to it during
`checkTypesAndAssign`. Previously, this check always bumped transparency
to `.default` (via `withInferTypeConfig`), which is overly permissive.
The new option uses `.instances` transparency instead (via
`withImplicitConfig`), matching the behavior already used for implicit
arguments.

The option defaults to `false` (preserving old behavior) until stage0 is
updated and breakage is assessed. If
`backward.isDefEq.respectTransparency` (already in v4.29) is set to
`false`, then `backward.isDefEq.respectTransparency.types` is
automatically treated as `false` too.

When `diagnostics` is enabled, a trace message is emitted if the
stricter transparency fails but `.default` would have succeeded, helping
identify affected code. To investigate failures when enabling
`backward.isDefEq.respectTransparency.types`, use:

```
set_option diagnostics true
set_option trace.diagnostics true
```

Also renames `withInstanceConfig` to `withImplicitConfig` since it now
serves implicit argument and type checking, not just instances.
Registers the `diagnostics` trace class in `CoreM`.

Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
2026-04-04 18:05:33 +00:00
Jason Yuen
3770b3dcb8 chore: fix spelling errors (#13274)
This PR fixed typos:

```
pip install codespell --upgrade
codespell --summary --ignore-words-list enew,forin,fro,happend,hge,ihs,iterm,spred --skip stage0 --check-filenames
codespell --summary --ignore-words-list enew,forin,fro,happend,hge,ihs,iterm,spred --skip stage0 --check-filenames --regex '[A-Z][a-z]*'
codespell --summary --ignore-words-list enew,forin,fro,happend,hge,ihs,iterm,spred --skip stage0 --check-filenames --regex "\b[a-z']*"
```
2026-04-04 07:34:34 +00:00
Leonardo de Moura
3c6ea49d0e feat: add mkAppNS, mkAppRevS, betaRevS, betaS, and related Sym functions (#13273)
This PR adds a comprehensive public API for constructing maximally
shared
expression applications and performing beta reduction in the `Sym`
framework.
These functions were previously defined locally in the VC generator and
cbv
tactic, and are needed by downstream `SymM`-based tools.

New functions in `Lean.Meta.Sym.Internal` (generic over
`MonadShareCommon`):
- `mkAppS₆` through `mkAppS₁₁` (higher-arity application builders)
- `mkAppRangeS`, `mkAppNS` (forward application over arrays/ranges)
- `mkAppRevRangeS`, `mkAppRevS` (reversed application over
arrays/ranges)

New public functions in `Lean.Meta.Sym` (`SymM`):
- `betaRevS` and `betaS` (beta reduction with max sharing)
- `mkForallFVarsS` (forall abstraction with max sharing)

The `AlphaShareBuilderM`-specific `mkAppRevRangeS` in
`InstantiateS.lean` is
replaced by the generic version from `Internal`, and the internal
`betaRevS`
is renamed to `betaRevS'`. The `Cbv.mkAppNS` now delegates to
`Internal.mkAppNS`.

---------

Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-04-04 02:31:16 +00:00
Leonardo de Moura
608e0d06a8 fix: extend sym canonicalizer reductions to value positions (#13272)
This PR extends the sym canonicalizer to apply reductions (projection,
match/ite/cond, Nat
arithmetic) in all positions, not just inside types. Previously, a value
`v` appearing in a
type `T(v)` could remain unreduced while `T(v)` was normalized, breaking
the invariant that
definitionally equal types are structurally identical after
canonicalization.

Changes:
- Remove `insideType` guards from `canonApp` and `canonProj`, so
reductions apply unconditionally
(eta reduction remains type-only, to preserve lambda structure for
`grind`)
- Add `canonInstDecCore` to handle `Decidable` instances in
`if-then-else` expressions, dispatching
`Grind.nestedDecidable` to `canonInstDec` and falling back silently for
other instances
- Add `report` parameter to `canonInstCore`/`canonInst'`/`canonInst` to
allow suppressing issue
reporting for propositional and decidable instances that cannot be
resynthesized (common with
`haveI`-provided instances that propagate into types through forward
dependencies)
- Update module documentation to reflect the new reduction scope and the
`haveI` reporting tradeoff

Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
2026-04-04 01:52:24 +00:00
Leonardo de Moura
5fdeaf0d5a fix: handle propositional and decidable instances in sym canonicalizer (#13271)
This PR refactors instance canonicalization in the sym canonicalizer to
properly handle
\`Grind.nestedProof\` and \`Grind.nestedDecidable\` markers. Previously,
the canonicalizer
would report an issue when it failed to resynthesize propositional
instances that were
provided by \`grind\` itself or by the user via \`haveI\`. Now,
resynthesis failure gracefully
falls back to the original instance in value positions, while remaining
strict inside types.

Changes:
- Extract \`canonInstCore\` as the shared resynthesis + defEq-check
logic
- Add \`canonInstProp\` for \`Grind.nestedProof\`: canonicalize the
proposition, attempt resynthesis, fall back silently (proof irrelevance
means no defEq check needed)
- Add \`canonInstDec\`/\`canonInstDec'\` for \`Grind.nestedDecidable\`:
canonicalize the proposition, attempt resynthesis with defEq guard, fall
back silently
- Remove the separate \`cacheInsts\` cache in favor of the existing
type/value caches via \`withCaching\`
- Update module-level documentation

Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
2026-04-04 00:40:39 +00:00
535 changed files with 3620 additions and 1546 deletions

View File

@@ -59,11 +59,11 @@ jobs:
with:
msystem: clang64
# `:` means do not prefix with msystem
pacboy: "make: python: cmake clang ccache gmp libuv openssl: git: zip: unzip: diffutils: binutils: tree: zstd tar:"
pacboy: "make: python: cmake clang ccache gmp libuv git: zip: unzip: diffutils: binutils: tree: zstd tar:"
if: runner.os == 'Windows'
- name: Install Brew Packages
run: |
brew install ccache tree zstd coreutils gmp libuv openssl
brew install ccache tree zstd coreutils gmp libuv
if: runner.os == 'macOS'
- name: Checkout
uses: actions/checkout@v6
@@ -92,7 +92,7 @@ jobs:
run: |
sudo dpkg --add-architecture i386
sudo apt-get update
sudo apt-get install -y gcc-multilib g++-multilib ccache libuv1-dev:i386 libssl-dev:i386 pkgconf:i386
sudo apt-get install -y gcc-multilib g++-multilib ccache libuv1-dev:i386 pkgconf:i386
if: matrix.cmultilib
- name: Restore Cache
id: restore-cache
@@ -131,7 +131,7 @@ jobs:
[ -d build ] || mkdir build
cd build
# arguments passed to `cmake`
OPTIONS=(-DLEAN_EXTRA_MAKE_OPTS=-DwarningAsError=true)
OPTIONS=(-DWFAIL=ON)
if [[ -n '${{ matrix.release }}' ]]; then
# this also enables githash embedding into stage 1 library, which prohibits reusing
# `.olean`s across commits, so we don't do it in the fast non-release CI

View File

@@ -77,7 +77,7 @@ jobs:
# sync options with `Linux Lake` to ensure cache reuse
run: |
mkdir -p build
cmake --preset release -B build -DLEAN_EXTRA_MAKE_OPTS=-DwarningAsError=true
cmake --preset release -B build -DWFAIL=ON
shell: 'nix develop -c bash -euxo pipefail {0}'
- if: env.should_update_stage0 == 'yes'
run: |

1
.gitignore vendored
View File

@@ -34,3 +34,4 @@ wdErr.txt
wdIn.txt
wdOut.txt
downstream_releases/
.claude/worktrees/

View File

@@ -9,7 +9,6 @@ Requirements
- [CMake](http://www.cmake.org)
- [GMP (GNU multiprecision library)](http://gmplib.org/)
- [LibUV](https://libuv.org/)
- [OpenSSL](https://www.openssl.org/)
Platform-Specific Setup
-----------------------

View File

@@ -32,7 +32,7 @@ MSYS2 has a package management system, [pacman][pacman].
Here are the commands to install all dependencies needed to compile Lean on your machine.
```bash
pacman -S make python mingw-w64-clang-x86_64-cmake mingw-w64-clang-x86_64-clang mingw-w64-clang-x86_64-ccache mingw-w64-clang-x86_64-libuv mingw-w64-clang-x86_64-gmp mingw-w64-clang-x86_64-openssl git unzip diffutils binutils
pacman -S make python mingw-w64-clang-x86_64-cmake mingw-w64-clang-x86_64-clang mingw-w64-clang-x86_64-ccache mingw-w64-clang-x86_64-libuv mingw-w64-clang-x86_64-gmp git unzip diffutils binutils
```
You should now be able to run these commands:

View File

@@ -32,13 +32,12 @@ following to use `g++`.
cmake -DCMAKE_CXX_COMPILER=g++ ...
```
## Required Packages: CMake, GMP, libuv, OpenSSL, pkgconf
## Required Packages: CMake, GMP, libuv, pkgconf
```bash
brew install cmake
brew install gmp
brew install libuv
brew install openssl
brew install pkgconf
```

View File

@@ -8,5 +8,5 @@ follow the [generic build instructions](index.md).
## Basic packages
```bash
sudo apt-get install git libgmp-dev libuv1-dev libssl-dev cmake ccache clang pkgconf
sudo apt-get install git libgmp-dev libuv1-dev cmake ccache clang pkgconf
```

View File

@@ -24,7 +24,7 @@
stdenv = pkgs.overrideCC pkgs.stdenv llvmPackages.clang;
} ({
buildInputs = with pkgs; [
cmake gmp libuv ccache pkg-config openssl openssl.dev
cmake gmp libuv ccache pkg-config
llvmPackages.bintools # wrapped lld
llvmPackages.llvm # llvm-symbolizer for asan/lsan
gdb
@@ -34,21 +34,7 @@
hardeningDisable = [ "all" ];
# more convenient `ctest` output
CTEST_OUTPUT_ON_FAILURE = 1;
} // pkgs.lib.optionalAttrs pkgs.stdenv.isLinux (let
# Build OpenSSL 3 statically using pkgsDist's old-glibc stdenv,
# so the resulting static libs don't require newer glibc symbols.
opensslForDist = pkgsDist.stdenv.mkDerivation {
name = "openssl-static-${pkgs.lib.getVersion pkgs.openssl.name}";
inherit (pkgs.openssl) src;
nativeBuildInputs = [ pkgsDist.perl ];
configurePhase = ''
patchShebangs .
./config --prefix=$out no-shared no-tests
'';
buildPhase = "make -j$NIX_BUILD_CORES";
installPhase = "make install_sw";
};
in {
} // pkgs.lib.optionalAttrs pkgs.stdenv.isLinux {
GMP = (pkgsDist.gmp.override { withStatic = true; }).overrideAttrs (attrs:
pkgs.lib.optionalAttrs (pkgs.stdenv.system == "aarch64-linux") {
# would need additional linking setup on Linux aarch64, we don't use it anywhere else either
@@ -67,15 +53,13 @@
};
doCheck = false;
});
OPENSSL = opensslForDist;
OPENSSL_DEV = opensslForDist;
GLIBC = pkgsDist.glibc;
GLIBC_DEV = pkgsDist.glibc.dev;
GCC_LIB = pkgsDist.gcc.cc.lib;
ZLIB = pkgsDist.zlib;
# for CI coredumps
GDB = pkgsDist.gdb;
}));
});
in {
devShells.${system} = {
# The default development shell for working on lean itself

View File

@@ -1,7 +1,7 @@
#!/usr/bin/env bash
set -euxo pipefail
# run from root build directory (from inside nix-shell or otherwise defining GLIBC/ZLIB/GMP/OPENSSL) as in
# run from root build directory (from inside nix-shell or otherwise defining GLIBC/ZLIB/GMP) as in
# ```
# eval cmake ../.. $(../../script/prepare-llvm-linux.sh ~/Downloads/lean-llvm-x86_64-linux-gnu.tar.zst)
# ```
@@ -42,8 +42,6 @@ $CP $GLIBC/lib/*crt* stage1/lib/
# runtime
(cd llvm; $CP --parents lib/clang/*/lib/*/{clang_rt.*.o,libclang_rt.builtins*} ../stage1)
$CP llvm/lib/*/lib{c++,c++abi,unwind}.* $GMP/lib/libgmp.a $LIBUV/lib/libuv.a stage1/lib/
# bundle OpenSSL static libs
cp $OPENSSL/lib/libssl.a $OPENSSL/lib/libcrypto.a stage1/lib/
# LLVM 19 appears to ship the dependencies in 'llvm/lib/<target-triple>/' and 'llvm/include/<target-triple>/'
# but clang-19 that we use to compile is linked against 'llvm/lib/' and 'llvm/include'
# https://github.com/llvm/llvm-project/issues/54955
@@ -59,7 +57,6 @@ for f in $GLIBC/lib/{ld,lib{c,dl,m,rt,pthread}}-*; do b=$(basename $f); cp $f st
OPTIONS=()
# We build cadical using the custom toolchain on Linux to avoid glibc versioning issues
echo -n " -DLEAN_STANDALONE=ON -DCADICAL_USE_CUSTOM_CXX=ON"
echo -n " -DOPENSSL_INCLUDE_DIR=$OPENSSL_DEV/include -DOPENSSL_SSL_LIBRARY=$OPENSSL/lib/libssl.a -DOPENSSL_CRYPTO_LIBRARY=$OPENSSL/lib/libcrypto.a"
echo -n " -DCMAKE_CXX_COMPILER=$PWD/llvm-host/bin/clang++ -DLEAN_CXX_STDLIB='-Wl,-Bstatic -lc++ -lc++abi -Wl,-Bdynamic'"
# these should also be used for cadical, so do not use `LEAN_EXTRA_CXX_FLAGS` here
echo -n " -DCMAKE_CXX_FLAGS='--sysroot $PWD/llvm -idirafter $GLIBC_DEV/include ${EXTRA_FLAGS:-}'"
@@ -77,8 +74,8 @@ fi
echo -n " -DLEANC_INTERNAL_FLAGS='--sysroot ROOT -nostdinc -isystem ROOT/include/clang' -DLEANC_CC=ROOT/bin/clang"
# ld.so is usually included by the libc.so linker script but we discard those. Make sure it is linked to only after `libc.so` like in the original
# linker script so that no libc symbols are bound to it instead.
echo -n " -DLEANC_INTERNAL_LINKER_FLAGS='--sysroot ROOT -L ROOT/lib -L ROOT/lib/glibc -lc -lc_nonshared -Wl,--as-needed -l:ld.so -Wl,--no-as-needed -lpthread_nonshared -Wl,--as-needed -Wl,-Bstatic -lgmp -lunwind -luv -lssl -lcrypto -Wl,-Bdynamic -Wl,--no-as-needed -fuse-ld=lld'"
# when not using the above flags, link GMP/libuv/OpenSSL dynamically/as usual
echo -n " -DLEAN_EXTRA_LINKER_FLAGS='-Wl,--as-needed -lgmp -luv -Wl,-Bstatic -lssl -lcrypto -Wl,-Bdynamic -lpthread -ldl -lrt -Wl,--no-as-needed'"
echo -n " -DLEANC_INTERNAL_LINKER_FLAGS='--sysroot ROOT -L ROOT/lib -L ROOT/lib/glibc -lc -lc_nonshared -Wl,--as-needed -l:ld.so -Wl,--no-as-needed -lpthread_nonshared -Wl,--as-needed -Wl,-Bstatic -lgmp -lunwind -luv -Wl,-Bdynamic -Wl,--no-as-needed -fuse-ld=lld'"
# when not using the above flags, link GMP dynamically/as usual
echo -n " -DLEAN_EXTRA_LINKER_FLAGS='-Wl,--as-needed -lgmp -luv -lpthread -ldl -lrt -Wl,--no-as-needed'"
# do not set `LEAN_CC` for tests
echo -n " -DLEAN_TEST_VARS=''"

View File

@@ -10,7 +10,6 @@ set -uxo pipefail
GMP=${GMP:-$(brew --prefix)}
LIBUV=${LIBUV:-$(brew --prefix)}
OPENSSL=${OPENSSL:-$(brew --prefix openssl@3)}
[[ -d llvm ]] || (mkdir llvm; gtar xf $1 --strip-components 1 --directory llvm)
[[ -d llvm-host ]] || if [[ "$#" -gt 1 ]]; then
@@ -42,7 +41,6 @@ gcp llvm/lib/libc++.dylib stage1/lib/libc
# and apparently since Sonoma does not do so implicitly either
install_name_tool -id /usr/lib/libc++.dylib stage1/lib/libc/libc++.dylib
echo -n " -DLEAN_STANDALONE=ON"
echo -n " -DOPENSSL_INCLUDE_DIR=$OPENSSL/include -DOPENSSL_SSL_LIBRARY=$OPENSSL/lib/libssl.a -DOPENSSL_CRYPTO_LIBRARY=$OPENSSL/lib/libcrypto.a"
# do not change C++ compiler; libc++ etc. being system libraries means there's no danger of conflicts,
# and the custom clang++ outputs a myriad of warnings when consuming the SDK
echo -n " -DLEAN_EXTRA_CXX_FLAGS='${EXTRA_FLAGS:-}'"
@@ -50,8 +48,7 @@ if [[ -L llvm-host ]]; then
echo -n " -DCMAKE_C_COMPILER=$PWD/stage1/bin/clang"
gcp $GMP/lib/libgmp.a stage1/lib/
gcp $LIBUV/lib/libuv.a stage1/lib/
gcp $OPENSSL/lib/libssl.a $OPENSSL/lib/libcrypto.a stage1/lib/
echo -n " -DLEAN_EXTRA_LINKER_FLAGS='-lgmp -luv -lssl -lcrypto'"
echo -n " -DLEAN_EXTRA_LINKER_FLAGS='-lgmp -luv'"
else
echo -n " -DCMAKE_C_COMPILER=$PWD/llvm-host/bin/clang -DLEANC_OPTS='--sysroot $PWD/stage1 -resource-dir $PWD/stage1/lib/clang/15.0.1 ${EXTRA_FLAGS:-}'"
fi

View File

@@ -40,14 +40,14 @@ cp /clang64/lib/{crtbegin,crtend,crt2,dllcrt2}.o stage1/lib/
# tells the compiler how to dynamically link against `bcrypt.dll` (which is located in the System32 folder).
# This distinction is relevant specifically for `libicu.a`/`icu.dll` because there we want updates to the time zone database to
# be delivered to users via Windows Update without having to recompile Lean or Lean programs.
cp /clang64/lib/lib{m,bcrypt,mingw32,moldname,mingwex,msvcrt,pthread,advapi32,shell32,user32,kernel32,ucrtbase,psapi,iphlpapi,userenv,ws2_32,dbghelp,ole32,icu,crypt32,gdi32}.* /clang64/lib/libgmp.a /clang64/lib/libuv.a /clang64/lib/libssl.a /clang64/lib/libcrypto.a llvm/lib/lib{c++,c++abi,unwind}.a stage1/lib/
cp /clang64/lib/lib{m,bcrypt,mingw32,moldname,mingwex,msvcrt,pthread,advapi32,shell32,user32,kernel32,ucrtbase,psapi,iphlpapi,userenv,ws2_32,dbghelp,ole32,icu}.* /clang64/lib/libgmp.a /clang64/lib/libuv.a llvm/lib/lib{c++,c++abi,unwind}.a stage1/lib/
echo -n " -DLEAN_STANDALONE=ON"
echo -n " -DCMAKE_C_COMPILER=$PWD/stage1/bin/clang.exe -DCMAKE_C_COMPILER_WORKS=1 -DCMAKE_CXX_COMPILER=$PWD/llvm/bin/clang++.exe -DCMAKE_CXX_COMPILER_WORKS=1 -DLEAN_CXX_STDLIB='-lc++ -lc++abi'"
echo -n " -DSTAGE0_CMAKE_C_COMPILER=clang -DSTAGE0_CMAKE_CXX_COMPILER=clang++"
echo -n " -DLEAN_EXTRA_CXX_FLAGS='--sysroot $PWD/llvm -idirafter /clang64/include/'"
echo -n " -DLEANC_INTERNAL_FLAGS='--sysroot ROOT -nostdinc -isystem ROOT/include/clang' -DLEANC_CC=ROOT/bin/clang.exe"
echo -n " -DLEANC_INTERNAL_LINKER_FLAGS='--sysroot ROOT -L ROOT/lib -Wl,-Bstatic -lgmp $(pkg-config --static --libs libuv) -lssl -lcrypto -lunwind -Wl,-Bdynamic -lcrypt32 -lgdi32 -fuse-ld=lld'"
# when not using the above flags, link GMP/libuv/OpenSSL dynamically/as usual. Always link ICU dynamically.
echo -n " -DLEAN_EXTRA_LINKER_FLAGS='-lgmp $(pkg-config --libs libuv) -lssl -lcrypto -lcrypt32 -lgdi32 -lucrtbase'"
echo -n " -DLEANC_INTERNAL_LINKER_FLAGS='--sysroot ROOT -L ROOT/lib -Wl,-Bstatic -lgmp $(pkg-config --static --libs libuv) -lunwind -Wl,-Bdynamic -fuse-ld=lld'"
# when not using the above flags, link GMP dynamically/as usual. Always link ICU dynamically.
echo -n " -DLEAN_EXTRA_LINKER_FLAGS='-lgmp $(pkg-config --libs libuv) -lucrtbase'"
# do not set `LEAN_CC` for tests
echo -n " -DLEAN_TEST_VARS=''"

View File

@@ -28,6 +28,14 @@ repositories:
branch: main
dependencies: []
- name: leansqlite
url: https://github.com/leanprover/leansqlite
toolchain-tag: true
stable-branch: false
branch: main
dependencies:
- plausible
- name: verso
url: https://github.com/leanprover/verso
toolchain-tag: true
@@ -100,7 +108,7 @@ repositories:
toolchain-tag: true
stable-branch: false
branch: main
dependencies: [lean4-cli, BibtexQuery, mathlib4]
dependencies: [lean4-cli, BibtexQuery, mathlib4, leansqlite]
- name: cslib
url: https://github.com/leanprover/cslib

View File

@@ -481,11 +481,9 @@ def execute_release_steps(repo, version, config):
run_command("lake update", cwd=repo_path, stream_output=True)
elif repo_name == "verso":
# verso has nested Lake projects in test-projects/ that each have their own
# lake-manifest.json with a subverso pin. After updating the root manifest via
# `lake update`, sync the de-modulized subverso rev into all sub-manifests.
# The sub-projects use an old toolchain (v4.21.0) that doesn't support module/prelude
# syntax, so they need the de-modulized version (tagged no-modules/<root-rev>).
# The "SubVerso version consistency" CI check accepts either the root or de-modulized rev.
# lake-manifest.json with a subverso pin and their own lean-toolchain.
# After updating the root manifest via `lake update`, sync the de-modulized
# subverso rev into all sub-manifests, and update their lean-toolchain files.
run_command("lake update", cwd=repo_path, stream_output=True)
print(blue("Syncing de-modulized subverso rev to test-project sub-manifests..."))
sync_script = (
@@ -498,6 +496,15 @@ def execute_release_steps(repo, version, config):
)
run_command(sync_script, cwd=repo_path)
print(green("Synced de-modulized subverso rev to all test-project sub-manifests"))
# Update all lean-toolchain files in test-projects/ to match the root
print(blue("Updating lean-toolchain files in test-projects/..."))
find_result = run_command("find test-projects -name lean-toolchain", cwd=repo_path)
for tc_path in find_result.stdout.strip().splitlines():
if tc_path:
tc_file = repo_path / tc_path
with open(tc_file, "w") as f:
f.write(f"leanprover/lean4:{version}\n")
print(green(f" Updated {tc_path}"))
elif dependencies:
run_command(f'perl -pi -e \'s/"v4\\.[0-9]+(\\.[0-9]+)?(-rc[0-9]+)?"/"' + version + '"/g\' lakefile.*', cwd=repo_path)
run_command("lake update", cwd=repo_path, stream_output=True)
@@ -659,56 +666,61 @@ def execute_release_steps(repo, version, config):
# Fetch latest changes to ensure we have the most up-to-date nightly-testing branch
print(blue("Fetching latest changes from origin..."))
run_command("git fetch origin", cwd=repo_path)
try:
print(blue("Merging origin/nightly-testing..."))
run_command("git merge origin/nightly-testing", cwd=repo_path)
print(green("Merge completed successfully"))
except subprocess.CalledProcessError:
# Merge failed due to conflicts - check which files are conflicted
print(blue("Merge conflicts detected, checking which files are affected..."))
# Get conflicted files using git status
status_result = run_command("git status --porcelain", cwd=repo_path)
conflicted_files = []
for line in status_result.stdout.splitlines():
if len(line) >= 2 and line[:2] in ['UU', 'AA', 'DD', 'AU', 'UA', 'DU', 'UD']:
# Extract filename (skip the first 3 characters which are status codes)
conflicted_files.append(line[3:])
# Filter out allowed files
allowed_patterns = ['lean-toolchain', 'lake-manifest.json']
problematic_files = []
for file in conflicted_files:
is_allowed = any(pattern in file for pattern in allowed_patterns)
if not is_allowed:
problematic_files.append(file)
if problematic_files:
# There are conflicts in non-allowed files - fail
print(red("❌ Merge failed!"))
print(red(f"Merging nightly-testing resulted in conflicts in:"))
for file in problematic_files:
print(red(f" - {file}"))
print(red("Please resolve these conflicts manually."))
return
else:
# Only allowed files are conflicted - resolve them automatically
print(green(f"✅ Only allowed files conflicted: {', '.join(conflicted_files)}"))
print(blue("Resolving conflicts automatically..."))
# For lean-toolchain and lake-manifest.json, keep our versions
# Check if nightly-testing branch exists on origin (use local ref after fetch for exact match)
nightly_check = run_command("git show-ref --verify --quiet refs/remotes/origin/nightly-testing", cwd=repo_path, check=False)
if nightly_check.returncode != 0:
print(yellow("No nightly-testing branch found on origin, skipping merge"))
else:
try:
print(blue("Merging origin/nightly-testing..."))
run_command("git merge origin/nightly-testing", cwd=repo_path)
print(green("Merge completed successfully"))
except subprocess.CalledProcessError:
# Merge failed due to conflicts - check which files are conflicted
print(blue("Merge conflicts detected, checking which files are affected..."))
# Get conflicted files using git status
status_result = run_command("git status --porcelain", cwd=repo_path)
conflicted_files = []
for line in status_result.stdout.splitlines():
if len(line) >= 2 and line[:2] in ['UU', 'AA', 'DD', 'AU', 'UA', 'DU', 'UD']:
# Extract filename (skip the first 3 characters which are status codes)
conflicted_files.append(line[3:])
# Filter out allowed files
allowed_patterns = ['lean-toolchain', 'lake-manifest.json']
problematic_files = []
for file in conflicted_files:
print(blue(f"Keeping our version of {file}"))
run_command(f"git checkout --ours {file}", cwd=repo_path)
# Complete the merge
run_command("git add .", cwd=repo_path)
run_command("git commit --no-edit", cwd=repo_path)
print(green("Merge completed successfully with automatic conflict resolution"))
is_allowed = any(pattern in file for pattern in allowed_patterns)
if not is_allowed:
problematic_files.append(file)
if problematic_files:
# There are conflicts in non-allowed files - fail
print(red("❌ Merge failed!"))
print(red(f"Merging nightly-testing resulted in conflicts in:"))
for file in problematic_files:
print(red(f" - {file}"))
print(red("Please resolve these conflicts manually."))
return
else:
# Only allowed files are conflicted - resolve them automatically
print(green(f"✅ Only allowed files conflicted: {', '.join(conflicted_files)}"))
print(blue("Resolving conflicts automatically..."))
# For lean-toolchain and lake-manifest.json, keep our versions
for file in conflicted_files:
print(blue(f"Keeping our version of {file}"))
run_command(f"git checkout --ours {file}", cwd=repo_path)
# Complete the merge
run_command("git add .", cwd=repo_path)
run_command("git commit --no-edit", cwd=repo_path)
print(green("✅ Merge completed successfully with automatic conflict resolution"))
# Build and test (skip for Mathlib)
if repo_name not in ["mathlib4"]:

View File

@@ -116,11 +116,19 @@ option(CHECK_OLEAN_VERSION "Only load .olean files compiled with the current ver
option(USE_LAKE "Use Lake instead of lean.mk for building core libs from language server" ON)
option(USE_LAKE_CACHE "Use the Lake artifact cache for stage 1 builds (requires USE_LAKE)" OFF)
set(LEAN_EXTRA_MAKE_OPTS "" CACHE STRING "extra options to lean --make")
set(LEAN_EXTRA_OPTS "" CACHE STRING "extra options to lean (via lake or make)")
set(LEAN_EXTRA_MAKE_OPTS "" CACHE STRING "extra options to leanmake")
set(LEANC_CC ${CMAKE_C_COMPILER} CACHE STRING "C compiler to use in `leanc`")
# Temporary, core-only flags. Must be synced with stdlib_flags.h.
string(APPEND LEAN_EXTRA_MAKE_OPTS " -Dbackward.do.legacy=false")
string(APPEND LEAN_EXTRA_OPTS " -Dbackward.do.legacy=false")
# option used by the CI to fail on warnings
option(WFAIL "Fail build if warnings are emitted by Lean" ON)
if(WFAIL MATCHES "ON")
string(APPEND LAKE_EXTRA_ARGS " --wfail")
string(APPEND LEAN_EXTRA_MAKE_OPTS " -DwarningAsError=true")
endif()
if(LAZY_RC MATCHES "ON")
set(LEAN_LAZY_RC "#define LEAN_LAZY_RC")
@@ -198,7 +206,7 @@ set(CMAKE_ARCHIVE_OUTPUT_DIRECTORY "${CMAKE_BINARY_DIR}/lib/lean")
# OSX default thread stack size is very small. Moreover, in Debug mode, each new stack frame consumes a lot of extra memory.
if((MULTI_THREAD MATCHES "ON") AND (CMAKE_SYSTEM_NAME MATCHES "Darwin"))
string(APPEND LEAN_EXTRA_MAKE_OPTS " -s40000")
string(APPEND LEAN_EXTRA_OPTS " -s40000")
endif()
# We want explicit stack probes in huge Lean stack frames for robust stack overflow detection
@@ -357,28 +365,6 @@ if(NOT LEAN_STANDALONE)
string(APPEND LEAN_EXTRA_LINKER_FLAGS " ${LIBUV_LDFLAGS}")
endif()
# OpenSSL
if(NOT "${CMAKE_SYSTEM_NAME}" MATCHES "Emscripten")
find_package(OpenSSL 3 REQUIRED)
set(OPENSSL_LIBRARIES ${OPENSSL_SSL_LIBRARY} ${OPENSSL_CRYPTO_LIBRARY})
include_directories(${OPENSSL_INCLUDE_DIR})
string(JOIN " " OPENSSL_LIBRARIES_STR ${OPENSSL_LIBRARIES})
if(NOT LEAN_STANDALONE)
string(APPEND LEAN_EXTRA_LINKER_FLAGS " ${OPENSSL_LIBRARIES_STR}")
endif()
endif()
if(CMAKE_SYSTEM_NAME MATCHES "Linux")
string(APPEND LEANSHARED_LINKER_FLAGS " -Wl,-rpath,\\$$ORIGIN")
endif()
if(CMAKE_SYSTEM_NAME MATCHES "Windows")
string(APPEND LEANSHARED_LINKER_FLAGS " -lcrypt32 -lgdi32")
if(NOT LEAN_STANDALONE)
string(APPEND LEAN_EXTRA_LINKER_FLAGS " -lcrypt32 -lgdi32")
endif()
endif()
# Windows SDK (for ICU)
if(CMAKE_SYSTEM_NAME MATCHES "Windows")
# Pass 'tools' to skip MSVC version check (as MSVC/Visual Studio is not necessarily installed)
@@ -494,17 +480,6 @@ endif()
string(APPEND TOOLCHAIN_STATIC_LINKER_FLAGS " ${LEAN_CXX_STDLIB}")
string(APPEND TOOLCHAIN_SHARED_LINKER_FLAGS " ${LEAN_CXX_STDLIB}")
# In standalone (release) builds, OpenSSL must be statically embedded in libleanshared.so.
# In non-standalone (dev/CI nix) builds, `find_package(OpenSSL)` resolves to shared libs from
# the nix store. Adding those to libleanshared.so would produce DT_NEEDED entries pointing into
# the nix store (including transitive libc++.so.1 because nix's libssl is built with libc++).
# Those paths don't exist in `out/` during `make run-local` tests, breaking them.
# For non-standalone builds, OpenSSL is instead provided via LEAN_EXTRA_LINKER_FLAGS (leanc.sh)
# and resolved at runtime from the loading executable.
if(DEFINED OPENSSL_LIBRARIES_STR AND LEAN_STANDALONE)
string(APPEND TOOLCHAIN_SHARED_LINKER_FLAGS " ${OPENSSL_LIBRARIES_STR}")
endif()
# in local builds, link executables and not just dynlibs against C++ stdlib as well,
# which is required for e.g. asan
if(NOT LEAN_STANDALONE)
@@ -703,6 +678,9 @@ else()
set(LEAN_PATH_SEPARATOR ":")
endif()
# inherit genral options for lean.mk.in and stdlib.make.in
string(APPEND LEAN_EXTRA_MAKE_OPTS " ${LEAN_EXTRA_OPTS}")
# Version
configure_file("${LEAN_SOURCE_DIR}/version.h.in" "${LEAN_BINARY_DIR}/include/lean/version.h")
if(STAGE EQUAL 0)
@@ -796,7 +774,7 @@ if(STAGE GREATER 1)
endif()
else()
add_subdirectory(runtime)
if("${CMAKE_SYSTEM_NAME}" MATCHES "Emscripten")
if(CMAKE_SYSTEM_NAME MATCHES "Emscripten")
add_dependencies(leanrt libuv)
add_dependencies(leanrt_initial-exec libuv)
endif()
@@ -1087,7 +1065,7 @@ string(REPLACE "ROOT" "${CMAKE_BINARY_DIR}" LEANC_CC "${LEANC_CC}")
string(REPLACE "ROOT" "${CMAKE_BINARY_DIR}" LEANC_INTERNAL_FLAGS "${LEANC_INTERNAL_FLAGS}")
string(REPLACE "ROOT" "${CMAKE_BINARY_DIR}" LEANC_INTERNAL_LINKER_FLAGS "${LEANC_INTERNAL_LINKER_FLAGS}")
toml_escape("${LEAN_EXTRA_MAKE_OPTS}" LEAN_EXTRA_OPTS_TOML)
toml_escape("${LEAN_EXTRA_OPTS}" LEAN_EXTRA_OPTS_TOML)
if(CMAKE_BUILD_TYPE MATCHES "Debug|Release|RelWithDebInfo|MinSizeRel")
set(CMAKE_BUILD_TYPE_TOML "${CMAKE_BUILD_TYPE}")

View File

@@ -1085,6 +1085,17 @@ Examples:
def sum {α} [Add α] [Zero α] : Array α α :=
foldr (· + ·) 0
/--
Computes the product of the elements of an array.
Examples:
* `#[a, b, c].prod = a * (b * (c * 1))`
* `#[1, 2, 5].prod = 10`
-/
@[inline, expose]
def prod {α} [Mul α] [One α] : Array α α :=
foldr (· * ·) 1
/--
Counts the number of elements in the array `as` that satisfy the Boolean predicate `p`.

View File

@@ -7,6 +7,7 @@ module
prelude
public import Init.Data.List.Int.Sum
public import Init.Data.List.Int.Prod
public import Init.Data.Array.MinMax
import Init.Data.Int.Lemmas
@@ -74,4 +75,17 @@ theorem sum_div_length_le_max_of_max?_eq_some_int {xs : Array Int} (h : xs.max?
simpa [List.max?_toArray, List.sum_toArray] using
List.sum_div_length_le_max_of_max?_eq_some_int (by simpa using h)
@[simp] theorem prod_replicate_int {n : Nat} {a : Int} : (replicate n a).prod = a ^ n := by
rw [ List.toArray_replicate, List.prod_toArray]
simp
theorem prod_append_int {as₁ as₂ : Array Int} : (as₁ ++ as₂).prod = as₁.prod * as₂.prod := by
simp [prod_append]
theorem prod_reverse_int (xs : Array Int) : xs.reverse.prod = xs.prod := by
simp [prod_reverse]
theorem prod_eq_foldl_int {xs : Array Int} : xs.prod = xs.foldl (init := 1) (· * ·) := by
simp only [foldl_eq_foldr_reverse, Int.mul_comm, prod_eq_foldr, prod_reverse_int]
end Array

View File

@@ -4380,6 +4380,47 @@ theorem sum_eq_foldl [Zero α] [Add α] [Std.Associative (α := α) (· + ·)]
xs.sum = xs.foldl (init := 0) (· + ·) := by
simp [ sum_toList, List.sum_eq_foldl]
/-! ### prod -/
@[simp, grind =] theorem prod_empty [Mul α] [One α] : (#[] : Array α).prod = 1 := rfl
theorem prod_eq_foldr [Mul α] [One α] {xs : Array α} :
xs.prod = xs.foldr (init := 1) (· * ·) :=
rfl
@[simp, grind =]
theorem prod_toList [Mul α] [One α] {as : Array α} : as.toList.prod = as.prod := by
cases as
simp [Array.prod, List.prod]
@[simp, grind =]
theorem prod_append [One α] [Mul α] [Std.Associative (α := α) (· * ·)]
[Std.LawfulLeftIdentity (α := α) (· * ·) 1]
{as₁ as₂ : Array α} : (as₁ ++ as₂).prod = as₁.prod * as₂.prod := by
simp [ prod_toList, List.prod_append]
@[simp, grind =]
theorem prod_singleton [Mul α] [One α] [Std.LawfulRightIdentity (· * ·) (1 : α)] {x : α} :
#[x].prod = x := by
simp [Array.prod_eq_foldr, Std.LawfulRightIdentity.right_id x]
@[simp, grind =]
theorem prod_push [Mul α] [One α] [Std.Associative (α := α) (· * ·)]
[Std.LawfulIdentity (· * ·) (1 : α)] {xs : Array α} {x : α} :
(xs.push x).prod = xs.prod * x := by
simp [Array.prod_eq_foldr, Std.LawfulRightIdentity.right_id, Std.LawfulLeftIdentity.left_id,
Array.foldr_assoc]
@[simp, grind =]
theorem prod_reverse [One α] [Mul α] [Std.Associative (α := α) (· * ·)]
[Std.Commutative (α := α) (· * ·)]
[Std.LawfulLeftIdentity (α := α) (· * ·) 1] (xs : Array α) : xs.reverse.prod = xs.prod := by
simp [ prod_toList, List.prod_reverse]
theorem prod_eq_foldl [One α] [Mul α] [Std.Associative (α := α) (· * ·)]
[Std.LawfulIdentity (· * ·) (1 : α)] {xs : Array α} :
xs.prod = xs.foldl (init := 1) (· * ·) := by
simp [ prod_toList, List.prod_eq_foldl]
theorem foldl_toList_eq_flatMap {l : List α} {acc : Array β}
{F : Array β α Array β} {G : α List β}
(H : acc a, (F acc a).toList = acc.toList ++ G a) :

View File

@@ -8,6 +8,7 @@ module
prelude
public import Init.Data.Array.MinMax
import Init.Data.List.Nat.Sum
import Init.Data.List.Nat.Prod
import Init.Data.Array.Lemmas
public section
@@ -81,4 +82,24 @@ theorem sum_div_length_le_max_of_max?_eq_some_nat {xs : Array Nat} (h : xs.max?
simpa [List.max?_toArray, List.sum_toArray] using
List.sum_div_length_le_max_of_max?_eq_some_nat (by simpa using h)
protected theorem prod_pos_iff_forall_pos_nat {xs : Array Nat} : 0 < xs.prod x xs, 0 < x := by
simp [ prod_toList, List.prod_pos_iff_forall_pos_nat]
protected theorem prod_eq_zero_iff_exists_zero_nat {xs : Array Nat} :
xs.prod = 0 x xs, x = 0 := by
simp [ prod_toList, List.prod_eq_zero_iff_exists_zero_nat]
@[simp] theorem prod_replicate_nat {n : Nat} {a : Nat} : (replicate n a).prod = a ^ n := by
rw [ List.toArray_replicate, List.prod_toArray]
simp
theorem prod_append_nat {as₁ as₂ : Array Nat} : (as₁ ++ as₂).prod = as₁.prod * as₂.prod := by
simp [prod_append]
theorem prod_reverse_nat (xs : Array Nat) : xs.reverse.prod = xs.prod := by
simp [prod_reverse]
theorem prod_eq_foldl_nat {xs : Array Nat} : xs.prod = xs.foldl (init := 1) (· * ·) := by
simp only [foldl_eq_foldr_reverse, Nat.mul_comm, prod_eq_foldr, prod_reverse_nat]
end Array

View File

@@ -271,7 +271,7 @@ private def optionPelim' {α : Type u_1} (t : Option α) {β : Sort u_2}
/--
Inserts an `Option` case distinction after the first computation of a call to `MonadAttach.pbind`.
This lemma is useful for simplifying the second computation, which often involes `match` expressions
This lemma is useful for simplifying the second computation, which often involves `match` expressions
that use `pbind`'s proof term.
-/
private theorem pbind_eq_pbind_if_isSome [Monad m] [MonadAttach m] (x : m (Option α)) (f : (_ : _) _ m β) :

View File

@@ -2056,6 +2056,20 @@ def sum {α} [Add α] [Zero α] : List αα :=
@[simp, grind =] theorem sum_cons [Add α] [Zero α] {a : α} {l : List α} : (a::l).sum = a + l.sum := rfl
theorem sum_eq_foldr [Add α] [Zero α] {l : List α} : l.sum = l.foldr (· + ·) 0 := rfl
/--
Computes the product of the elements of a list.
Examples:
* `[a, b, c].prod = a * (b * (c * 1))`
* `[1, 2, 5].prod = 10`
-/
def prod {α} [Mul α] [One α] : List α α :=
foldr (· * ·) 1
@[simp, grind =] theorem prod_nil [Mul α] [One α] : ([] : List α).prod = 1 := rfl
@[simp, grind =] theorem prod_cons [Mul α] [One α] {a : α} {l : List α} : (a::l).prod = a * l.prod := rfl
theorem prod_eq_foldr [Mul α] [One α] {l : List α} : l.prod = l.foldr (· * ·) 1 := rfl
/-! ### range -/
/--

View File

@@ -7,3 +7,4 @@ module
prelude
public import Init.Data.List.Int.Sum
public import Init.Data.List.Int.Prod

View File

@@ -0,0 +1,31 @@
/-
Copyright (c) 2026 Lean FRO, LLC. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Kim Morrison
-/
module
prelude
import Init.Data.List.Lemmas
import Init.Data.Int.Lemmas
public import Init.Data.Int.Pow
public import Init.Data.List.Basic
public section
set_option linter.listVariables true -- Enforce naming conventions for `List`/`Array`/`Vector` variables.
set_option linter.indexVariables true -- Enforce naming conventions for index variables.
namespace List
@[simp]
theorem prod_replicate_int {n : Nat} {a : Int} : (replicate n a).prod = a ^ n := by
induction n <;> simp_all [replicate_succ, Int.pow_succ, Int.mul_comm]
theorem prod_append_int {l₁ l₂ : List Int} : (l₁ ++ l₂).prod = l₁.prod * l₂.prod := by
simp [prod_append]
theorem prod_reverse_int (xs : List Int) : xs.reverse.prod = xs.prod := by
simp [prod_reverse]
end List

View File

@@ -1878,6 +1878,24 @@ theorem sum_reverse [Zero α] [Add α] [Std.Associative (α := α) (· + ·)]
simp_all [sum_append, Std.Commutative.comm (α := α) _ 0,
Std.LawfulLeftIdentity.left_id, Std.Commutative.comm]
@[simp, grind =]
theorem prod_append [Mul α] [One α] [Std.LawfulLeftIdentity (α := α) (· * ·) 1]
[Std.Associative (α := α) (· * ·)] {l₁ l₂ : List α} : (l₁ ++ l₂).prod = l₁.prod * l₂.prod := by
induction l₁ generalizing l₂ <;> simp_all [Std.Associative.assoc, Std.LawfulLeftIdentity.left_id]
@[simp, grind =]
theorem prod_singleton [Mul α] [One α] [Std.LawfulRightIdentity (· * ·) (1 : α)] {x : α} :
[x].prod = x := by
simp [List.prod_eq_foldr, Std.LawfulRightIdentity.right_id x]
@[simp, grind =]
theorem prod_reverse [One α] [Mul α] [Std.Associative (α := α) (· * ·)]
[Std.Commutative (α := α) (· * ·)]
[Std.LawfulLeftIdentity (α := α) (· * ·) 1] (xs : List α) : xs.reverse.prod = xs.prod := by
induction xs <;>
simp_all [prod_append, Std.Commutative.comm (α := α) _ 1,
Std.LawfulLeftIdentity.left_id, Std.Commutative.comm]
/-! ### concat
Note that `concat_eq_append` is a `@[simp]` lemma, so `concat` should usually not appear in goals.
@@ -2784,6 +2802,11 @@ theorem sum_eq_foldl [Zero α] [Add α] [Std.Associative (α := α) (· + ·)]
xs.sum = xs.foldl (init := 0) (· + ·) := by
simp [sum_eq_foldr, foldl_eq_apply_foldr, Std.LawfulLeftIdentity.left_id]
theorem prod_eq_foldl [One α] [Mul α] [Std.Associative (α := α) (· * ·)]
[Std.LawfulIdentity (· * ·) (1 : α)] {xs : List α} :
xs.prod = xs.foldl (init := 1) (· * ·) := by
simp [prod_eq_foldr, foldl_eq_apply_foldr, Std.LawfulLeftIdentity.left_id]
-- The argument `f : α₁ → α₂` is intentionally explicit, as it is sometimes not found by unification.
theorem foldl_hom (f : α₁ α₂) {g₁ : α₁ β α₁} {g₂ : α₂ β α₂} {l : List β} {init : α₁}
(H : x y, g₂ (f x) y = f (g₁ x y)) : l.foldl g₂ (f init) = f (l.foldl g₁ init) := by

View File

@@ -13,6 +13,7 @@ public import Init.Data.List.Nat.Sublist
public import Init.Data.List.Nat.TakeDrop
public import Init.Data.List.Nat.Count
public import Init.Data.List.Nat.Sum
public import Init.Data.List.Nat.Prod
public import Init.Data.List.Nat.Erase
public import Init.Data.List.Nat.Find
public import Init.Data.List.Nat.BEq

View File

@@ -0,0 +1,50 @@
/-
Copyright (c) 2026 Lean FRO, LLC. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Kim Morrison
-/
module
prelude
import Init.Data.List.Lemmas
public import Init.BinderPredicates
public import Init.NotationExtra
import Init.Data.Nat.Lemmas
public section
set_option linter.listVariables true -- Enforce naming conventions for `List`/`Array`/`Vector` variables.
set_option linter.indexVariables true -- Enforce naming conventions for index variables.
namespace List
protected theorem prod_eq_zero_iff_exists_zero_nat {l : List Nat} : l.prod = 0 x l, x = 0 := by
induction l with
| nil => simp
| cons x xs ih =>
simp [Nat.mul_eq_zero, ih, eq_comm (a := (0 : Nat))]
protected theorem prod_pos_iff_forall_pos_nat {l : List Nat} : 0 < l.prod x l, 0 < x := by
induction l with
| nil => simp
| cons x xs ih =>
simp only [prod_cons, mem_cons, forall_eq_or_imp, ih]
constructor
· intro h
exact Nat.pos_of_mul_pos_right h, Nat.pos_of_mul_pos_left h
· exact fun hx, hxs => Nat.mul_pos hx hxs
@[simp]
theorem prod_replicate_nat {n : Nat} {a : Nat} : (replicate n a).prod = a ^ n := by
induction n <;> simp_all [replicate_succ, Nat.pow_succ, Nat.mul_comm]
theorem prod_append_nat {l₁ l₂ : List Nat} : (l₁ ++ l₂).prod = l₁.prod * l₂.prod := by
simp [prod_append]
theorem prod_reverse_nat (xs : List Nat) : xs.reverse.prod = xs.prod := by
simp [prod_reverse]
theorem prod_eq_foldl_nat {xs : List Nat} : xs.prod = xs.foldl (init := 1) (· * ·) := by
simp only [foldl_eq_foldr_reverse, Nat.mul_comm, prod_eq_foldr, prod_reverse_nat]
end List

View File

@@ -606,6 +606,13 @@ theorem sum_nat {l₁ l₂ : List Nat} (h : l₁ ~ l₂) : l₁.sum = l₂.sum :
| swap => simpa [List.sum_cons] using Nat.add_left_comm ..
| trans _ _ ih₁ ih₂ => simp [ih₁, ih₂]
theorem prod_nat {l₁ l₂ : List Nat} (h : l₁ ~ l₂) : l₁.prod = l₂.prod := by
induction h with
| nil => simp
| cons _ _ ih => simp [ih]
| swap => simpa [List.prod_cons] using Nat.mul_left_comm ..
| trans _ _ ih₁ ih₂ => simp [ih₁, ih₂]
theorem all_eq {l₁ l₂ : List α} {f : α Bool} (hp : l₁.Perm l₂) : l₁.all f = l₂.all f := by
rw [Bool.eq_iff_iff]; simp [hp.mem_iff]
@@ -615,6 +622,9 @@ theorem any_eq {l₁ l₂ : List α} {f : α → Bool} (hp : l₁.Perm l₂) : l
grind_pattern Perm.sum_nat => l₁ ~ l₂, l₁.sum
grind_pattern Perm.sum_nat => l₁ ~ l₂, l₂.sum
grind_pattern Perm.prod_nat => l₁ ~ l₂, l₁.prod
grind_pattern Perm.prod_nat => l₁ ~ l₂, l₂.prod
end Perm
end List

View File

@@ -213,6 +213,9 @@ theorem forM_toArray [Monad m] (l : List α) (f : α → m PUnit) :
@[simp, grind =] theorem sum_toArray [Add α] [Zero α] (l : List α) : l.toArray.sum = l.sum := by
simp [Array.sum, List.sum]
@[simp, grind =] theorem prod_toArray [Mul α] [One α] (l : List α) : l.toArray.prod = l.prod := by
simp [Array.prod, List.prod]
@[simp, grind =] theorem append_toArray (l₁ l₂ : List α) :
l₁.toArray ++ l₂.toArray = (l₁ ++ l₂).toArray := by
apply ext'

View File

@@ -60,7 +60,7 @@ theorem gcd_def (x y : Nat) : gcd x y = if x = 0 then y else gcd (y % x) x := by
cases n with
| zero => simp
| succ n =>
-- `simp [gcd_succ]` produces an invalid term unless `gcd_succ` is proved with `id rfl` instead
-- `simp [gcd_succ]` produces an invalid term unless `gcd_succ` is proved with `(rfl)` instead
rw [gcd_succ]
exact gcd_zero_left _
instance : Std.LawfulIdentity gcd 0 where

View File

@@ -1718,7 +1718,7 @@ theorem toArray_roc_append_toArray_roc {l m n : Nat} (h : l ≤ m) (h' : m ≤ n
@[simp]
theorem getElem_toArray_roc {m n i : Nat} (_h : i < (m<...=n).toArray.size) :
(m<...=n).toArray[i]'_h = m + 1 + i := by
simp [toArray_roc_eq_toArray_rco]
simp [toArray_roc_eq_toArray_rco]
theorem getElem?_toArray_roc {m n i : Nat} :
(m<...=n).toArray[i]? = if i < n - m then some (m + 1 + i) else none := by

View File

@@ -363,7 +363,7 @@ theorem toBitVec_eq_of_parseFirstByte_eq_threeMore {b : UInt8} (h : parseFirstBy
public def isInvalidContinuationByte (b : UInt8) : Bool :=
b &&& 0xc0 != 0x80
theorem isInvalidContinutationByte_eq_false_iff {b : UInt8} :
theorem isInvalidContinuationByte_eq_false_iff {b : UInt8} :
isInvalidContinuationByte b = false b &&& 0xc0 = 0x80 := by
simp [isInvalidContinuationByte]

View File

@@ -31,7 +31,7 @@ namespace Slice
/--
A list of all positions starting at {name}`p`.
This function is not meant to be used in actual progams. Actual programs should use
This function is not meant to be used in actual programs. Actual programs should use
{name}`Slice.positionsFrom` or {name}`Slice.positions`.
-/
protected def Model.positionsFrom {s : Slice} (p : s.Pos) : List { p : s.Pos // p s.endPos } :=
@@ -206,7 +206,7 @@ end Slice
/--
A list of all positions starting at {name}`p`.
This function is not meant to be used in actual progams. Actual programs should use
This function is not meant to be used in actual programs. Actual programs should use
{name}`Slice.positionsFrom` or {name}`Slice.positions`.
-/
protected def Model.positionsFrom {s : String} (p : s.Pos) : List { p : s.Pos // p s.endPos } :=

View File

@@ -23,7 +23,7 @@ Given a {name}`Slice` {name}`s`, the type {lean}`s.Subslice` is the type of half
in {name}`s` delineated by a valid position on both sides.
This type is useful to track regions of interest within some larger slice that is also of interest.
In contrast, {name}`Slice` is used to track regions of interest whithin some larger string that is
In contrast, {name}`Slice` is used to track regions of interest within some larger string that is
not or no longer relevant.
Equality on {name}`Subslice` is somewhat better behaved than on {name}`Slice`, but note that there

View File

@@ -506,6 +506,16 @@ Examples:
@[inline, expose] def sum [Add α] [Zero α] (xs : Vector α n) : α :=
xs.toArray.sum
/--
Computes the product of the elements of a vector.
Examples:
* `#v[a, b, c].prod = a * (b * (c * 1))`
* `#v[1, 2, 5].prod = 10`
-/
@[inline, expose] def prod [Mul α] [One α] (xs : Vector α n) : α :=
xs.toArray.prod
/--
Pad a vector on the left with a given element.

View File

@@ -30,4 +30,16 @@ theorem sum_reverse_int (xs : Vector Int n) : xs.reverse.sum = xs.sum := by
theorem sum_eq_foldl_int {xs : Vector Int n} : xs.sum = xs.foldl (b := 0) (· + ·) := by
simp only [foldl_eq_foldr_reverse, Int.add_comm, sum_eq_foldr, sum_reverse_int]
@[simp] theorem prod_replicate_int {n : Nat} {a : Int} : (replicate n a).prod = a ^ n := by
simp [ prod_toArray, Array.prod_replicate_int]
theorem prod_append_int {as₁ as₂ : Vector Int n} : (as₁ ++ as₂).prod = as₁.prod * as₂.prod := by
simp [ prod_toArray]
theorem prod_reverse_int (xs : Vector Int n) : xs.reverse.prod = xs.prod := by
simp [prod_reverse]
theorem prod_eq_foldl_int {xs : Vector Int n} : xs.prod = xs.foldl (b := 1) (· * ·) := by
simp only [foldl_eq_foldr_reverse, Int.mul_comm, prod_eq_foldr, prod_reverse_int]
end Vector

View File

@@ -278,6 +278,12 @@ theorem toArray_mk {xs : Array α} (h : xs.size = n) : (Vector.mk xs h).toArray
@[simp, grind =] theorem sum_toArray [Add α] [Zero α] {xs : Vector α n} :
xs.toArray.sum = xs.sum := rfl
@[simp] theorem prod_mk [Mul α] [One α] {xs : Array α} (h : xs.size = n) :
(Vector.mk xs h).prod = xs.prod := rfl
@[simp, grind =] theorem prod_toArray [Mul α] [One α] {xs : Vector α n} :
xs.toArray.prod = xs.prod := rfl
@[simp] theorem eq_mk : xs = Vector.mk as h xs.toArray = as := by
cases xs
simp
@@ -551,6 +557,10 @@ theorem toArray_toList {xs : Vector α n} : xs.toList.toArray = xs.toArray := rf
xs.toList.sum = xs.sum := by
rw [ toList_toArray, Array.sum_toList, sum_toArray]
@[simp, grind =] theorem prod_toList [Mul α] [One α] {xs : Vector α n} :
xs.toList.prod = xs.prod := by
rw [ toList_toArray, Array.prod_toList, prod_toArray]
@[simp] theorem getElem_toList {xs : Vector α n} {i : Nat} (h : i < xs.toList.length) :
xs.toList[i] = xs[i]'(by simpa using h) := by
cases xs
@@ -3134,3 +3144,39 @@ theorem sum_eq_foldl [Zero α] [Add α]
{xs : Vector α n} :
xs.sum = xs.foldl (b := 0) (· + ·) := by
simp [ sum_toList, List.sum_eq_foldl]
/-! ### prod -/
@[simp, grind =] theorem prod_empty [Mul α] [One α] : (#v[] : Vector α 0).prod = 1 := rfl
theorem prod_eq_foldr [Mul α] [One α] {xs : Vector α n} :
xs.prod = xs.foldr (b := 1) (· * ·) :=
rfl
@[simp, grind =]
theorem prod_append [One α] [Mul α] [Std.Associative (α := α) (· * ·)]
[Std.LeftIdentity (α := α) (· * ·) 1] [Std.LawfulLeftIdentity (α := α) (· * ·) 1]
{as₁ as₂ : Vector α n} : (as₁ ++ as₂).prod = as₁.prod * as₂.prod := by
simp [ prod_toList, List.prod_append]
@[simp, grind =]
theorem prod_singleton [Mul α] [One α] [Std.LawfulRightIdentity (· * ·) (1 : α)] {x : α} :
#v[x].prod = x := by
simp [ prod_toList, Std.LawfulRightIdentity.right_id x]
@[simp, grind =]
theorem prod_push [Mul α] [One α] [Std.Associative (α := α) (· * ·)]
[Std.LawfulIdentity (· * ·) (1 : α)] {xs : Vector α n} {x : α} :
(xs.push x).prod = xs.prod * x := by
simp [ prod_toArray]
@[simp, grind =]
theorem prod_reverse [One α] [Mul α] [Std.Associative (α := α) (· * ·)]
[Std.Commutative (α := α) (· * ·)]
[Std.LawfulLeftIdentity (α := α) (· * ·) 1] (xs : Vector α n) : xs.reverse.prod = xs.prod := by
simp [ prod_toList, List.prod_reverse]
theorem prod_eq_foldl [One α] [Mul α]
[Std.Associative (α := α) (· * ·)] [Std.LawfulIdentity (· * ·) (1 : α)]
{xs : Vector α n} :
xs.prod = xs.foldl (b := 1) (· * ·) := by
simp [ prod_toList, List.prod_eq_foldl]

View File

@@ -37,4 +37,23 @@ theorem sum_reverse_nat (xs : Vector Nat n) : xs.reverse.sum = xs.sum := by
theorem sum_eq_foldl_nat {xs : Vector Nat n} : xs.sum = xs.foldl (b := 0) (· + ·) := by
simp only [foldl_eq_foldr_reverse, Nat.add_comm, sum_eq_foldr, sum_reverse_nat]
protected theorem prod_pos_iff_forall_pos_nat {xs : Vector Nat n} : 0 < xs.prod x xs, 0 < x := by
simp [ prod_toArray, Array.prod_pos_iff_forall_pos_nat]
protected theorem prod_eq_zero_iff_exists_zero_nat {xs : Vector Nat n} :
xs.prod = 0 x xs, x = 0 := by
simp [ prod_toArray, Array.prod_eq_zero_iff_exists_zero_nat]
@[simp] theorem prod_replicate_nat {n : Nat} {a : Nat} : (replicate n a).prod = a ^ n := by
simp [ prod_toArray, Array.prod_replicate_nat]
theorem prod_append_nat {as₁ as₂ : Vector Nat n} : (as₁ ++ as₂).prod = as₁.prod * as₂.prod := by
simp [ prod_toArray]
theorem prod_reverse_nat (xs : Vector Nat n) : xs.reverse.prod = xs.prod := by
simp [prod_reverse]
theorem prod_eq_foldl_nat {xs : Vector Nat n} : xs.prod = xs.foldl (b := 1) (· * ·) := by
simp only [foldl_eq_foldr_reverse, Nat.mul_comm, prod_eq_foldr, prod_reverse_nat]
end Vector

View File

@@ -564,6 +564,28 @@ end Ring
end IsCharP
/--
`PowIdentity α p` states that `x ^ p = x` holds for all elements of `α`.
The primary source of instances is Fermat's little theorem: for a finite field with `q` elements,
`x ^ q = x` for every `x`. For `Fin p` or `ZMod p` with prime `p`, this gives `x ^ p = x`.
The `grind` ring solver uses this typeclass to add the relation `x ^ p - x = 0` to the
Groebner basis, which allows it to reduce high-degree polynomials. Mathlib can provide
instances for general finite fields via `FiniteField.pow_card`.
-/
class PowIdentity (α : Type u) [CommSemiring α] (p : outParam Nat) : Prop where
/-- Every element satisfies `x ^ p = x`. -/
pow_eq (x : α) : x ^ p = x
namespace PowIdentity
variable [CommSemiring α] [PowIdentity α p]
theorem pow (x : α) : x ^ p = x := pow_eq x
end PowIdentity
open AddCommGroup
theorem no_int_zero_divisors {α : Type u} [IntModule α] [NoNatZeroDivisors α] {k : Int} {a : α}

View File

@@ -193,7 +193,7 @@ theorem mul_assoc (a b c : Q α) : mul (mul a b) c = mul a (mul b c) := by
simp [Semiring.left_distrib, Semiring.right_distrib]; refine 0, ?_; ac_rfl
theorem mul_one (a : Q α) : mul a (natCast 1) = a := by
obtain _, _ := a; simp
obtain _, _ := a; simp
theorem one_mul (a : Q α) : mul (natCast 1) a = a := by
obtain _, _ := a; simp

View File

@@ -156,6 +156,12 @@ instance [i : NeZero n] : ToInt.Pow (Fin n) (.co 0 n) where
rw [pow_succ, ToInt.Mul.toInt_mul, ih, ToInt.wrap_toInt,
IntInterval.wrap_mul (by simp), Int.pow_succ, ToInt.wrap_toInt]
instance : PowIdentity (Fin 2) 2 where
pow_eq x := by
match x with
| 0, _ => rfl
| 1, _ => rfl
end Fin
end Lean.Grind

View File

@@ -145,7 +145,7 @@ Examples:
The constant function that ignores its argument.
If `a : α`, then `Function.const β a : β → α` is the “constant function with value `a`”. For all
arguments `b : β`, `Function.const β a b = a`.
arguments `b : β`, `Function.const β a b = a`. It is often written directly as `fun _ => a`.
Examples:
* `Function.const Bool 10 true = 10`
@@ -3754,7 +3754,7 @@ class Functor (f : Type u → Type v) : Type (max (u+1) v) where
/--
Mapping a constant function.
Given `a : α` and `v : f α`, `mapConst a v` is equivalent to `Function.const _ a <$> v`. For some
Given `a : α` and `v : f β`, `mapConst a v` is equivalent to `(fun _ => a) <$> v`. For some
functors, this can be implemented more efficiently; for all other functors, the default
implementation may be used.
-/

View File

@@ -1880,3 +1880,12 @@ lead to undefined behavior.
-/
@[extern "lean_runtime_forget"]
def Runtime.forget (a : α) : BaseIO Unit := return
set_option linter.unusedVariables false in
/--
Ensures `a` remains at least alive until the call site by holding a reference to `a`. This can be useful
for unsafe code (such as an FFI) that relies on a Lean object not being freed until after some point
in the program. At runtime, this will be a no-op as the C compiler will optimize away this call.
-/
@[extern "lean_runtime_hold"]
def Runtime.hold (a : @& α) : BaseIO Unit := return

View File

@@ -9,6 +9,7 @@ prelude
public import Lean.Meta.Sorry
public import Lean.Util.CollectAxioms
public import Lean.OriginalConstKind
import Lean.Compiler.MetaAttr
import all Lean.OriginalConstKind -- for accessing `privateConstKindsExt`
public section
@@ -208,8 +209,12 @@ where
catch _ => pure ()
def addAndCompile (decl : Declaration) (logCompileErrors : Bool := true) : CoreM Unit := do
def addAndCompile (decl : Declaration) (logCompileErrors : Bool := true)
(markMeta : Bool := false) : CoreM Unit := do
addDecl decl
if markMeta then
for n in decl.getNames do
modifyEnv (Lean.markMeta · n)
compileDecl decl (logErrors := logCompileErrors)
end Lean

View File

@@ -67,7 +67,7 @@ structure ParamMap where
The set of fvars that were already annotated as borrowed before arriving at this pass. We try to
preserve the annotations here if possible.
-/
annoatedBorrows : Std.HashSet FVarId := {}
annotatedBorrows : Std.HashSet FVarId := {}
namespace ParamMap
@@ -95,7 +95,7 @@ where
modify fun m =>
{ m with
map := m.map.insert (.decl decl.name) (initParamsIfNotExported exported decl.params),
annoatedBorrows := decl.params.foldl (init := m.annoatedBorrows) fun acc p =>
annotatedBorrows := decl.params.foldl (init := m.annotatedBorrows) fun acc p =>
if p.borrow then acc.insert p.fvarId else acc
}
goCode decl.name code
@@ -116,7 +116,7 @@ where
modify fun m =>
{ m with
map := m.map.insert (.jp declName decl.fvarId) (initParams decl.params),
annoatedBorrows := decl.params.foldl (init := m.annoatedBorrows) fun acc p =>
annotatedBorrows := decl.params.foldl (init := m.annotatedBorrows) fun acc p =>
if p.borrow then acc.insert p.fvarId else acc
}
goCode declName decl.value
@@ -286,7 +286,7 @@ where
ownFVar (fvarId : FVarId) (reason : OwnReason) : InferM Unit := do
unless ( get).owned.contains fvarId do
if !reason.isForced && ( get).paramMap.annoatedBorrows.contains fvarId then
if !reason.isForced && ( get).paramMap.annotatedBorrows.contains fvarId then
trace[Compiler.inferBorrow] "user annotation blocked owning {← PP.run <| PP.ppFVar fvarId}: {← reason.toString}"
else
trace[Compiler.inferBorrow] "own {← PP.run <| PP.ppFVar fvarId}: {← reason.toString}"

View File

@@ -121,7 +121,7 @@ def mkPerDeclaration (name : Name) (phase : Phase)
occurrence := occurrence
phase := phase
name := name
run := fun xs => xs.mapM run
run := fun xs => xs.mapM fun decl => do checkSystem "LCNF compiler"; run decl
end Pass

View File

@@ -28,7 +28,7 @@ inserts addition instructions to attempt to reuse the memory right away instead
allocator.
For this the paper defines three functions:
- `R` (called `Decl.insertResetReuse` here) which looks for candidates that might be elligible for
- `R` (called `Decl.insertResetReuse` here) which looks for candidates that might be eligible for
reuse. For these variables it invokes `D`.
- `D` which looks for code regions in which the target variable is dead (i.e. no longer read from),
it then invokes `S`. If `S` succeeds it inserts a `reset` instruction to match the `reuse`

View File

@@ -217,6 +217,8 @@ Simplify `code`
-/
partial def simp (code : Code .pure) : SimpM (Code .pure) := withIncRecDepth do
incVisited
if ( get).visited % 128 == 0 then
checkSystem "LCNF simp"
match code with
| .let decl k =>
let baseDecl := decl

View File

@@ -24,7 +24,7 @@ In particular we perform:
- folding of the most common cases arm into the default arm if possible
Note: Currently the simplifier still contains almost equivalent simplifications to the ones shown
here. We know that this causes unforseen behavior and do plan on changing it eventually.
here. We know that this causes unforeseen behavior and do plan on changing it eventually.
-/
-- TODO: the following functions are duplicated from simp and should be deleted in simp once we

View File

@@ -171,7 +171,7 @@ def toDecl (declName : Name) : CompilerM (Decl .pure) := do
if compiler.ignoreBorrowAnnotation.get ( getOptions) then
decl := { decl with params := decl.params.mapM (·.updateBorrow false) }
if isExport env decl.name && decl.params.any (·.borrow) then
throwError m!" Declaration {decl.name} is marked as `export` but some of its parameters have borrow annotations.\n Consider using `set_option compiler.ignoreBorrowAnnotation true in` to supress the borrow annotations in its type.\n If the declaration is part of an `export`/`extern` pair make sure to also supress the annotations at the `extern` declaration."
throwError m!" Declaration {decl.name} is marked as `export` but some of its parameters have borrow annotations.\n Consider using `set_option compiler.ignoreBorrowAnnotation true in` to suppress the borrow annotations in its type.\n If the declaration is part of an `export`/`extern` pair make sure to also suppress the annotations at the `extern` declaration."
return decl
end Lean.Compiler.LCNF

View File

@@ -20,6 +20,8 @@ register_builtin_option diagnostics : Bool := {
descr := "collect diagnostic information"
}
builtin_initialize registerTraceClass `diagnostics
register_builtin_option diagnostics.threshold : Nat := {
defValue := 20
descr := "only diagnostic counters above this threshold are reported by the definitional equality"
@@ -444,10 +446,6 @@ Note that the value of `ctx.initHeartbeats` is ignored and replaced with `IO.get
@[inline] def CoreM.toIO' (x : CoreM α) (ctx : Context) (s : State) : IO α :=
(·.1) <$> x.toIO ctx s
/-- withIncRecDepth for a monad `m` such that `[MonadControlT CoreM n]`. -/
protected def withIncRecDepth [Monad m] [MonadControlT CoreM m] (x : m α) : m α :=
controlAt CoreM fun runInBase => withIncRecDepth (runInBase x)
/--
Throws an internal interrupt exception if cancellation has been requested. The exception is not
caught by `try catch` but is intended to be caught by `Command.withLoggingExceptions` at the top
@@ -462,6 +460,12 @@ heartbeat tracking (e.g. `SynthInstance`).
if ( tk.isSet) then
throwInterruptException
/-- withIncRecDepth for a monad `m` such that `[MonadControlT CoreM n]`.
Also checks for cancellation, so that recursive elaboration functions
(inferType, whnf, isDefEq, …) respond promptly to interrupt requests. -/
protected def withIncRecDepth [Monad m] [MonadControlT CoreM m] (x : m α) : m α :=
controlAt CoreM fun runInBase => do checkInterrupted; withIncRecDepth (runInBase x)
register_builtin_option debug.moduleNameAtTimeout : Bool := {
defValue := true
descr := "include module name in deterministic timeout error messages.\nRemark: we set this option to false to increase the stability of our test suite"

View File

@@ -111,8 +111,14 @@ open Lean.Meta
for x in loopMutVars do
let defn getLocalDeclFromUserName x.getId
Term.addTermInfo' x defn.toExpr
-- ForIn forces all mut vars into the same universe: that of the do block result type.
discard <| Term.ensureHasType (mkSort (mi.u.succ)) defn.type
-- ForIn forces the mut tuple into the universe mi.u: that of the do block result type.
-- If we don't do this, then we are stuck on solving constraints such as
-- `max ?u.46 ?u.47 =?= max (max ?u.22 ?u.46) ?u.47`
-- It's important we do this as a separate isLevelDefEq check on the decremented level because
-- otherwise (`ensureHasType (mkSort mi.u.succ)`) we are stuck on constraints like
-- `max (?u+1) (?v+1) =?= ?u+1`
let u getDecLevel defn.type
discard <| isLevelDefEq u mi.u
defs := defs.push defn.toExpr
if info.returnsEarly && loopMutVars.isEmpty then
defs := defs.push (mkConst ``Unit.unit)

View File

@@ -233,27 +233,41 @@ def processDefDeriving (view : DerivingClassView) (decl : Expr) (isNoncomputable
finally
Core.setMessageLog (msgLog ++ ( Core.getMessageLog))
let env getEnv
let hints := ReducibilityHints.regular (getMaxHeight env result.value + 1)
let decl mkDefinitionValInferringUnsafe instName result.levelParams.toList result.type result.value hints
-- Pre-check: if the instance value depends on noncomputable definitions and the user didn't write
-- `noncomputable`, give an actionable error with a `Try this:` suggestion.
unless isNoncomputable || ( read).isNoncomputableSection || ( isProp result.type) do
let noncompRef? := preNormValue.foldConsts none fun n acc =>
acc <|> if Lean.isNoncomputable (asyncMode := .local) env n then some n else none
if let some noncompRef := noncompRef? then
if let some cmdRef := cmdRef? then
if let some origText := cmdRef.reprint then
let newText := (origText.replace "deriving instance " "deriving noncomputable instance ").trimAscii
logInfoAt cmdRef m!"Try this: {newText}"
throwError "failed to derive instance because it depends on \
`{.ofConstName noncompRef}`, which is noncomputable"
if isNoncomputable || ( read).isNoncomputableSection then
addDecl <| Declaration.defnDecl decl
modifyEnv (addNoncomputable · instName)
let isPropType isProp result.type
if isPropType then
let decl mkThmOrUnsafeDef {
name := instName, levelParams := result.levelParams.toList,
type := result.type, value := result.value
}
addDecl decl
else
addAndCompile <| Declaration.defnDecl decl
let hints := ReducibilityHints.regular (getMaxHeight env result.value + 1)
let decl mkDefinitionValInferringUnsafe instName result.levelParams.toList result.type result.value hints
-- Pre-check: if the instance value depends on noncomputable definitions and the user didn't write
-- `noncomputable`, give an actionable error with a `Try this:` suggestion.
unless isNoncomputable || ( read).isNoncomputableSection do
let noncompRef? := preNormValue.foldConsts none fun n acc =>
acc <|> if Lean.isNoncomputable (asyncMode := .local) env n then some n else none
if let some noncompRef := noncompRef? then
if let some cmdRef := cmdRef? then
if let some origText := cmdRef.reprint then
let newText := (origText.replace "deriving instance " "deriving noncomputable instance ").trimAscii
logInfoAt cmdRef m!"Try this: {newText}"
throwError "failed to derive instance because it depends on \
`{.ofConstName noncompRef}`, which is noncomputable"
if isNoncomputable || ( read).isNoncomputableSection then
addDecl <| Declaration.defnDecl decl
modifyEnv (addNoncomputable · instName)
else
addAndCompile <| Declaration.defnDecl decl
trace[Elab.Deriving] "Derived instance `{.ofConstName instName}`"
registerInstance instName AttributeKind.global (eval_prio default)
-- For Prop-typed instances (theorems), skip `implicit_reducible` since reducibility hints are
-- irrelevant for theorems. This matches the behavior of the handwritten `instance` command
-- (see `MutualDef.lean`).
if isPropType then
addInstance instName AttributeKind.global (eval_prio default)
else
registerInstance instName AttributeKind.global (eval_prio default)
addDeclarationRangesFromSyntax instName ( getRef)
end Term

View File

@@ -111,7 +111,7 @@ def mkMatchNew (ctx : Context) (header : Header) (indVal : InductiveVal) : TermE
let x1 := mkIdent header.targetNames[0]!
let x2 := mkIdent header.targetNames[1]!
let ctorIdxName := mkCtorIdxName indVal.name
-- NB: the getMatcherInfo? assumes all mathcers are called `match_`
-- NB: the getMatcherInfo? assumes all matchers are called `match_`
let casesOnSameCtorName mkFreshUserName (indVal.name ++ `match_on_same_ctor)
mkCasesOnSameCtor casesOnSameCtorName indVal.name
let alts Array.ofFnM (n := indVal.numCtors) fun ctorIdx, _ => do

View File

@@ -28,7 +28,9 @@ def HeaderSyntax.isModule (header : HeaderSyntax) : Bool :=
def HeaderSyntax.imports (stx : HeaderSyntax) (includeInit : Bool := true) : Array Import :=
match stx with
| `(Parser.Module.header| $[module%$moduleTk]? $[prelude%$preludeTk]? $importsStx*) =>
let imports := if preludeTk.isNone && includeInit then #[{ module := `Init : Import }] else #[]
let imports := if preludeTk.isNone && includeInit then
#[{ module := `Init : Import }, { module := `Init, isMeta := true : Import }]
else #[]
imports ++ importsStx.map fun
| `(Parser.Module.import| $[public%$publicTk]? $[meta%$metaTk]? import $[all%$allTk]? $n) =>
{ module := n.getId, importAll := allTk.isSome

View File

@@ -233,7 +233,7 @@ def setImportAll : Parser := fun _ s =>
def main : Parser :=
keywordCore "module" (setIsModule false) (setIsModule true) >>
keywordCore "prelude" (fun _ s => s.pushImport `Init) skip >>
keywordCore "prelude" (fun _ s => (s.pushImport `Init).pushImport { module := `Init, isMeta := true }) skip >>
manyImports (atomic (keywordCore "public" skip setExported >>
keywordCore "meta" skip setMeta >>
keyword "import") >>

View File

@@ -36,7 +36,7 @@ def mkContext (lratPath : System.FilePath) (cfg : BVDecideConfig) : TermElabM Ta
TacticContext.new lratPath cfg
/--
Prepare an `Expr` that proves `bvExpr.unsat` using native evalution.
Prepare an `Expr` that proves `bvExpr.unsat` using native evaluation.
-/
def lratChecker (ctx : TacticContext) (reflectionResult : ReflectionResult) : MetaM Expr := do
let cert LratCert.ofFile ctx.lratPath ctx.config.trimProofs

View File

@@ -357,6 +357,7 @@ def reflectBV (g : MVarId) : M ReflectionResult := g.withContext do
let mut sats := #[]
let mut unusedHypotheses := {}
for hyp in hyps do
checkSystem "bv_decide"
if let (some reflected, lemmas) (SatAtBVLogical.of (mkFVar hyp)).run then
sats := (sats ++ lemmas).push reflected
else

View File

@@ -33,6 +33,7 @@ where
Reify `x`, returns `none` if the reification procedure failed.
-/
go (origExpr : Expr) : LemmaM (Option ReifiedBVExpr) := do
checkSystem "bv_decide"
match_expr origExpr with
| BitVec.ofNat _ _ => goBvLit origExpr
| HAnd.hAnd _ _ _ _ lhsExpr rhsExpr =>
@@ -340,6 +341,7 @@ where
Reify `t`, returns `none` if the reification procedure failed.
-/
go (origExpr : Expr) : LemmaM (Option ReifiedBVLogical) := do
checkSystem "bv_decide"
match_expr origExpr with
| Bool.true => ReifiedBVLogical.mkBoolConst true
| Bool.false => ReifiedBVLogical.mkBoolConst false

View File

@@ -159,6 +159,7 @@ Repeatedly run a list of `Pass` until they either close the goal or an iteration
the goal anymore.
-/
partial def fixpointPipeline (passes : List Pass) (goal : MVarId) : PreProcessM (Option MVarId) := do
checkSystem "bv_decide"
let mut newGoal := goal
for pass in passes do
if let some nextGoal pass.run newGoal then

View File

@@ -18,3 +18,4 @@ public import Lean.Linter.List
public import Lean.Linter.Sets
public import Lean.Linter.UnusedSimpArgs
public import Lean.Linter.Coe
public import Lean.Linter.GlobalAttributeIn

View File

@@ -0,0 +1,59 @@
/-
Copyright (c) 2026 Lean FRO, LLC. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Wojciech Różowski
-/
module
prelude
public import Lean.Elab.Command
public import Lean.Linter.Basic
namespace Lean.Linter
open Elab.Command
private structure TopDownSkipQuot where
stx : Syntax
def topDownSkipQuot (stx : Syntax) : TopDownSkipQuot := stx
partial instance [Monad m] : ForIn m TopDownSkipQuot Syntax where
forIn := fun stx init f => do
let rec @[specialize] loop stx b [Inhabited (type_of% b)] := do
if stx.isQuot then return ForInStep.yield b
match ( f stx b) with
| ForInStep.yield b' =>
let mut b := b'
if let Syntax.node _ _ args := stx then
for arg in args do
match ( loop arg b) with
| ForInStep.yield b' => b := b'
| ForInStep.done b' => return ForInStep.done b'
return ForInStep.yield b
| ForInStep.done b => return ForInStep.done b
match ( @loop stx init init) with
| ForInStep.yield b => return b
| ForInStep.done b => return b
def getGlobalAttributesIn? : Syntax Option (Ident × Array (TSyntax `attr))
| `(attribute [$x,*] $id in $_) =>
let xs := x.getElems.filterMap fun a => match a.raw with
| `(Parser.Command.eraseAttr| -$_) => none
| `(Parser.Term.attrInstance| local $_attr:attr) => none
| `(Parser.Term.attrInstance| scoped $_attr:attr) => none
| `(attr| $a) => some a
(id, xs)
| _ => default
def globalAttributeIn : Linter where run := withSetOptionIn fun stx => do
for s in topDownSkipQuot stx do
if let some (id, nonScopedNorLocal) := getGlobalAttributesIn? s then
for attr in nonScopedNorLocal do
logErrorAt attr
m!"Despite the `in`, the attribute {attr} is added globally to {id}\n\
please remove the `in` or make this a `local {attr}`"
builtin_initialize addLinter globalAttributeIn
end Lean.Linter

View File

@@ -48,6 +48,16 @@ register_builtin_option backward.isDefEq.respectTransparency : Bool := {
when checking whether implicit arguments are definitionally equal"
}
/--
Controls the transparency used to check whether the type of metavariable matches the type of the
term being assigned to it.
-/
register_builtin_option backward.isDefEq.respectTransparency.types : Bool := {
defValue := false -- TODO: replace with `true` after we fix stage0
descr := "if true, do not bump transparency to `.default` \
when checking whether the type of a metavariable matches the type of the term being assigned to it."
}
/--
Controls whether *all* implicit arguments (not just instance-implicit `[..]`) get their
transparency bumped to `TransparencyMode.instances` during `isDefEq`.
@@ -335,10 +345,10 @@ private def isDefEqArgsFirstPass
/--
Ensure `MetaM` configuration is strong enough for checking definitional equality of
instances. For example, we must be able to unfold instances, `beta := true`, `proj := .yesWithDelta`
are essential.
implicit arguments (e.g., instances) and types.
For example, we must be able to unfold instances, `beta := true`, `proj := .yesWithDelta` are essential.
-/
@[inline] def withInstanceConfig (x : MetaM α) : MetaM α :=
@[inline] def withImplicitConfig (x : MetaM α) : MetaM α :=
withAtLeastTransparency .instances do
let cfg getConfig
if cfg.beta && cfg.iota && cfg.zeta && cfg.zetaHave && cfg.zetaDelta && cfg.proj == .yesWithDelta then
@@ -382,7 +392,7 @@ private partial def isDefEqArgs (f : Expr) (args₁ args₂ : Array Expr) : Meta
-- Bump to `.instances` so that `[implicit_reducible]` definitions (instances, `Nat.add`,
-- `Array.size`, etc.) are unfolded. The user doesn't choose implicit arguments directly,
-- so Lean should try harder than the caller's transparency to make them match.
unless ( withInstanceConfig <| Meta.isExprDefEqAux a₁ a₂) do return false
unless ( withImplicitConfig <| Meta.isExprDefEqAux a₁ a₂) do return false
else if respectTransparency then
unless ( Meta.isExprDefEqAux a₁ a₂) do return false
else
@@ -392,7 +402,7 @@ private partial def isDefEqArgs (f : Expr) (args₁ args₂ : Array Expr) : Meta
let a₁ := args₁[i]!
let a₂ := args₂[i]!
if respectTransparency && (implicitBump || finfo.paramInfo[i]!.isInstance) then
unless ( withInstanceConfig <| Meta.isExprDefEqAux a₁ a₂) do return false
unless ( withImplicitConfig <| Meta.isExprDefEqAux a₁ a₂) do return false
else if !respectTransparency && finfo.paramInfo[i]!.isInstance then
-- Old behavior
unless ( withInferTypeConfig <| Meta.isExprDefEqAux a₁ a₂) do return false
@@ -454,6 +464,19 @@ private partial def isDefEqBindingAux (lctx : LocalContext) (fvars : Array Expr)
let lctx getLCtx
isDefEqBindingAux lctx #[] a b #[]
/--
Returns `true` if both `backward.isDefEq.respectTransparency` and `backward.isDefEq.respectTransparency.types` is true.
The option `backward.isDefEq.respectTransparency.types` is newer than ``backward.isDefEq.respectTransparency`,
and is used to enable the transparency bump when checking metavariable assignments.
If `backward.isDefEq.respectTransparency` is `false`, then we automatically disable
`backward.isDefEq.respectTransparency.types` too.
-/
abbrev respectTransparencyAtTypes : CoreM Bool := do
let opts getOptions
return backward.isDefEq.respectTransparency.types.get opts && backward.isDefEq.respectTransparency.get opts
private def checkTypesAndAssign (mvar : Expr) (v : Expr) : MetaM Bool :=
withTraceNodeBefore `Meta.isDefEq.assign.checkTypes (fun _ => return m!"({mvar} : {← inferType mvar}) := ({v} : {← inferType v})") do
if !mvar.isMVar then
@@ -462,14 +485,24 @@ private def checkTypesAndAssign (mvar : Expr) (v : Expr) : MetaM Bool :=
else
-- must check whether types are definitionally equal or not, before assigning and returning true
let mvarType inferType mvar
-- **TODO**: avoid transparency bump. Let's fix other issues first
withInferTypeConfig do
let vType inferType v
if ( Meta.isExprDefEqAux mvarType vType) then
mvar.mvarId!.assign v
pure true
else
pure false
let vType inferType v
if ( respectTransparencyAtTypes) then
withImplicitConfig do
if ( Meta.isExprDefEqAux mvarType vType) then
mvar.mvarId!.assign v
return true
else
if ( isDiagnosticsEnabled) then withInferTypeConfig do
if ( Meta.isExprDefEqAux mvarType vType) then
trace[diagnostics] "failure when assigning metavariable with type{indentExpr mvarType}\nwhich is not definitionally equal to{indentExpr vType}\nwhen using `.instances` transparency, but it is with `.default`.\nWorkaround: `set_option backward.isDefEq.respectTransparency.types false`"
return false
else
withInferTypeConfig do
if ( Meta.isExprDefEqAux mvarType vType) then
mvar.mvarId!.assign v
return true
else
return false
/--
Auxiliary method for solving constraints of the form `?m xs := v`.
@@ -2062,7 +2095,7 @@ private def isDefEqProj : Expr → Expr → MetaM Bool
for instance-implicit parameters. -/
let fromClass := isClass ( getEnv) m
let isDefEqStructArgs (x : MetaM Bool) : MetaM Bool :=
if fromClass then withInstanceConfig x else x
if fromClass then withImplicitConfig x else x
if ( read).inTypeClassResolution then
-- See comment at `inTypeClassResolution`
pure (i == j && m == n) <&&> isDefEqStructArgs (Meta.isExprDefEqAux t s)

View File

@@ -33,12 +33,12 @@ The high-level overview of moves are
* If there is an alternative, solve its constraints
* Else use `contradiction` to prove completeness of the match
* Process “independent prefixes” of patterns. These are patterns that can be processed without
affecting the aother alternatives, and without side effects in the sense of updating the `mvarId`.
affecting the other alternatives, and without side effects in the sense of updating the `mvarId`.
These are
- variable patterns; substitute
- inaccessible patterns; add equality constraints
- as-patterns: substitute value and equality
After thes have been processed, we use `.inaccessible x` where `x` is the variable being matched
After these have been processed, we use `.inaccessible x` where `x` is the variable being matched
to mark them as “done”.
* If all patterns start with “done”, drop the first variable
* The first alt has only “done” patterns, drop remaining alts (they're overlapped)
@@ -1108,6 +1108,9 @@ def mkMatcherAuxDefinition (name : Name) (type : Expr) (value : Expr) (isSplitte
-- matcher bodies should always be exported, if not private anyway
withExporting do
addDecl decl
-- if `matcher` is not private, we mark it as `implicit_reducible` too
unless isPrivateName name do
setReducibilityStatus name .implicitReducible
unless isSplitter do
modifyEnv fun env => matcherExt.modifyState env fun s => s.insert key name
addMatcherInfo name mi

View File

@@ -17,7 +17,7 @@ namespace Lean.Meta
/--
Tries to rewrite the `ite`, `dite` or `cond` expression `e` with the hypothesis `hc`.
If it fails, it returns a rewrite with `proof? := none` and unchaged expression.
If it fails, it returns a rewrite with `proof? := none` and unchanged expression.
-/
def rwIfWith (hc : Expr) (e : Expr) : MetaM Simp.Result := do
match_expr e with

View File

@@ -22,9 +22,9 @@ of that computation as an axiom towards the logic.
-/
public inductive NativeEqTrueResult where
/-- The given expression `e` evalutes to true. `prf` is a proof of `e = true`. -/
/-- The given expression `e` evaluates to true. `prf` is a proof of `e = true`. -/
| success (prf : Expr)
/-- The given expression `e` evalutes to false. -/
/-- The given expression `e` evaluates to false. -/
| notTrue
/--

View File

@@ -14,7 +14,7 @@ This module contains utilities for dealing with equalities between constructor a
in particular about which fields must be the same a-priori for the equality to type check.
Users include (or will include) the injectivity theorems, the per-constructor no-confusion
construction and deriving type classes lik `BEq`, `DecidableEq` or `Ord`.
construction and deriving type classes like `BEq`, `DecidableEq` or `Ord`.
-/
namespace Lean.Meta

View File

@@ -149,17 +149,15 @@ partial def mkSizeOfFn (recName : Name) (declName : Name): MetaM Unit := do
trace[Meta.sizeOf] "declName: {declName}"
trace[Meta.sizeOf] "type: {sizeOfType}"
trace[Meta.sizeOf] "val: {sizeOfValue}"
-- We expose the `sizeOf` functions so that the `spec` theorems can be publicly `defeq`
withExporting do
addDecl <| Declaration.defnDecl {
name := declName
levelParams := levelParams
type := sizeOfType
value := sizeOfValue
safety := DefinitionSafety.safe
hints := ReducibilityHints.abbrev
}
enableRealizationsForConst declName
addDecl <| Declaration.defnDecl {
name := declName
levelParams := levelParams
type := sizeOfType
value := sizeOfValue
safety := DefinitionSafety.safe
hints := ReducibilityHints.abbrev
}
enableRealizationsForConst declName
/--
Create `sizeOf` functions for all inductive datatypes in the mutual inductive declaration containing `typeName`
@@ -469,7 +467,6 @@ private def mkSizeOfSpecTheorem (indInfo : InductiveVal) (sizeOfFns : Array Name
type := thmType
value := thmValue
}
inferDefEqAttr thmName
simpAttr.add thmName default .global
grindAttr.add thmName grindAttrStx .global

View File

@@ -25,6 +25,7 @@ public import Lean.Meta.Sym.Simp
public import Lean.Meta.Sym.Util
public import Lean.Meta.Sym.Eta
public import Lean.Meta.Sym.Canon
public import Lean.Meta.Sym.Arith
public import Lean.Meta.Sym.Grind
public import Lean.Meta.Sym.SynthInstance

View File

@@ -97,4 +97,16 @@ public def mkLambdaFVarsS (xs : Array Expr) (e : Expr) : SymM Expr := do
let type abstractFVarsRange decl.type i xs
mkLambdaS decl.userName decl.binderInfo type b
/--
Similar to `mkForallFVars`, but uses the more efficient `abstractFVars` and `abstractFVarsRange`,
and makes the same assumption made by these functions.
-/
public def mkForallFVarsS (xs : Array Expr) (e : Expr) : SymM Expr := do
let b abstractFVars e xs
xs.size.foldRevM (init := b) fun i _ b => do
let x := xs[i]
let decl x.fvarId!.getDecl
let type abstractFVarsRange decl.type i xs
mkForallS decl.userName decl.binderInfo type b
end Lean.Meta.Sym

View File

@@ -189,4 +189,48 @@ def mkAppS₄ (f a₁ a₂ a₃ a₄ : Expr) : m Expr := do
def mkAppS₅ (f a₁ a₂ a₃ a₄ a₅ : Expr) : m Expr := do
mkAppS ( mkAppS₄ f a₁ a₂ a₃ a₄) a₅
def mkAppS₆ (f a₁ a₂ a₃ a₄ a₅ a₆ : Expr) : m Expr := do
mkAppS ( mkAppS₅ f a₁ a₂ a₃ a₄ a₅) a₆
def mkAppS₇ (f a₁ a₂ a₃ a₄ a₅ a₆ a₇ : Expr) : m Expr := do
mkAppS ( mkAppS₆ f a₁ a₂ a₃ a₄ a₅ a₆) a₇
def mkAppS₈ (f a₁ a₂ a₃ a₄ a₅ a₆ a₇ a₈ : Expr) : m Expr := do
mkAppS ( mkAppS₇ f a₁ a₂ a₃ a₄ a₅ a₆ a₇) a₈
def mkAppS₉ (f a₁ a₂ a₃ a₄ a₅ a₆ a₇ a₈ a₉ : Expr) : m Expr := do
mkAppS ( mkAppS₈ f a₁ a₂ a₃ a₄ a₅ a₆ a₇ a₈) a₉
def mkAppS₁₀ (f a₁ a₂ a₃ a₄ a₅ a₆ a₇ a₈ a₉ a₁₀ : Expr) : m Expr := do
mkAppS ( mkAppS₉ f a₁ a₂ a₃ a₄ a₅ a₆ a₇ a₈ a₉) a₁₀
def mkAppS₁₁ (f a₁ a₂ a₃ a₄ a₅ a₆ a₇ a₈ a₉ a₁₀ a₁₁ : Expr) : m Expr := do
mkAppS ( mkAppS₁₀ f a₁ a₂ a₃ a₄ a₅ a₆ a₇ a₈ a₉ a₁₀) a₁₁
/-- `mkAppRangeS f i j #[a₀, ..., aᵢ, ..., aⱼ, ...]` ==> `f aᵢ ... aⱼ₋₁` with max sharing. -/
partial def mkAppRangeS (f : Expr) (beginIdx endIdx : Nat) (args : Array Expr) : m Expr :=
go endIdx f beginIdx
where
go (endIdx : Nat) (b : Expr) (i : Nat) : m Expr := do
if endIdx i then return b
else go endIdx ( mkAppS b args[i]!) (i + 1)
/-- `mkAppNS f #[a₀, ..., aₙ]` constructs `f a₀ ... aₙ` with max sharing. -/
def mkAppNS (f : Expr) (args : Array Expr) : m Expr :=
mkAppRangeS f 0 args.size args
/-- `mkAppRevRangeS f b e revArgs` ==> `mkAppRev f (revArgs.extract b e)` with max sharing. -/
partial def mkAppRevRangeS (f : Expr) (beginIdx endIdx : Nat) (revArgs : Array Expr) : m Expr :=
go revArgs beginIdx f endIdx
where
go (revArgs : Array Expr) (start : Nat) (b : Expr) (i : Nat) : m Expr := do
if i start then return b
else
let i := i - 1
go revArgs start ( mkAppS b revArgs[i]!) i
/-- Same as `mkAppS f args` but reversing `args`, with max sharing. -/
def mkAppRevS (f : Expr) (revArgs : Array Expr) : m Expr :=
mkAppRevRangeS f 0 revArgs.size revArgs
end Lean.Meta.Sym.Internal

View File

@@ -0,0 +1,20 @@
/-
Copyright (c) 2026 Amazon.com, Inc. or its affiliates. All Rights Reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Leonardo de Moura
-/
module
prelude
public import Lean.Meta.Sym.Arith.Types
public import Lean.Meta.Sym.Arith.EvalNum
public import Lean.Meta.Sym.Arith.Classify
public import Lean.Meta.Sym.Arith.MonadCanon
public import Lean.Meta.Sym.Arith.MonadRing
public import Lean.Meta.Sym.Arith.MonadSemiring
public import Lean.Meta.Sym.Arith.MonadVar
public import Lean.Meta.Sym.Arith.Functions
public import Lean.Meta.Sym.Arith.Reify
public import Lean.Meta.Sym.Arith.DenoteExpr
public import Lean.Meta.Sym.Arith.ToExpr
public import Lean.Meta.Sym.Arith.VarRename
public import Lean.Meta.Sym.Arith.Poly

View File

@@ -0,0 +1,165 @@
/-
Copyright (c) 2026 Amazon.com, Inc. or its affiliates. All Rights Reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Leonardo de Moura
-/
module
prelude
public import Lean.Meta.Sym.Arith.EvalNum
import Lean.Meta.Sym.SynthInstance
import Lean.Meta.Sym.Canon
import Lean.Meta.DecLevel
import Init.Grind.Ring
public section
namespace Lean.Meta.Sym.Arith
/-!
# Algebraic structure classification
Detects the strongest algebraic structure available for a type and caches
the classification in `Arith.State.typeClassify`. The detection order is:
1. `Grind.CommRing` (includes `Field` check)
2. `Grind.Ring` (non-commutative)
3. `Grind.CommSemiring` (via `OfSemiring.Q` envelope)
4. `Grind.Semiring` (non-commutative)
Results (including failures) are cached in a single `PHashMap ExprPtr ClassifyResult`
to avoid repeated synthesis attempts.
-/
private def getIsCharInst? (u : Level) (type : Expr) (semiringInst : Expr) : SymM (Option (Expr × Nat)) := do
withNewMCtxDepth do
let n mkFreshExprMVar (mkConst ``Nat)
let charType := mkApp3 (mkConst ``Grind.IsCharP [u]) type semiringInst n
let some charInst Sym.synthInstance? charType | return none
let n instantiateMVars n
let some n evalNat? n | return none
return some (charInst, n)
private def getNoZeroDivInst? (u : Level) (type : Expr) : SymM (Option Expr) := do
let natModuleType := mkApp (mkConst ``Grind.NatModule [u]) type
let some natModuleInst Sym.synthInstance? natModuleType | return none
let noZeroDivType := mkApp2 (mkConst ``Grind.NoNatZeroDivisors [u]) type natModuleInst
Sym.synthInstance? noZeroDivType
private def getPowIdentityInst? (u : Level) (type : Expr) : SymM (Option (Expr × Expr × Nat)) := do withNewMCtxDepth do
-- We use a fresh metavar for `CommSemiring` (unlike `getIsCharInst?` which pins the semiring)
-- because `PowIdentity` instances may be declared against a canonical `CommSemiring` instance
-- that is not definitionally equal to `CommRing.toCommSemiring`. The synthesized `csInst` is
-- stored and used in proof terms to ensure type-correctness.
let csInst mkFreshExprMVar (mkApp (mkConst ``Grind.CommSemiring [u]) type)
let p mkFreshExprMVar (mkConst ``Nat)
let powIdentityType := mkApp3 (mkConst ``Grind.PowIdentity [u]) type csInst p
let some inst synthInstance? powIdentityType | return none
let csInst instantiateMVars csInst
let p instantiateMVars p
let some pVal evalNat? p | return none
return some (inst, csInst, pVal)
private def mkAddRightCancelInst? (u : Level) (type : Expr) : SymM (Option Expr) := do
let add := mkApp (mkConst ``Add [u]) type
let some addInst synthInstance? add | return none
let addRightCancel := mkApp2 (mkConst ``Grind.AddRightCancel [u]) type addInst
synthInstance? addRightCancel
/-- Try to classify `type` as a `CommRing`. Returns the ring id on success. -/
private def tryCommRing? (type : Expr) : SymM (Option Nat) := do
let u getDecLevel type
let commRing := mkApp (mkConst ``Grind.CommRing [u]) type
let some commRingInst Sym.synthInstance? commRing | return none
let ringInst := mkApp2 (mkConst ``Grind.CommRing.toRing [u]) type commRingInst
let semiringInst := mkApp2 (mkConst ``Grind.Ring.toSemiring [u]) type ringInst
let commSemiringInst := mkApp2 (mkConst ``Grind.CommRing.toCommSemiring [u]) type semiringInst
let charInst? getIsCharInst? u type semiringInst
let noZeroDivInst? getNoZeroDivInst? u type
let fieldInst? Sym.synthInstance? <| mkApp (mkConst ``Grind.Field [u]) type
let powIdentityInst? getPowIdentityInst? u type
let semiringId? := none
let id := ( getArithState).rings.size
let ring : CommRing := {
id, semiringId?, type, u, semiringInst, ringInst, commSemiringInst,
commRingInst, charInst?, noZeroDivInst?, fieldInst?, powIdentityInst?
}
modifyArithState fun s => { s with rings := s.rings.push ring }
return some id
/-- Try to classify `type` as a non-commutative `Ring`. -/
private def tryNonCommRing? (type : Expr) : SymM (Option Nat) := do
let u getDecLevel type
let ring := mkApp (mkConst ``Grind.Ring [u]) type
let some ringInst Sym.synthInstance? ring | return none
let semiringInst := mkApp2 (mkConst ``Grind.Ring.toSemiring [u]) type ringInst
let charInst? getIsCharInst? u type semiringInst
let id := ( getArithState).ncRings.size
let ring : Ring := {
id, type, u, semiringInst, ringInst, charInst?
}
modifyArithState fun s => { s with ncRings := s.ncRings.push ring }
return some id
/-- Helper function for `tryCommSemiring? -/
private def tryCacheAndCommRing? (type : Expr) : SymM (Option Nat) := do
if let some result := ( getArithState).typeClassify.find? { expr := type } then
let .commRing id := result | return none
return id
let id? tryCommRing? type
let result := match id? with
| none => .none
| some id => .commRing id
modifyArithState fun s => { s with typeClassify := s.typeClassify.insert { expr := type } result }
return id?
/-- Try to classify `type` as a `CommSemiring`. Creates the `OfSemiring.Q` envelope ring. -/
private def tryCommSemiring? (type : Expr) : SymM (Option Nat) := do
let u getDecLevel type
let commSemiring := mkApp (mkConst ``Grind.CommSemiring [u]) type
let some commSemiringInst Sym.synthInstance? commSemiring | return none
let semiringInst := mkApp2 (mkConst ``Grind.CommSemiring.toSemiring [u]) type commSemiringInst
let addRightCancelInst? mkAddRightCancelInst? u type
let q shareCommon ( Sym.canon (mkApp2 (mkConst ``Grind.Ring.OfSemiring.Q [u]) type semiringInst))
-- The envelope `Q` type must be classifiable as a CommRing.
let some ringId tryCacheAndCommRing? q
| reportIssue! "unexpected failure initializing ring{indentExpr q}"; return none
let id := ( getArithState).semirings.size
let semiring : CommSemiring := {
id, type, ringId, u, semiringInst, commSemiringInst, addRightCancelInst?
}
modifyArithState fun s => { s with semirings := s.semirings.push semiring }
-- Link the envelope ring back to this semiring
modifyArithState fun s =>
let rings := s.rings.modify ringId fun r => { r with semiringId? := some id }
{ s with rings }
return some id
/-- Try to classify `type` as a non-commutative `Semiring`. -/
private def tryNonCommSemiring? (type : Expr) : SymM (Option Nat) := do
let u getDecLevel type
let semiring := mkApp (mkConst ``Grind.Semiring [u]) type
let some semiringInst Sym.synthInstance? semiring | return none
let id := ( getArithState).ncSemirings.size
let semiring : Semiring := { id, type, u, semiringInst }
modifyArithState fun s => { s with ncSemirings := s.ncSemirings.push semiring }
return some id
/--
Classify the algebraic structure of `type`, trying the strongest first:
CommRing > Ring > CommSemiring > Semiring.
Results are cached in `Arith.State.typeClassify`.
-/
def classify? (type : Expr) : SymM ClassifyResult := do
if let some result := ( getArithState).typeClassify.find? { expr := type } then
return result
let result go
modifyArithState fun s => { s with typeClassify := s.typeClassify.insert { expr := type } result }
return result
where
go : SymM ClassifyResult := do
if let some id tryCommRing? type then return .commRing id
if let some id tryNonCommRing? type then return .nonCommRing id
if let some id tryCommSemiring? type then return .commSemiring id
if let some id tryNonCommSemiring? type then return .nonCommSemiring id
return .none
end Lean.Meta.Sym.Arith

View File

@@ -0,0 +1,93 @@
/-
Copyright (c) 2026 Amazon.com, Inc. or its affiliates. All Rights Reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Leonardo de Moura
-/
module
prelude
public import Lean.Meta.Sym.Arith.Functions
public import Lean.Meta.Sym.Arith.MonadVar
public section
namespace Lean.Meta.Sym.Arith
/-!
# Denotation of reified expressions
Converts reified `RingExpr`, `Poly`, `Mon`, `Power` back into Lean `Expr`s using
the ring's cached operator functions and variable array.
-/
variable [Monad m] [MonadError m] [MonadLiftT MetaM m] [MonadCanon m] [MonadRing m]
/-- Convert an integer to a numeral expression in the ring. Negative values use `getNegFn`. -/
def denoteNum (k : Int) : m Expr := do
let ring getRing
let n := mkRawNatLit k.natAbs
let ofNatInst if let some inst MonadCanon.synthInstance? (mkApp2 (mkConst ``OfNat [ring.u]) ring.type n) then
pure inst
else
pure <| mkApp3 (mkConst ``Grind.Semiring.ofNat [ring.u]) ring.type ring.semiringInst n
let e := mkApp3 (mkConst ``OfNat.ofNat [ring.u]) ring.type n ofNatInst
if k < 0 then
return mkApp ( getNegFn) e
else
return e
/-- Denote a `Power` (variable raised to a power). -/
def denotePower [MonadGetVar m] (pw : Power) : m Expr := do
let x getVar pw.x
if pw.k == 1 then
return x
else
return mkApp2 ( getPowFn) x (toExpr pw.k)
/-- Denote a `Mon` (product of powers). -/
def denoteMon [MonadGetVar m] (mn : Mon) : m Expr := do
match mn with
| .unit => denoteNum 1
| .mult pw mn => go mn ( denotePower pw)
where
go (mn : Mon) (acc : Expr) : m Expr := do
match mn with
| .unit => return acc
| .mult pw mn => go mn (mkApp2 ( getMulFn) acc ( denotePower pw))
/-- Denote a `Poly` (sum of coefficient × monomial terms). -/
def denotePoly [MonadGetVar m] (p : Poly) : m Expr := do
match p with
| .num k => denoteNum k
| .add k mn p => go p ( denoteTerm k mn)
where
denoteTerm (k : Int) (mn : Mon) : m Expr := do
if k == 1 then
denoteMon mn
else
return mkApp2 ( getMulFn) ( denoteNum k) ( denoteMon mn)
go (p : Poly) (acc : Expr) : m Expr := do
match p with
| .num 0 => return acc
| .num k => return mkApp2 ( getAddFn) acc ( denoteNum k)
| .add k mn p => go p (mkApp2 ( getAddFn) acc ( denoteTerm k mn))
/-- Denote a `RingExpr` using a variable lookup function. -/
@[specialize]
private def denoteRingExprCore (getVarExpr : Nat Expr) (e : RingExpr) : m Expr := do
go e
where
go : RingExpr m Expr
| .num k => denoteNum k
| .natCast k => return mkApp ( getNatCastFn) (mkNatLit k)
| .intCast k => return mkApp ( getIntCastFn) (mkIntLit k)
| .var x => return getVarExpr x
| .add a b => return mkApp2 ( getAddFn) ( go a) ( go b)
| .sub a b => return mkApp2 ( getSubFn) ( go a) ( go b)
| .mul a b => return mkApp2 ( getMulFn) ( go a) ( go b)
| .pow a k => return mkApp2 ( getPowFn) ( go a) (toExpr k)
| .neg a => return mkApp ( getNegFn) ( go a)
/-- Denote a `RingExpr` using an explicit variable array. -/
def denoteRingExpr (vars : Array Expr) (e : RingExpr) : m Expr := do
denoteRingExprCore (fun x => vars[x]!) e
end Lean.Meta.Sym.Arith

View File

@@ -0,0 +1,90 @@
/-
Copyright (c) 2026 Amazon.com, Inc. or its affiliates. All Rights Reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Leonardo de Moura
-/
module
prelude
public import Lean.Meta.Sym.Arith.Types
import Lean.Meta.Sym.LitValues
import Lean.Meta.IntInstTesters
import Lean.Meta.NatInstTesters
public section
namespace Lean.Meta.Sym.Arith
/-!
Functions for evaluating simple `Nat` and `Int` expressions that appear in type classes
(e.g., `ToInt` and `IsCharP`). Using `whnf` for this purpose is too expensive and can
exhaust the stack. We considered `evalExpr` as an alternative, but it introduces
considerable overhead. We may use `evalExpr` as a fallback in the future.
-/
def checkExp (k : Nat) : OptionT SymM Unit := do
let exp getExpThreshold
if k > exp then
reportIssue! "exponent {k} exceeds threshold for exponentiation `(exp := {exp})`"
failure
/-
**Note**: It is safe to use (the more efficient) structural instance tests here because
`Sym.Canon` has already run.
-/
open Structural in
mutual
private partial def evalNatCore (e : Expr) : OptionT SymM Nat := do
match_expr e with
| Nat.zero => return 0
| Nat.succ a => return ( evalNatCore a) + 1
| Int.toNat a => return ( evalIntCore a).toNat
| Int.natAbs a => return ( evalIntCore a).natAbs
| HAdd.hAdd _ _ _ inst a b => guard ( isInstHAddNat inst); return ( evalNatCore a) + ( evalNatCore b)
| HMul.hMul _ _ _ inst a b => guard ( isInstHMulNat inst); return ( evalNatCore a) * ( evalNatCore b)
| HSub.hSub _ _ _ inst a b => guard ( isInstHSubNat inst); return ( evalNatCore a) - ( evalNatCore b)
| HDiv.hDiv _ _ _ inst a b => guard ( isInstHDivNat inst); return ( evalNatCore a) / ( evalNatCore b)
| HMod.hMod _ _ _ inst a b => guard ( isInstHModNat inst); return ( evalNatCore a) % ( evalNatCore b)
| OfNat.ofNat _ _ _ =>
let some n := Sym.getNatValue? e |>.run | failure
return n
| HPow.hPow _ _ _ inst a k =>
guard ( isInstHPowNat inst)
let k evalNatCore k
checkExp k
let a evalNatCore a
return a ^ k
| _ => failure
private partial def evalIntCore (e : Expr) : OptionT SymM Int := do
match_expr e with
| Neg.neg _ i a => guard ( isInstNegInt i); return - ( evalIntCore a)
| HAdd.hAdd _ _ _ i a b => guard ( isInstHAddInt i); return ( evalIntCore a) + ( evalIntCore b)
| HSub.hSub _ _ _ i a b => guard ( isInstHSubInt i); return ( evalIntCore a) - ( evalIntCore b)
| HMul.hMul _ _ _ i a b => guard ( isInstHMulInt i); return ( evalIntCore a) * ( evalIntCore b)
| HDiv.hDiv _ _ _ i a b => guard ( isInstHDivInt i); return ( evalIntCore a) / ( evalIntCore b)
| HMod.hMod _ _ _ i a b => guard ( isInstHModInt i); return ( evalIntCore a) % ( evalIntCore b)
| HPow.hPow _ _ _ i a k =>
guard ( isInstHPowInt i)
let a evalIntCore a
let k evalNatCore k
checkExp k
return a ^ k
| OfNat.ofNat _ _ _ =>
let some n := Sym.getIntValue? e |>.run | failure
return n
| NatCast.natCast _ i a =>
let_expr instNatCastInt i | failure
return ( evalNatCore a)
| Nat.cast _ i a =>
let_expr instNatCastInt i | failure
return ( evalNatCore a)
| _ => failure
end
def evalNat? (e : Expr) : SymM (Option Nat) :=
evalNatCore e |>.run
def evalInt? (e : Expr) : SymM (Option Int) :=
evalIntCore e |>.run
end Lean.Meta.Sym.Arith

View File

@@ -1,52 +1,62 @@
/-
Copyright (c) 2025 Amazon.com, Inc. or its affiliates. All Rights Reserved.
Copyright (c) 2026 Amazon.com, Inc. or its affiliates. All Rights Reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Leonardo de Moura
-/
module
prelude
public import Lean.Meta.Tactic.Grind.Arith.CommRing.MonadRing
public import Lean.Meta.Sym.Arith.MonadRing
public import Lean.Meta.Sym.Arith.MonadSemiring
public section
namespace Lean.Meta.Grind.Arith.CommRing
namespace Lean.Meta.Sym.Arith
/-!
# Cached function expressions for arithmetic operators
Synthesizes and caches the canonical Lean expressions for arithmetic operators
(`+`, `*`, `-`, `^`, `intCast`, `natCast`, etc.). These cached expressions are used
during reification to validate instances via pointer equality (`isSameExpr`).
Each getter checks the cache field first. On a miss, it synthesizes the instance,
verifies it against the expected instance from the ring structure using `isDefEqI`,
canonicalizes the result via `canonExpr`, and stores it.
-/
variable [MonadLiftT MetaM m] [MonadError m] [Monad m] [MonadCanon m]
section
variable [MonadRing m]
private def checkInst (declName : Name) (inst inst' : Expr) : MetaM Unit := do
unless ( withReducibleAndInstances <| isDefEq inst inst') do
throwError "error while initializing arithmetic operators:\ninstance for `{declName}` {indentExpr inst}\nis not definitionally equal to the expected one {indentExpr inst'}\nwhen only reducible definitions and instances are reduced"
def checkInst (declName : Name) (inst inst' : Expr) : MetaM Unit := do
unless ( isDefEqI inst inst') do
throwError "error while initializing `grind ring` operators:\ninstance for `{declName}` {indentExpr inst}\nis not definitionally equal to the expected one {indentExpr inst'}\nwhen only reducible definitions and instances are reduced"
def mkUnaryFn (type : Expr) (u : Level) (instDeclName : Name) (declName : Name) (expectedInst : Expr) : m Expr := do
private def mkUnaryFn (type : Expr) (u : Level) (instDeclName : Name) (declName : Name) (expectedInst : Expr) : m Expr := do
let inst MonadCanon.synthInstance <| mkApp (mkConst instDeclName [u]) type
checkInst declName inst expectedInst
canonExpr <| mkApp2 (mkConst declName [u]) type inst
def mkBinHomoFn (type : Expr) (u : Level) (instDeclName : Name) (declName : Name) (expectedInst : Expr) : m Expr := do
private def mkBinHomoFn (type : Expr) (u : Level) (instDeclName : Name) (declName : Name) (expectedInst : Expr) : m Expr := do
let inst MonadCanon.synthInstance <| mkApp3 (mkConst instDeclName [u, u, u]) type type type
checkInst declName inst expectedInst
canonExpr <| mkApp4 (mkConst declName [u, u, u]) type type type inst
def mkPowFn (u : Level) (type : Expr) (semiringInst : Expr) : m Expr := do
private def mkPowFn (u : Level) (type : Expr) (semiringInst : Expr) : m Expr := do
let inst MonadCanon.synthInstance <| mkApp3 (mkConst ``HPow [u, 0, u]) type Nat.mkType type
let inst' := mkApp2 (mkConst ``Grind.Semiring.npow [u]) type semiringInst
checkInst ``HPow.hPow inst inst'
canonExpr <| mkApp4 (mkConst ``HPow.hPow [u, 0, u]) type Nat.mkType type inst
def mkNatCastFn (u : Level) (type : Expr) (semiringInst : Expr) : m Expr := do
private def mkNatCastFn (u : Level) (type : Expr) (semiringInst : Expr) : m Expr := do
let inst' := mkApp2 (mkConst ``Grind.Semiring.natCast [u]) type semiringInst
let instType := mkApp (mkConst ``NatCast [u]) type
-- Note that `Semiring.natCast` is not registered as a global instance
-- (to avoid introducing unwanted coercions)
-- so merely having a `Semiring α` instance
-- does not guarantee that an `NatCast α` will be available.
-- When both are present we verify that they are defeq,
-- and otherwise fall back to the field of the `Semiring α` instance that we already have.
-- Note: `Semiring.natCast` is not a global instance, so `NatCast α` may not be available.
-- When present, verify defeq; otherwise fall back to the semiring field.
let inst match ( MonadCanon.synthInstance? instType) with
| none => pure inst'
| some inst => checkInst ``NatCast.natCast inst inst'; pure inst
canonExpr <| mkApp2 (mkConst ``NatCast.natCast [u]) type inst
section RingFns
variable [MonadRing m]
def getAddFn : m Expr := do
let ring getRing
if let some addFn := ring.addFn? then return addFn
@@ -55,14 +65,6 @@ def getAddFn : m Expr := do
modifyRing fun s => { s with addFn? := some addFn }
return addFn
def getSubFn : m Expr := do
let ring getRing
if let some subFn := ring.subFn? then return subFn
let expectedInst := mkApp2 (mkConst ``instHSub [ring.u]) ring.type <| mkApp2 (mkConst ``Grind.Ring.toSub [ring.u]) ring.type ring.ringInst
let subFn mkBinHomoFn ring.type ring.u ``HSub ``HSub.hSub expectedInst
modifyRing fun s => { s with subFn? := some subFn }
return subFn
def getMulFn : m Expr := do
let ring getRing
if let some mulFn := ring.mulFn? then return mulFn
@@ -71,6 +73,14 @@ def getMulFn : m Expr := do
modifyRing fun s => { s with mulFn? := some mulFn }
return mulFn
def getSubFn : m Expr := do
let ring getRing
if let some subFn := ring.subFn? then return subFn
let expectedInst := mkApp2 (mkConst ``instHSub [ring.u]) ring.type <| mkApp2 (mkConst ``Grind.Ring.toSub [ring.u]) ring.type ring.ringInst
let subFn mkBinHomoFn ring.type ring.u ``HSub ``HSub.hSub expectedInst
modifyRing fun s => { s with subFn? := some subFn }
return subFn
def getNegFn : m Expr := do
let ring getRing
if let some negFn := ring.negFn? then return negFn
@@ -91,12 +101,7 @@ def getIntCastFn : m Expr := do
if let some intCastFn := ring.intCastFn? then return intCastFn
let inst' := mkApp2 (mkConst ``Grind.Ring.intCast [ring.u]) ring.type ring.ringInst
let instType := mkApp (mkConst ``IntCast [ring.u]) ring.type
-- Note that `Ring.intCast` is not registered as a global instance
-- (to avoid introducing unwanted coercions)
-- so merely having a `Ring α` instance
-- does not guarantee that an `IntCast α` will be available.
-- When both are present we verify that they are defeq,
-- and otherwise fall back to the field of the `Ring α` instance that we already have.
-- Note: `Ring.intCast` is not a global instance. Same pattern as `mkNatCastFn`.
let inst match ( MonadCanon.synthInstance? instType) with
| none => pure inst'
| some inst => checkInst ``Int.cast inst inst'; pure inst
@@ -111,32 +116,56 @@ def getNatCastFn : m Expr := do
modifyRing fun s => { s with natCastFn? := some natCastFn }
return natCastFn
private def mkOne (u : Level) (type : Expr) (semiringInst : Expr) : m Expr := do
let n := mkRawNatLit 1
let ofNatInst := mkApp3 (mkConst ``Grind.Semiring.ofNat [u]) type semiringInst n
canonExpr <| mkApp3 (mkConst ``OfNat.ofNat [u]) type n ofNatInst
end RingFns
def getOne [MonadLiftT GoalM m] : m Expr := do
let ring getRing
if let some one := ring.one? then return one
let one mkOne ring.u ring.type ring.semiringInst
modifyRing fun s => { s with one? := some one }
internalize one 0
return one
end
section
section CommRingFns
variable [MonadCommRing m]
def getInvFn : m Expr := do
let ring getCommRing
let some fieldInst := ring.fieldInst?
| throwError "`grind` internal error, type is not a field{indentExpr ring.type}"
| throwError "internal error: type is not a field{indentExpr ring.type}"
if let some invFn := ring.invFn? then return invFn
let expectedInst := mkApp2 (mkConst ``Grind.Field.toInv [ring.u]) ring.type fieldInst
let invFn mkUnaryFn ring.type ring.u ``Inv ``Inv.inv expectedInst
modifyCommRing fun s => { s with invFn? := some invFn }
return invFn
end
end Lean.Meta.Grind.Arith.CommRing
end CommRingFns
section SemiringFns
variable [MonadSemiring m]
def getAddFn' : m Expr := do
let sr getSemiring
if let some addFn := sr.addFn? then return addFn
let expectedInst := mkApp2 (mkConst ``instHAdd [sr.u]) sr.type <| mkApp2 (mkConst ``Grind.Semiring.toAdd [sr.u]) sr.type sr.semiringInst
let addFn mkBinHomoFn sr.type sr.u ``HAdd ``HAdd.hAdd expectedInst
modifySemiring fun s => { s with addFn? := some addFn }
return addFn
def getMulFn' : m Expr := do
let sr getSemiring
if let some mulFn := sr.mulFn? then return mulFn
let expectedInst := mkApp2 (mkConst ``instHMul [sr.u]) sr.type <| mkApp2 (mkConst ``Grind.Semiring.toMul [sr.u]) sr.type sr.semiringInst
let mulFn mkBinHomoFn sr.type sr.u ``HMul ``HMul.hMul expectedInst
modifySemiring fun s => { s with mulFn? := some mulFn }
return mulFn
def getPowFn' : m Expr := do
let sr getSemiring
if let some powFn := sr.powFn? then return powFn
let powFn mkPowFn sr.u sr.type sr.semiringInst
modifySemiring fun s => { s with powFn? := some powFn }
return powFn
def getNatCastFn' : m Expr := do
let sr getSemiring
if let some natCastFn := sr.natCastFn? then return natCastFn
let natCastFn mkNatCastFn sr.u sr.type sr.semiringInst
modifySemiring fun s => { s with natCastFn? := some natCastFn }
return natCastFn
end SemiringFns
end Lean.Meta.Sym.Arith

View File

@@ -0,0 +1,42 @@
/-
Copyright (c) 2026 Amazon.com, Inc. or its affiliates. All Rights Reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Leonardo de Moura
-/
module
prelude
public import Lean.Meta.Sym.Arith.Types
public import Lean.Meta.Sym.Canon
public import Lean.Meta.Sym.SynthInstance
public section
namespace Lean.Meta.Sym.Arith
class MonadCanon (m : Type Type) where
/--
Canonicalize an expression (types, instances, support arguments).
In `SymM`, this is `Sym.canon`. In `PP.M` (diagnostics), this is the identity.
-/
canonExpr : Expr m Expr
/--
Synthesize an instance, returning `none` on failure.
In `SymM`, this is `Sym.synthInstance?`. In `PP.M`, this is `Meta.synthInstance?`.
-/
synthInstance? : Expr m (Option Expr)
export MonadCanon (canonExpr)
@[always_inline]
instance (m n) [MonadLift m n] [MonadCanon m] : MonadCanon n where
canonExpr e := liftM (canonExpr e : m Expr)
synthInstance? e := liftM (MonadCanon.synthInstance? e : m (Option Expr))
def MonadCanon.synthInstance [Monad m] [MonadError m] [MonadCanon m] (type : Expr) : m Expr := do
let some inst synthInstance? type
| throwError "failed to find instance{indentExpr type}"
return inst
instance : MonadCanon SymM where
canonExpr := canon
synthInstance? := Sym.synthInstance?
end Lean.Meta.Sym.Arith

View File

@@ -1,13 +1,13 @@
/-
Copyright (c) 2025 Amazon.com, Inc. or its affiliates. All Rights Reserved.
Copyright (c) 2026 Amazon.com, Inc. or its affiliates. All Rights Reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Leonardo de Moura
-/
module
prelude
public import Lean.Meta.Tactic.Grind.Arith.CommRing.MonadCanon
public import Lean.Meta.Sym.Arith.MonadCanon
public section
namespace Lean.Meta.Grind.Arith.CommRing
namespace Lean.Meta.Sym.Arith
class MonadRing (m : Type Type) where
getRing : m Ring
@@ -36,4 +36,4 @@ instance (m) [Monad m] [MonadCommRing m] : MonadRing m where
getRing := return ( getCommRing).toRing
modifyRing f := modifyCommRing fun s => { s with toRing := f s.toRing }
end Lean.Meta.Grind.Arith.CommRing
end Lean.Meta.Sym.Arith

View File

@@ -1,13 +1,13 @@
/-
Copyright (c) 2025 Amazon.com, Inc. or its affiliates. All Rights Reserved.
Copyright (c) 2026 Amazon.com, Inc. or its affiliates. All Rights Reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Leonardo de Moura
-/
module
prelude
public import Lean.Meta.Tactic.Grind.Arith.CommRing.MonadCanon
public import Lean.Meta.Sym.Arith.MonadCanon
public section
namespace Lean.Meta.Grind.Arith.CommRing
namespace Lean.Meta.Sym.Arith
class MonadSemiring (m : Type Type) where
getSemiring : m Semiring
@@ -36,4 +36,4 @@ instance (m) [Monad m] [MonadCommSemiring m] : MonadSemiring m where
getSemiring := return ( getCommSemiring).toSemiring
modifySemiring f := modifyCommSemiring fun s => { s with toSemiring := f s.toSemiring }
end Lean.Meta.Grind.Arith.CommRing
end Lean.Meta.Sym.Arith

View File

@@ -0,0 +1,32 @@
/-
Copyright (c) 2026 Amazon.com, Inc. or its affiliates. All Rights Reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Leonardo de Moura
-/
module
prelude
public import Lean.Meta.Sym.Arith.Types
public section
namespace Lean.Meta.Sym.Arith
/-- Read a variable's Lean expression by index. Used by `DenoteExpr` and diagnostics (PP). -/
class MonadGetVar (m : Type Type) where
getVar : Var m Expr
export MonadGetVar (getVar)
@[always_inline]
instance (m n) [MonadLift m n] [MonadGetVar m] : MonadGetVar n where
getVar x := liftM (getVar x : m Expr)
/-- Create or lookup a variable for a Lean expression. Used by reification. -/
class MonadMkVar (m : Type Type) where
mkVar : Expr (gen : Nat) m Var
export MonadMkVar (mkVar)
@[always_inline]
instance (m n) [MonadLift m n] [MonadMkVar m] : MonadMkVar n where
mkVar e gen := liftM (mkVar e gen : m Var)
end Lean.Meta.Sym.Arith

View File

@@ -1,16 +1,36 @@
/-
Copyright (c) 2025 Amazon.com, Inc. or its affiliates. All Rights Reserved.
Copyright (c) 2026 Amazon.com, Inc. or its affiliates. All Rights Reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Leonardo de Moura
-/
module
prelude
public import Lean.Meta.Tactic.Grind.Arith.CommRing.NonCommRingM
public import Lean.Meta.Tactic.Grind.Arith.CommRing.NonCommSemiringM
public import Lean.Meta.Sym.Arith.Functions
public import Lean.Meta.Sym.Arith.MonadVar
public import Lean.Meta.Sym.LitValues
public section
namespace Lean.Meta.Grind.Arith.CommRing
namespace Lean.Meta.Sym.Arith
open Sym.Arith (MonadCanon)
variable [MonadLiftT MetaM m] [MonadError m] [Monad m] [MonadCanon m] [MonadRing m]
/-!
# Reification of arithmetic expressions
Converts Lean expressions into `CommRing.Expr` (ring) or `CommSemiring.Expr`
(semiring) for reflection-based normalization.
Instance validation uses pointer equality (`isSameExpr`) against cached function
expressions from `Functions.lean`.
## Differences from grind's `Reify.lean`
- Uses `MonadMkVar` for variable creation instead of grind's `internalize` + `mkVarCore`
- Uses `Sym.getNatValue?`/`Sym.getIntValue?` (pure) instead of `MetaM` versions
- No `MonadSetTermId` term-to-ring-id tracking is grind-specific
-/
section RingReify
variable [MonadLiftT SymM m] [MonadLiftT MetaM m] [MonadError m] [Monad m] [MonadCanon m] [MonadRing m] [MonadMkVar m]
def isAddInst (inst : Expr) : m Bool :=
return isSameExpr ( getAddFn).appArg! inst
@@ -27,29 +47,22 @@ def isIntCastInst (inst : Expr) : m Bool :=
def isNatCastInst (inst : Expr) : m Bool :=
return isSameExpr ( getNatCastFn).appArg! inst
private def reportAppIssue (e : Expr) : GoalM Unit := do
private def reportRingAppIssue [MonadLiftT SymM m] (e : Expr) : m Unit := do
reportIssue! "ring term with unexpected instance{indentExpr e}"
variable [MonadLiftT GoalM m] [MonadSetTermId m]
/--
Converts a Lean expression `e` in the `CommRing` into a `CommRing.Expr` object.
Converts a Lean expression `e` into a `RingExpr`.
If `skipVar` is `true`, then the result is `none` if `e` is not an interpreted `CommRing` term.
We use `skipVar := false` when processing inequalities, and `skipVar := true` for equalities and disequalities
If `skipVar` is `true`, returns `none` if `e` is not an interpreted ring term
(used for equalities/disequalities). If `false`, treats non-interpreted terms
as variables (used for inequalities).
-/
partial def reifyCore? (e : Expr) (skipVar : Bool) (gen : Nat) : m (Option RingExpr) := do
let mkVar (e : Expr) : m Var := do
if ( alreadyInternalized e) then
mkVarCore e
else
internalize e gen
mkVarCore e
partial def reifyRing? (e : Expr) (skipVar : Bool := true) (gen : Nat) : m (Option RingExpr) := do
let toVar (e : Expr) : m RingExpr := do
return .var ( mkVar e)
return .var ( mkVar e gen)
let asVar (e : Expr) : m RingExpr := do
reportAppIssue e
return .var ( mkVar e)
reportRingAppIssue e
return .var ( mkVar e gen)
let rec go (e : Expr) : m RingExpr := do
match_expr e with
| HAdd.hAdd _ _ _ i a b =>
@@ -59,27 +72,33 @@ partial def reifyCore? (e : Expr) (skipVar : Bool) (gen : Nat) : m (Option RingE
| HSub.hSub _ _ _ i a b =>
if ( isSubInst i) then return .sub ( go a) ( go b) else asVar e
| HPow.hPow _ _ _ i a b =>
let some k getNatValue? b | toVar e
let some k := Sym.getNatValue? b |>.run | toVar e
if ( isPowInst i) then return .pow ( go a) k else asVar e
| Neg.neg _ i a =>
if ( isNegInst i) then return .neg ( go a) else asVar e
| IntCast.intCast _ i a =>
if ( isIntCastInst i) then
let some k getIntValue? a | toVar e
let some k := Sym.getIntValue? a |>.run | toVar e
return .intCast k
else
asVar e
| NatCast.natCast _ i a =>
if ( isNatCastInst i) then
let some k getNatValue? a | toVar e
let some k := Sym.getNatValue? a |>.run | toVar e
return .natCast k
else
asVar e
| OfNat.ofNat _ n _ =>
let some k getNatValue? n | toVar e
/-
**Note**: We extract `n` directly as a raw nat literal. The grind version uses `MetaM`'s
`getNatValue?` which handles multiple encodings (raw literals, nested `OfNat`, etc.).
In `SymM`, we assume terms have been canonicalized by `Sym.canon` before reification,
so `OfNat.ofNat _ n _` always has a raw nat literal at position 1.
-/
let .lit (.natVal k) := n | toVar e
return .num k
| BitVec.ofNat _ n =>
let some k getNatValue? n | toVar e
let .lit (.natVal k) := n | toVar e
return .num k
| _ => toVar e
let toTopVar (e : Expr) : m (Option RingExpr) := do
@@ -88,7 +107,7 @@ partial def reifyCore? (e : Expr) (skipVar : Bool) (gen : Nat) : m (Option RingE
else
return some ( toVar e)
let asTopVar (e : Expr) : m (Option RingExpr) := do
reportAppIssue e
reportRingAppIssue e
toTopVar e
match_expr e with
| HAdd.hAdd _ _ _ i a b =>
@@ -98,54 +117,46 @@ partial def reifyCore? (e : Expr) (skipVar : Bool) (gen : Nat) : m (Option RingE
| HSub.hSub _ _ _ i a b =>
if ( isSubInst i) then return some (.sub ( go a) ( go b)) else asTopVar e
| HPow.hPow _ _ _ i a b =>
let some k getNatValue? b | asTopVar e
let some k := Sym.getNatValue? b |>.run | asTopVar e
if ( isPowInst i) then return some (.pow ( go a) k) else asTopVar e
| Neg.neg _ i a =>
if ( isNegInst i) then return some (.neg ( go a)) else asTopVar e
| IntCast.intCast _ i a =>
if ( isIntCastInst i) then
let some k getIntValue? a | toTopVar e
let some k := Sym.getIntValue? a |>.run | toTopVar e
return some (.intCast k)
else
asTopVar e
| NatCast.natCast _ i a =>
if ( isNatCastInst i) then
let some k getNatValue? a | toTopVar e
let some k := Sym.getNatValue? a |>.run | toTopVar e
return some (.natCast k)
else
asTopVar e
| OfNat.ofNat _ n _ =>
let some k getNatValue? n | asTopVar e
let .lit (.natVal k) := n | asTopVar e
return some (.num k)
| _ => toTopVar e
/-- Reify ring expression. -/
def reify? (e : Expr) (skipVar := true) (gen : Nat := 0) : RingM (Option RingExpr) := do
reifyCore? e skipVar gen
end RingReify
/-- Reify non-commutative ring expression. -/
def ncreify? (e : Expr) (skipVar := true) (gen : Nat := 0) : NonCommRingM (Option RingExpr) := do
reifyCore? e skipVar gen
section SemiringReify
private def reportSAppIssue (e : Expr) : GoalM Unit := do
variable [MonadLiftT SymM m] [MonadLiftT MetaM m] [MonadError m] [Monad m] [MonadCanon m] [MonadSemiring m] [MonadMkVar m]
private def reportSemiringAppIssue [MonadLiftT SymM m] (e : Expr) : m Unit := do
reportIssue! "semiring term with unexpected instance{indentExpr e}"
section
variable [MonadLiftT GoalM m] [MonadError m] [Monad m] [MonadCanon m] [MonadSemiring m] [MonadSetTermId m]
/--
Similar to `reify?` but for `CommSemiring`
Converts a Lean expression `e` into a `SemiringExpr`.
Only recognizes `add`, `mul`, `pow`, `natCast`, and numerals (no `sub`, `neg`, `intCast`).
-/
partial def sreifyCore? (e : Expr) : m (Option SemiringExpr) := do
let mkVar (e : Expr) : m Var := do
unless ( alreadyInternalized e) do
internalize e 0
mkSVarCore e
partial def reifySemiring? (e : Expr) (gen : Nat) : m (Option SemiringExpr) := do
let toVar (e : Expr) : m SemiringExpr := do
return .var ( mkVar e)
return .var ( mkVar e gen)
let asVar (e : Expr) : m SemiringExpr := do
reportSAppIssue e
return .var ( mkVar e)
reportSemiringAppIssue e
return .var ( mkVar e gen)
let rec go (e : Expr) : m SemiringExpr := do
match_expr e with
| HAdd.hAdd _ _ _ i a b =>
@@ -153,22 +164,22 @@ partial def sreifyCore? (e : Expr) : m (Option SemiringExpr) := do
| HMul.hMul _ _ _ i a b =>
if isSameExpr ( getMulFn').appArg! i then return .mul ( go a) ( go b) else asVar e
| HPow.hPow _ _ _ i a b =>
let some k getNatValue? b | toVar e
let some k := Sym.getNatValue? b |>.run | toVar e
if isSameExpr ( getPowFn').appArg! i then return .pow ( go a) k else asVar e
| NatCast.natCast _ i a =>
if isSameExpr ( getNatCastFn').appArg! i then
let some k getNatValue? a | toVar e
let some k := Sym.getNatValue? a |>.run | toVar e
return .num k
else
asVar e
| OfNat.ofNat _ n _ =>
let some k getNatValue? n | toVar e
let .lit (.natVal k) := n | toVar e
return .num k
| _ => toVar e
let toTopVar (e : Expr) : m (Option SemiringExpr) := do
return some ( toVar e)
let asTopVar (e : Expr) : m (Option SemiringExpr) := do
reportSAppIssue e
reportSemiringAppIssue e
toTopVar e
match_expr e with
| HAdd.hAdd _ _ _ i a b =>
@@ -176,27 +187,19 @@ partial def sreifyCore? (e : Expr) : m (Option SemiringExpr) := do
| HMul.hMul _ _ _ i a b =>
if isSameExpr ( getMulFn').appArg! i then return some (.mul ( go a) ( go b)) else asTopVar e
| HPow.hPow _ _ _ i a b =>
let some k getNatValue? b | return none
let some k := Sym.getNatValue? b |>.run | return none
if isSameExpr ( getPowFn').appArg! i then return some (.pow ( go a) k) else asTopVar e
| NatCast.natCast _ i a =>
if isSameExpr ( getNatCastFn').appArg! i then
let some k getNatValue? a | toTopVar e
let some k := Sym.getNatValue? a |>.run | toTopVar e
return some (.num k)
else
asTopVar e
| OfNat.ofNat _ n _ =>
let some k getNatValue? n | asTopVar e
let .lit (.natVal k) := n | asTopVar e
return some (.num k)
| _ => toTopVar e
end
end SemiringReify
/-- Reify semiring expression. -/
def sreify? (e : Expr) : SemiringM (Option SemiringExpr) := do
sreifyCore? e
/-- Reify non-commutative semiring expression. -/
def ncsreify? (e : Expr) : NonCommSemiringM (Option SemiringExpr) := do
sreifyCore? e
end Lean.Meta.Grind.Arith.CommRing
end Lean.Meta.Sym.Arith

View File

@@ -8,7 +8,7 @@ prelude
public import Init.Grind.Ring.CommSemiringAdapter
public import Lean.ToExpr
public section
namespace Lean.Meta.Grind.Arith.CommRing
namespace Lean.Meta.Sym.Arith
open Grind.CommRing
/-!
`ToExpr` instances for `CommRing.Poly` types.
@@ -57,4 +57,4 @@ instance : ToExpr CommRing.Expr where
toExpr := ofRingExpr
toTypeExpr := mkConst ``CommRing.Expr
end Lean.Meta.Grind.Arith.CommRing
end Lean.Meta.Sym.Arith

View File

@@ -0,0 +1,160 @@
/-
Copyright (c) 2026 Amazon.com, Inc. or its affiliates. All Rights Reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Leonardo de Moura
-/
module
prelude
public import Init.Grind.Ring.CommSemiringAdapter
public import Lean.Meta.Sym.SymM
public section
namespace Lean.Meta.Sym.Arith
export Lean.Grind.CommRing (Var Power Mon Poly)
abbrev RingExpr := Grind.CommRing.Expr
/-
**Note**: recall that we use ring expressions to represent semiring expressions,
and ignore non-applicable constructors.
-/
abbrev SemiringExpr := Grind.CommRing.Expr
/-- Classification state for a type with a `Semiring` instance. -/
structure Semiring where
id : Nat
type : Expr
/-- Cached `getDecLevel type` -/
u : Level
/-- `Semiring` instance for `type` -/
semiringInst : Expr
addFn? : Option Expr := none
mulFn? : Option Expr := none
powFn? : Option Expr := none
natCastFn? : Option Expr := none
deriving Inhabited
/-- Classification state for a type with a `Ring` instance. -/
structure Ring where
id : Nat
type : Expr
/-- Cached `getDecLevel type` -/
u : Level
/-- `Ring` instance for `type` -/
ringInst : Expr
/-- `Semiring` instance for `type` -/
semiringInst : Expr
/-- `IsCharP` instance for `type` if available. -/
charInst? : Option (Expr × Nat)
addFn? : Option Expr := none
mulFn? : Option Expr := none
subFn? : Option Expr := none
negFn? : Option Expr := none
powFn? : Option Expr := none
intCastFn? : Option Expr := none
natCastFn? : Option Expr := none
one? : Option Expr := none
deriving Inhabited
/-- Classification state for a type with a `CommRing` instance. -/
structure CommRing extends Ring where
/-- Inverse function if `fieldInst?` is `some inst` -/
invFn? : Option Expr := none
/--
If this is a `OfSemiring.Q α` ring, this field contains the
`semiringId` for `α`.
-/
semiringId? : Option Nat
/-- `CommSemiring` instance for `type` -/
commSemiringInst : Expr
/-- `CommRing` instance for `type` -/
commRingInst : Expr
/-- `NoNatZeroDivisors` instance for `type` if available. -/
noZeroDivInst? : Option Expr
/-- `Field` instance for `type` if available. -/
fieldInst? : Option Expr
/-- `PowIdentity` instance, the synthesized `CommSemiring` instance, and exponent `p` if available. -/
powIdentityInst? : Option (Expr × Expr × Nat) := none
deriving Inhabited
/--
Classification state for a type with a `CommSemiring` instance.
Recall that `CommSemiring` types are normalized using the `OfSemiring.Q` envelope.
-/
structure CommSemiring extends Semiring where
/-- Id of the envelope ring `OfSemiring.Q type` -/
ringId : Nat
/-- `CommSemiring` instance for `type` -/
commSemiringInst : Expr
/-- `AddRightCancel` instance for `type` if available. -/
addRightCancelInst? : Option Expr := none
toQFn? : Option Expr := none
deriving Inhabited
/-- Result of classifying a type's algebraic structure. -/
inductive ClassifyResult where
| commRing (id : Nat)
| nonCommRing (id : Nat)
| commSemiring (id : Nat)
| nonCommSemiring (id : Nat)
| /-- No algebraic structure found. -/ none
deriving Inhabited
def ClassifyResult.toOption : ClassifyResult Option Nat
| .commRing id | .nonCommRing id
| .commSemiring id | .nonCommSemiring id => some id
| .none => .none
/-- Arith type classification state, stored as a `SymExtension`. -/
structure State where
/-- Exponent threshold for `HPow` evaluation. -/
exp : Nat := 8
/-- Commutative rings. -/
rings : Array CommRing := {}
/-- Commutative semirings. -/
semirings : Array CommSemiring := {}
/-- Non-commutative rings. -/
ncRings : Array Ring := {}
/-- Non-commutative semirings. -/
ncSemirings : Array Semiring := {}
/-- Mapping from types to their classification result. Caches failures as `.none`. -/
typeClassify : PHashMap ExprPtr ClassifyResult := {}
deriving Inhabited
builtin_initialize arithExt : SymExtension State registerSymExtension (return {})
def getArithState : SymM State :=
arithExt.getState
@[inline] def modifyArithState (f : State State) : SymM Unit :=
arithExt.modifyState f
/-- Get the exponent threshold. -/
def getExpThreshold : SymM Nat :=
return ( getArithState).exp
/-- Set the exponent threshold. -/
def setExpThreshold (exp : Nat) : SymM Unit :=
modifyArithState fun s => { s with exp }
/-- Run `k` with a temporary exponent threshold. -/
def withExpThreshold (exp : Nat) (k : SymM α) : SymM α := do
let oldExp := ( getArithState).exp
setExpThreshold exp
try k finally setExpThreshold oldExp
def getCommRingOfId (id : Nat) : SymM CommRing := do
let s getArithState
return s.rings[id]!
def getNonCommRingOfId (id : Nat) : SymM Ring := do
let s getArithState
return s.ncRings[id]!
def getCommSemiringOfId (id : Nat) : SymM CommSemiring := do
let s getArithState
return s.semirings[id]!
def getNonCommSemiringOfId (id : Nat) : SymM Semiring := do
let s getArithState
return s.ncSemirings[id]!
end Lean.Meta.Sym.Arith

View File

@@ -24,10 +24,19 @@ builtin_initialize registerTraceClass `sym.debug.canon
Canonicalizes expressions by normalizing types and instances. At the top level, it traverses
applications, foralls, lambdas, and let-bindings, classifying each argument as a type, instance,
implicit, or value using `shouldCanon`. Values are recursively visited but not normalized.
Types and instances receive targeted reductions.
implicit, or value using `shouldCanon`. Targeted reductions (projection, match/ite/cond, Nat
arithmetic) are applied to all positions; instances are re-synthesized.
## Reductions (applied only in type positions)
**Note about types:** `grind` is not built for reasoning about types that are not propositions.
We assume that definitionally equal types will be structurally identical after we apply the
canonicalizer. We also erase most of the subsingleton markers occurring inside types.
## Reductions
It also normalizes expressions using the following reductions. We can view it as a cheap/custom `dsimp`.
We used to reduce only terms inside types, but it restricted important normalizations that were important
when, for example, a term had a forward dependency. That is, the term is not directly in a type, but
there is a type that depends on it.
- **Eta**: `fun x => f x` → `f`
- **Projection**: `⟨a, b⟩.1` → `a` (structure projections, not class projections)
@@ -35,11 +44,30 @@ Types and instances receive targeted reductions.
- **Nat arithmetic**: ground evaluation (`2 + 1` → `3`) and offset normalization
(`n.succ + 1` → `n + 2`)
**Note**: Eta is applied only if the lambda is occurring inside of a type. For lambdas occurring
in non type positions, we want to leverage the support in `grind` for lambda-expressions.
## Instance canonicalization
Instances are re-synthesized via `synthInstance`. The instance type is first normalized
using the type-level reductions above, so that `OfNat (Fin (2+1)) 0` and `OfNat (Fin 3) 0`
produce the same canonical instance.
produce the same canonical instance. Two special cases:
- **`Decidable` instances** (`Grind.nestedDecidable`): the proposition is recursively
canonicalized, then the `Decidable` instance is re-synthesized. If resynthesis fails,
the original instance is kept (users often provide these via `haveI`).
A `checkDefEqInst` guard is required because structurally different `Decidable` instances
are not necessarily definitionally equal.
- **Propositional instances** (`Grind.nestedProof`): the proposition is recursively
canonicalized, then the proof is re-synthesized. If resynthesis fails, the original
proof is kept. No definitional-equality check is needed thanks to proof irrelevance.
In both cases, resynthesis failure is silent — the original instance or proof is kept.
Ideally we would report an issue when resynthesis fails inside a type (where structural
agreement matters most), but in practice users provide non-synthesizable instances via `haveI`,
and these instances propagate into types through forward dependencies. Reporting failures
for such instances produces noise that obscures real issues.
## Two caches
@@ -213,7 +241,7 @@ def checkDefEqInst (e : Expr) (inst : Expr) : SymM Expr := do
return e
return inst
/-- Canonicalize `e`. Applies targeted reductions in type positions; recursively visits value positions. -/
/-- Canonicalize `e`. Applies targeted reductions and re-synthesizes instances. -/
partial def canon (e : Expr) : CanonM Expr := do
match e with
| .forallE .. => withCaching e <| canonForall #[] e
@@ -246,23 +274,91 @@ where
else
withReader (fun ctx => { ctx with insideType := true }) <| canon e
canonInst (e : Expr) : CanonM Expr := do
if let some inst := ( get).canon.cacheInsts.get? e then
checkDefEqInst e inst
/--
Canonicalize `e : type` where `e` is an instance by trying to resynthesize `type`.
If `report` is `true`, we report an issue when the instance cannot be resynthesized.
-/
canonInstCore (e : Expr) (type : Expr) (report := true) : CanonM Expr := do
let some inst Sym.synthInstance? type |
if report then
reportIssue! "failed to canonicalize instance{indentExpr e}\nfailed to synthesize{indentExpr type}"
return e
checkDefEqInst e inst
/--
Canonicalize an instance by trying to resynthesize it without caching.
Recall that we have special support for `Decidable` and propositional instances.
-/
canonInst' (e : Expr) (report := true) : CanonM Expr := do
/-
We normalize the type to make sure `OfNat (Fin (2+1)) 1` and `OfNat (Fin 3) 1` will produce
the same instances.
-/
let type inferType e
let type' canonInsideType' type
canonInstCore e type' report
/-- `withCaching` + `canonInst'` -/
canonInst (e : Expr) (report := true) : CanonM Expr := withCaching e do
canonInst' e report
/--
Canonicalize a proposition that is also a term instance.
Given a term `e` of the form `@Grind.nestedProof prop h`, where `g` is the constant `Grind.nestedProof`,
we canonicalize it as follows:
1- We recursively canonicalize the proposition `prop`.
2- Try to resynthesize the instance, but keep the original one in case of failure since users often
provide them using `haveI`.
-/
canonInstProp (g : Expr) (prop : Expr) (h : Expr) (e : Expr) : CanonM Expr := withCaching e do
let prop' canon prop
if ( read).insideType then
/- We suppress reporting here because `haveI`-provided instances propagate into types
through forward dependencies, and reporting them produces noise. See module doc. -/
canonInstCore h prop' (report := false)
else
/-
We normalize the type to make sure `OfNat (Fin (2+1)) 1` and `OfNat (Fin 3) 1` will produce
the same instances.
**Note**: We try to resynthesize the proposition, but if it fails we keep the current one.
This may happen because propositional instances are often provided manually using `haveI`.
-/
let type inferType e
let type' canonInsideType' type
let some inst Sym.synthInstance? type' |
reportIssue! "failed to canonicalize instance{indentExpr e}\nfailed to synthesize{indentExpr type'}"
return e
let inst checkDefEqInst e inst
-- Remark: we cache result using the type **before** canonicalization.
modify fun s => { s with canon.cacheInsts := s.canon.cacheInsts.insert e inst }
return inst
let h' := ( Sym.synthInstance? prop').getD h
/- **Note**: We don't need to check whether `h` and `h'` are definitionally equal because of proof irrelevance. -/
return if isSameExpr prop prop' && isSameExpr h h' then e else mkApp2 g prop' h'
/--
Canonicalize a decidable instance without checking the cache.
Given a term `e` of the form `@Grind.nestedDecidable prop inst`, where `g` is the constant `Grind.nestedDecidable`,
we canonicalize it as follows:
1- We recursively canonicalize the proposition `prop`.
2- Try to resynthesize the instance, but keep the original one in case of failure since users often
provide them using `haveI`.
-/
canonInstDec' (g : Expr) (prop : Expr) (inst : Expr) (e : Expr) : CanonM Expr := do
let prop' canon prop
let type := mkApp (mkConst ``Decidable) prop'
if ( read).insideType then
/- See comment in `canonInstProp` for why we suppress reporting here. -/
canonInstCore inst type (report := false)
else
/-
**Note**: We try to resynthesize the instance, but if it fails we keep the current one.
We use `checkDefEqInst` here because two structurally different decidable instances are not necessarily
definitionally equal.
This may happen because propositional instances are often provided manually using `haveI`.
-/
let inst' := ( Sym.synthInstance? type).getD inst
let inst' checkDefEqInst inst inst'
return if isSameExpr prop prop' && isSameExpr inst inst' then e else mkApp2 g prop' inst'
/-- `withCaching` + `canonInstDec'` -/
canonInstDec (g : Expr) (prop : Expr) (h : Expr) (e : Expr) : CanonM Expr := withCaching e do
canonInstDec' g prop h e
/-- `canonInstDec` variant for normalizing `if-then-else` expressions. -/
canonInstDecCore (e : Expr) : CanonM Expr := do
match_expr e with
| g@Grind.nestedDecidable prop inst => canonInstDec g prop inst e
| _ => canonInst e (report := false)
canonLambda (e : Expr) : CanonM Expr := do
if ( read).insideType then
@@ -295,60 +391,56 @@ where
mkLetFVars (generalizeNondepLet := false) fvars ( canon (e.instantiateRev fvars))
canonAppDefault (e : Expr) : CanonM Expr := e.withApp fun f args => do
if f.isConstOf ``Grind.nestedProof && args.size == 2 then
let prop := args[0]!
let prop' canon prop
let e' := if isSameExpr prop prop' then e else mkAppN f (args.set! 0 prop')
return e'
else if f.isConstOf ``Grind.nestedDecidable && args.size == 2 then
let prop := args[0]!
let prop' canon prop
let e' := if isSameExpr prop prop' then e else mkAppN f (args.set! 0 prop')
return e'
if args.size == 2 then
if f.isConstOf ``Grind.nestedProof then
/- **Note**: We don't have special treatment if `e` inside a type. -/
let prop := args[0]!
let prop' canon prop
let e' := if isSameExpr prop prop' then e else mkApp2 f prop' args[1]!
return e'
else if f.isConstOf ``Grind.nestedDecidable then
return ( canonInstDec' f args[0]! args[1]! e)
let mut modified := false
let args if f.isConstOf ``OfNat.ofNat then
let some args normOfNatArgs? args | pure args
modified := true
pure args
else
let mut modified := false
let args if f.isConstOf ``OfNat.ofNat then
let some args normOfNatArgs? args | pure args
pure args
let mut f := f
let f' canon f
unless isSameExpr f f' do
f := f'
modified := true
let pinfos := ( getFunInfo f).paramInfo
let mut args := args.toVector
for h : i in *...args.size do
let arg := args[i]
trace[sym.debug.canon] "[{repr (← shouldCanon pinfos i arg)}]: {arg} : {← inferType arg}"
let arg' match ( shouldCanon pinfos i arg) with
| .canonType =>
/-
The type may have nested propositions and terms that may need to be canonicalized too.
So, we must recurse over it. See issue #10232
-/
canonInsideType' arg
| .canonImplicit => canon arg
| .visit => canon arg
| .canonInst =>
match_expr arg with
| g@Grind.nestedDecidable prop h => canonInstDec g prop h arg
| g@Grind.nestedProof prop h => canonInstProp g prop h arg
| _ => canonInst arg
unless isSameExpr arg arg' do
args := args.set i arg'
modified := true
pure args
else
pure args
let mut f := f
let f' canon f
unless isSameExpr f f' do
f := f'
modified := true
let pinfos := ( getFunInfo f).paramInfo
let mut args := args.toVector
for h : i in *...args.size do
let arg := args[i]
trace[sym.debug.canon] "[{repr (← shouldCanon pinfos i arg)}]: {arg} : {← inferType arg}"
let arg' match ( shouldCanon pinfos i arg) with
| .canonType =>
/-
The type may have nested propositions and terms that may need to be canonicalized too.
So, we must recurse over it. See issue #10232
-/
canonInsideType' arg
| .canonImplicit => canon arg
| .visit => canon arg
| .canonInst =>
if arg.isAppOfArity ``Grind.nestedDecidable 2 then
let prop := arg.appFn!.appArg!
let prop' canon prop
if isSameExpr prop prop' then pure arg else pure (mkApp2 arg.appFn!.appFn! prop' arg.appArg!)
else
canonInst arg
unless isSameExpr arg arg' do
args := args.set i arg'
modified := true
return if modified then mkAppN f args.toArray else e
return if modified then mkAppN f args.toArray else e
canonIte (f : Expr) (α c inst a b : Expr) : CanonM Expr := do
let c canon c
if isTrueCond c then canon a
else if isFalseCond c then canon b
else return mkApp5 f ( canonInsideType α) c ( canonInst inst) ( canon a) ( canon b)
else return mkApp5 f ( canonInsideType α) c ( canonInstDecCore inst) ( canon a) ( canon b)
canonCond (f : Expr) (α c a b : Expr) : CanonM Expr := do
let c canon c
@@ -389,30 +481,24 @@ where
return e
canonApp (e : Expr) : CanonM Expr := do
if ( read).insideType then
match_expr e with
| f@ite α c i a b => canonIte f α c i a b
| f@cond α c a b => canonCond f α c a b
-- Remark: We currently don't normalize dependent-if-then-else occurring in types.
| _ =>
let f := e.getAppFn
let .const declName _ := f | canonAppAndPost e
if ( isMatcher declName) then
canonMatch e
else
canonAppAndPost e
else
canonAppDefault e
match_expr e with
| f@ite α c i a b => canonIte f α c i a b
| f@cond α c a b => canonCond f α c a b
-- Remark: We currently don't normalize dependent-if-then-else occurring in types.
| _ =>
let f := e.getAppFn
let .const declName _ := f | canonAppAndPost e
if ( isMatcher declName) then
canonMatch e
else
canonAppAndPost e
canonProj (e : Expr) : CanonM Expr := do
let e := e.updateProj! ( canon e.projExpr!)
if ( read).insideType then
return ( reduceProj? e).getD e
else
return e
return ( reduceProj? e).getD e
/--
Returns `true` if `shouldCannon pinfos i arg` is not `.visit`.
Returns `true` if `shouldCanon pinfos i arg` is not `.visit`.
This is a helper function used to implement mbtc.
-/
public def isSupport (pinfos : Array ParamInfo) (i : Nat) (arg : Expr) : MetaM Bool := do
@@ -423,8 +509,8 @@ end Canon
/--
Canonicalize `e` by normalizing types, instances, and support arguments.
Types receive targeted reductions (eta, projection, match/ite, Nat arithmetic).
Instances are re-synthesized. Values are traversed but not reduced.
Applies targeted reductions (projection, match/ite/cond, Nat arithmetic) in all positions;
eta reduction is applied only inside types. Instances are re-synthesized.
Runs at reducible transparency.
-/
public def canon (e : Expr) : SymM Expr := do profileitM Exception "sym canon" ( getOptions) do

View File

@@ -86,22 +86,8 @@ It assumes the input is maximally shared, and ensures the output is too.
public def instantiateS (e : Expr) (subst : Array Expr) : SymM Expr :=
liftBuilderM <| instantiateS' e subst
/-- `mkAppRevRangeS f b e args == mkAppRev f (revArgs.extract b e)` -/
def mkAppRevRangeS (f : Expr) (beginIdx endIdx : Nat) (revArgs : Array Expr) : AlphaShareBuilderM Expr :=
loop revArgs beginIdx f endIdx
where
loop (revArgs : Array Expr) (start : Nat) (b : Expr) (i : Nat) : AlphaShareBuilderM Expr := do
if i start then
return b
else
let i := i - 1
loop revArgs start ( mkAppS b revArgs[i]!) i
/--
Beta-reduces `f` applied to reversed arguments `revArgs`, ensuring maximally shared terms.
`betaRevS f #[a₃, a₂, a₁]` computes the beta-normal form of `f a₁ a₂ a₃`.
-/
partial def betaRevS (f : Expr) (revArgs : Array Expr) : AlphaShareBuilderM Expr :=
/-- Internal variant of `betaRevS` that runs in `AlphaShareBuilderM`. -/
private partial def betaRevS' (f : Expr) (revArgs : Array Expr) : AlphaShareBuilderM Expr :=
if revArgs.size == 0 then
return f
else
@@ -173,7 +159,7 @@ where
| .bvar bidx =>
let f' visitBVar f bidx offset
if modified || !isSameExpr f f' then
betaRevS f' argsRev
betaRevS' f' argsRev
else
return e
| _ => unreachable!
@@ -215,4 +201,18 @@ public def instantiateRevBetaS (e : Expr) (subst : Array Expr) : SymM Expr := do
if !e.hasLooseBVars || subst.isEmpty then return e
else liftBuilderM <| instantiateRevBetaS' e subst
/--
Beta-reduces `f` applied to reversed arguments `revArgs`, ensuring maximally shared terms.
`betaRevS f #[a₃, a₂, a₁]` computes the beta-normal form of `f a₁ a₂ a₃`.
-/
public def betaRevS (f : Expr) (revArgs : Array Expr) : SymM Expr :=
liftBuilderM <| betaRevS' f revArgs
/--
Apply the given arguments to `f`, beta-reducing if `f` is a lambda expression,
ensuring maximally shared terms. See `betaRevS` for details.
-/
public def betaS (f : Expr) (args : Array Expr) : SymM Expr :=
betaRevS f args.reverse
end Lean.Meta.Sym

View File

@@ -152,8 +152,6 @@ structure Canon.State where
cache : Std.HashMap Expr Expr := {}
/-- Cache for type-level canonicalization (reductions applied). -/
cacheInType : Std.HashMap Expr Expr := {}
/-- Cache mapping instances to their canonical synthesized instances. -/
cacheInsts : Std.HashMap Expr Expr := {}
/-- Mutable state for the symbolic computation framework. -/
structure State where

View File

@@ -44,7 +44,7 @@ def isCbvNoncomputable (p : Name) : CoreM Bool := do
return evalLemmas.isNone && Lean.isNoncomputable ( getEnv) p
/--
Attemps to synthesize `Decidable p` instance and guards against picking up a `noncomputable` instance
Attempts to synthesize `Decidable p` instance and guards against picking up a `noncomputable` instance
-/
def trySynthComputableInstance (p : Expr) : SymM <| Option Expr := do
let .some inst' trySynthInstance (mkApp (mkConst ``Decidable) p) | return .none
@@ -112,7 +112,7 @@ builtin_cbv_simproc ↓ simpIteCbv (@ite _ _ _ _ _) := fun e => do
else if ( isFalseExpr c') then
return .step b (mkApp (e.replaceFn ``ite_cond_eq_false) h) (contextDependent := cd)
else
-- If we got stuck with simplifying `p` then let's try evaluating the original isntance
-- If we got stuck with simplifying `p` then let's try evaluating the original instance
simpAndMatchIteDecidable f α c inst a b do
-- If we get stuck here, maybe the problem is that we need to look at `Decidable c'`
let inst' := mkApp4 (mkConst ``decidable_of_decidable_of_eq) c c' inst h
@@ -317,7 +317,7 @@ public def reduceRecMatcher : Simproc := fun e => do
else
return .rfl
builtin_cbv_simproc simpDecidableRec (@Decidable.rec _ _ _ _ _) :=
builtin_cbv_simproc simpDecidableRec (@Decidable.rec _ _ _ _ _) :=
(simpInterlaced · #[false,false,false,false,true]) >> reduceRecMatcher
def tryMatchEquations (appFn : Name) : Simproc := fun e => do

View File

@@ -272,7 +272,7 @@ def handleProj : Simproc := fun e => do
let reduced Sym.share reduced
return .step reduced ( Sym.mkEqRefl reduced)
| .none =>
-- If we failed to reduce it, we turn to a last resort; we try use heterogenous congruence lemma that we then try to turn into an equality.
-- If we failed to reduce it, we turn to a last resort; we try use heterogeneous congruence lemma that we then try to turn into an equality.
unless ( isDefEq struct e') do
-- If we rewrote the projection body using something that holds up to propositional equality, then there is nothing we can do.
-- TODO: Check if there is a need to report this to a user, or shall we fail silently.
@@ -283,6 +283,7 @@ def handleProj : Simproc := fun e => do
let newProof mkEqOfHEq newProof (check := false)
return .step ( Lean.Expr.updateProjS! e e') newProof
open Sym.Internal in
/--
For an application whose head is neither a constant nor a lambda (e.g. a projection
like `p.1 x`), simplify the function head and lift the proof via `congrArg`.

View File

@@ -24,9 +24,6 @@ namespace Lean.Meta.Tactic.Cbv
open Lean.Meta.Sym.Simp
public def mkAppNS (f : Expr) (args : Array Expr) : Sym.SymM Expr := do
args.foldlM Sym.Internal.mkAppS f
abbrev isNatValue (e : Expr) : Bool := (Sym.getNatValue? e).isSome
abbrev isStringValue (e : Expr) : Bool := (Sym.getStringValue? e).isSome
abbrev isIntValue (e : Expr) : Bool := (Sym.getIntValue? e).isSome

View File

@@ -5,28 +5,19 @@ Authors: Leonardo de Moura
-/
module
prelude
public import Lean.Meta.Tactic.Grind.Arith.CommRing.Poly
public import Lean.Meta.Tactic.Grind.Arith.CommRing.Types
public import Lean.Meta.Tactic.Grind.Arith.CommRing.RingId
public import Lean.Meta.Tactic.Grind.Arith.CommRing.Internalize
public import Lean.Meta.Tactic.Grind.Arith.CommRing.ToExpr
public import Lean.Meta.Tactic.Grind.Arith.CommRing.RingM
public import Lean.Meta.Tactic.Grind.Arith.CommRing.SemiringM
public import Lean.Meta.Tactic.Grind.Arith.CommRing.NonCommRingM
public import Lean.Meta.Tactic.Grind.Arith.CommRing.NonCommSemiringM
public import Lean.Meta.Tactic.Grind.Arith.CommRing.Functions
public import Lean.Meta.Tactic.Grind.Arith.CommRing.Reify
public import Lean.Meta.Tactic.Grind.Arith.CommRing.EqCnstr
public import Lean.Meta.Tactic.Grind.Arith.CommRing.Proof
public import Lean.Meta.Tactic.Grind.Arith.CommRing.DenoteExpr
public import Lean.Meta.Tactic.Grind.Arith.CommRing.Inv
public import Lean.Meta.Tactic.Grind.Arith.CommRing.PP
public import Lean.Meta.Tactic.Grind.Arith.CommRing.VarRename
public import Lean.Meta.Tactic.Grind.Arith.CommRing.MonadCanon
public import Lean.Meta.Tactic.Grind.Arith.CommRing.MonadRing
public import Lean.Meta.Tactic.Grind.Arith.CommRing.MonadSemiring
public import Lean.Meta.Tactic.Grind.Arith.CommRing.Action
public import Lean.Meta.Tactic.Grind.Arith.CommRing.Power
public import Lean.Meta.Tactic.Grind.Arith.CommRing.Denote
public section
namespace Lean.Meta.Grind.Arith.CommRing
builtin_initialize registerTraceClass `grind.ring

View File

@@ -0,0 +1,32 @@
/-
Copyright (c) 2026 Amazon.com, Inc. or its affiliates. All Rights Reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Leonardo de Moura
-/
module
prelude
public import Lean.Meta.Tactic.Grind.Arith.CommRing.Types
public import Lean.Meta.Sym.Arith.DenoteExpr
namespace Lean.Meta.Grind.Arith.CommRing
open Sym.Arith
/-!
Helper functions for converting reified terms back into their denotations.
-/
variable [Monad M] [MonadGetVar M] [MonadError M] [MonadLiftT MetaM M] [MonadCanon M] [MonadRing M]
def mkEq (a b : Expr) : M Expr := do
let r getRing
return mkApp3 (mkConst ``Eq [r.u.succ]) r.type a b
public def EqCnstr.denoteExpr (c : EqCnstr) : M Expr := do
mkEq ( denotePoly c.p) ( denoteNum 0)
public def PolyDerivation.denoteExpr (d : PolyDerivation) : M Expr := do
denotePoly d.p
public def DiseqCnstr.denoteExpr (c : DiseqCnstr) : M Expr := do
return mkNot ( mkEq ( c.d.denoteExpr) ( denoteNum 0))
end Lean.Meta.Grind.Arith.CommRing

View File

@@ -1,99 +0,0 @@
/-
Copyright (c) 2025 Amazon.com, Inc. or its affiliates. All Rights Reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Leonardo de Moura
-/
module
prelude
public import Lean.Meta.Tactic.Grind.Arith.CommRing.Functions
public section
namespace Lean.Meta.Grind.Arith.CommRing
/-!
Helper functions for converting reified terms back into their denotations.
-/
variable [Monad M] [MonadError M] [MonadLiftT MetaM M] [MonadCanon M] [MonadRing M]
def denoteNum (k : Int) : M Expr := do
let ring getRing
let n := mkRawNatLit k.natAbs
let ofNatInst if let some inst synthInstance? (mkApp2 (mkConst ``OfNat [ring.u]) ring.type n) then
pure inst
else
pure <| mkApp3 (mkConst ``Grind.Semiring.ofNat [ring.u]) ring.type ring.semiringInst n
let n := mkApp3 (mkConst ``OfNat.ofNat [ring.u]) ring.type n ofNatInst
if k < 0 then
return mkApp ( getNegFn) n
else
return n
def _root_.Lean.Grind.CommRing.Power.denoteExpr (pw : Power) : M Expr := do
let x := ( getRing).vars[pw.x]!
if pw.k == 1 then
return x
else
return mkApp2 ( getPowFn) x (toExpr pw.k)
def _root_.Lean.Grind.CommRing.Mon.denoteExpr (m : Mon) : M Expr := do
match m with
| .unit => denoteNum 1
| .mult pw m => go m ( pw.denoteExpr)
where
go (m : Mon) (acc : Expr) : M Expr := do
match m with
| .unit => return acc
| .mult pw m => go m (mkApp2 ( getMulFn) acc ( pw.denoteExpr))
def _root_.Lean.Grind.CommRing.Poly.denoteExpr (p : Poly) : M Expr := do
match p with
| .num k => denoteNum k
| .add k m p => go p ( denoteTerm k m)
where
denoteTerm (k : Int) (m : Mon) : M Expr := do
if k == 1 then
m.denoteExpr
else
return mkApp2 ( getMulFn) ( denoteNum k) ( m.denoteExpr)
go (p : Poly) (acc : Expr) : M Expr := do
match p with
| .num 0 => return acc
| .num k => return mkApp2 ( getAddFn) acc ( denoteNum k)
| .add k m p => go p (mkApp2 ( getAddFn) acc ( denoteTerm k m))
@[specialize]
private def denoteExprCore (getVar : Nat Expr) (e : RingExpr) : M Expr := do
go e
where
go : RingExpr M Expr
| .num k => denoteNum k
| .natCast k => return mkApp ( getNatCastFn) (mkNatLit k)
| .intCast k => return mkApp ( getIntCastFn) (mkIntLit k)
| .var x => return getVar x
| .add a b => return mkApp2 ( getAddFn) ( go a) ( go b)
| .sub a b => return mkApp2 ( getSubFn) ( go a) ( go b)
| .mul a b => return mkApp2 ( getMulFn) ( go a) ( go b)
| .pow a k => return mkApp2 ( getPowFn) ( go a) (toExpr k)
| .neg a => return mkApp ( getNegFn) ( go a)
def _root_.Lean.Grind.CommRing.Expr.denoteExpr (e : RingExpr) : M Expr := do
let ring getRing
denoteExprCore (fun x => ring.vars[x]!) e
def _root_.Lean.Grind.CommRing.Expr.denoteExpr' (vars : Array Expr) (e : RingExpr) : M Expr := do
denoteExprCore (fun x => vars[x]!) e
private def mkEq (a b : Expr) : M Expr := do
let r getRing
return mkApp3 (mkConst ``Eq [r.u.succ]) r.type a b
def EqCnstr.denoteExpr (c : EqCnstr) : M Expr := do
mkEq ( c.p.denoteExpr) ( denoteNum 0)
def PolyDerivation.denoteExpr (d : PolyDerivation) : M Expr := do
d.p.denoteExpr
def DiseqCnstr.denoteExpr (c : DiseqCnstr) : M Expr := do
return mkNot ( mkEq ( c.d.denoteExpr) ( denoteNum 0))
end Lean.Meta.Grind.Arith.CommRing

View File

@@ -9,9 +9,7 @@ public import Lean.Meta.Tactic.Grind.Arith.CommRing.RingId
import Lean.Meta.Tactic.Grind.Diseq
import Lean.Meta.Tactic.Grind.Arith.Util
import Lean.Meta.Tactic.Grind.Arith.CommRing.Proof
import Lean.Meta.Tactic.Grind.Arith.CommRing.DenoteExpr
import Lean.Meta.Tactic.Grind.Arith.CommRing.Inv
import Lean.Meta.Tactic.Grind.Arith.CommRing.Reify
import Lean.Meta.Tactic.Grind.Arith.CommRing.SafePoly
public section
namespace Lean.Meta.Grind.Arith.CommRing

View File

@@ -6,12 +6,16 @@ Authors: Leonardo de Moura
module
prelude
public import Lean.Meta.Tactic.Grind.Arith.CommRing.RingId
import Lean.Meta.Sym.Arith.Reify
import Lean.Meta.Sym.Arith.DenoteExpr
import Lean.Meta.Tactic.Grind.Arith.CommRing.SemiringM
import Lean.Meta.Tactic.Grind.Arith.CommRing.NonCommRingM
import Lean.Meta.Tactic.Grind.Arith.CommRing.NonCommSemiringM
import Lean.Meta.Tactic.Grind.Simp
import Lean.Meta.Tactic.Grind.Arith.Util
import Lean.Meta.Tactic.Grind.Arith.CommRing.Reify
import Lean.Meta.Tactic.Grind.Arith.CommRing.DenoteExpr
public section
namespace Lean.Meta.Grind.Arith.CommRing
open Sym Arith
/-- If `e` is a function application supported by the `CommRing` module, return its type. -/
private def getType? (e : Expr) : Option Expr :=
@@ -80,8 +84,8 @@ private def processInv (e inst a : Expr) : RingM Unit := do
unless ( isInvInst inst) do return ()
let ring getCommRing
let some fieldInst := ring.fieldInst? | return ()
if ( getCommRing).invSet.contains a then return ()
modifyCommRing fun s => { s with invSet := s.invSet.insert a }
if ( getCommRingEntry).invSet.contains a then return ()
modifyCommRingEntry fun s => { s with invSet := s.invSet.insert a }
if let some k toInt? a then
if k == 0 then
/-
@@ -112,6 +116,26 @@ private def processInv (e inst a : Expr) : RingM Unit := do
return ()
pushNewFact <| mkApp3 (mkConst ``Grind.CommRing.inv_split [ring.u]) ring.type fieldInst a
/--
For each new variable `x` in a ring with `PowIdentity α p`,
push the equation `x ^ p = x` as a new fact into grind.
-/
private def processPowIdentityVars : RingM Unit := do
let ring getCommRing
let some (powIdentityInst, csInst, p) := ring.powIdentityInst? | return ()
let ringEntry getCommRingEntry
let startIdx := ringEntry.powIdentityVarCount
let vars := ringEntry.vars
if startIdx >= vars.size then return ()
for i in [startIdx:vars.size] do
let x := vars[i]!
trace_goal[grind.ring] "PowIdentity: pushing x^{p} = x for {x}"
-- Construct proof: @PowIdentity.pow_eq α csInst p powIdentityInst x
let proof := mkApp5 (mkConst ``Grind.PowIdentity.pow_eq [ring.u])
ring.type csInst (mkNatLit p) powIdentityInst x
pushNewFact proof
modifyCommRingEntry fun s => { s with powIdentityVarCount := vars.size }
/-- Returns `true` if `e` is a term `a⁻¹`. -/
private def internalizeInv (e : Expr) : GoalM Bool := do
match_expr e with
@@ -129,32 +153,34 @@ def internalize (e : Expr) (parent? : Option Expr) : GoalM Unit := do
if ( internalizeInv e) then return ()
let some type := getType? e | return ()
if isForbiddenParent parent? then return ()
let gen getGeneration e
if let some ringId getCommRingId? type then RingM.run ringId do
let some re reify? e | return ()
let some re reifyRing? e (gen := gen) | return ()
trace_goal[grind.ring.internalize] "[{ringId}]: {e}"
setTermRingId e
ringExt.markTerm e
modifyCommRing fun s => { s with
modifyCommRingEntry fun s => { s with
denote := s.denote.insert { expr := e } re
denoteEntries := s.denoteEntries.push (e, re)
}
processPowIdentityVars
else if let some semiringId getCommSemiringId? type then SemiringM.run semiringId do
let some re sreify? e | return ()
let some re reifySemiring? e (gen := gen) | return ()
trace_goal[grind.ring.internalize] "semiring [{semiringId}]: {e}"
setTermSemiringId e
ringExt.markTerm e
modifySemiring fun s => { s with denote := s.denote.insert { expr := e } re }
modifyCommSemiringEntry fun s => { s with denote := s.denote.insert { expr := e } re }
else if let some ncRingId getNonCommRingId? type then NonCommRingM.run ncRingId do
let some re ncreify? e | return ()
let some re reifyRing? e (gen := gen) | return ()
trace_goal[grind.ring.internalize] "(non-comm) ring [{ncRingId}]: {e}"
setTermNonCommRingId e
ringExt.markTerm e
modifyRing fun s => { s with denote := s.denote.insert { expr := e } re }
modifyRingEntry fun s => { s with denote := s.denote.insert { expr := e } re }
else if let some ncSemiringId getNonCommSemiringId? type then NonCommSemiringM.run ncSemiringId do
let some re ncsreify? e | return ()
let some re reifySemiring? e (gen := gen) | return ()
trace_goal[grind.ring.internalize] "(non-comm) semiring [{ncSemiringId}]: {e}"
setTermNonCommSemiringId e
ringExt.markTerm e
modifySemiring fun s => { s with denote := s.denote.insert { expr := e } re }
modifySemiringEntry fun s => { s with denote := s.denote.insert { expr := e } re }
end Lean.Meta.Grind.Arith.CommRing

View File

@@ -6,10 +6,10 @@ Authors: Leonardo de Moura
module
prelude
public import Lean.Meta.Tactic.Grind.Arith.CommRing.RingM
import Lean.Meta.Tactic.Grind.Arith.CommRing.Poly
import Lean.Meta.Sym.Arith.Poly
public section
namespace Lean.Meta.Grind.Arith.CommRing
/-
private def checkVars : RingM Unit := do
let s ← getRing
let mut num := 0
@@ -42,11 +42,14 @@ private def checkDiseqs : RingM Unit := do
for c in (← getCommRing).diseqs do
checkPoly c.d.p
-/
private def checkRingInvs : RingM Unit := do
checkVars
checkBasis
checkQueue
checkDiseqs
-- checkVars
-- checkBasis
-- checkQueue
-- checkDiseqs
return ()
def checkInvariants : GoalM Unit := do
if ( isDebugEnabled) then

View File

@@ -1,37 +0,0 @@
/-
Copyright (c) 2025 Amazon.com, Inc. or its affiliates. All Rights Reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Leonardo de Moura
-/
module
prelude
public import Lean.Meta.Tactic.Grind.Arith.CommRing.Types
public section
namespace Lean.Meta.Grind.Arith.CommRing
class MonadCanon (m : Type Type) where
/--
Helper function for removing dependency on `GoalM`.
In `RingM` and `SemiringM`, this is just `sharedCommon (← canon e)`
When printing counterexamples, we are at `MetaM`, and this is just the identity function.
-/
canonExpr : Expr m Expr
/--
Helper function for removing dependency on `GoalM`. During search we
want to track the instances synthesized by `grind`, and this is `Grind.synthInstance`.
-/
synthInstance? : Expr m (Option Expr)
export MonadCanon (canonExpr)
@[always_inline]
instance (m n) [MonadLift m n] [MonadCanon m] : MonadCanon n where
canonExpr e := liftM (canonExpr e : m Expr)
synthInstance? e := liftM (MonadCanon.synthInstance? e : m (Option Expr))
def MonadCanon.synthInstance [Monad m] [MonadError m] [MonadCanon m] (type : Expr) : m Expr := do
let some inst synthInstance? type
| throwError "`grind` failed to find instance{indentExpr type}"
return inst
end Lean.Meta.Grind.Arith.CommRing

View File

@@ -8,31 +8,31 @@ prelude
public import Lean.Meta.Tactic.Grind.Arith.CommRing.RingM
public section
namespace Lean.Meta.Grind.Arith.CommRing
open Sym.Arith
structure NonCommRingM.Context where
ringId : Nat
symRingId : Nat
/-- We don't want to keep carrying the `RingId` around. -/
abbrev NonCommRingM := ReaderT NonCommRingM.Context GoalM
abbrev NonCommRingM.run (ringId : Nat) (x : NonCommRingM α) : GoalM α :=
x { ringId }
instance : MonadCanon NonCommRingM where
canonExpr e := do shareCommon ( canon e)
synthInstance? e := Grind.synthInstance? e
abbrev NonCommRingM.run (ringId : Nat) (x : NonCommRingM α) : GoalM α := do
let s get'
let symRingId := s.ncRings[ringId]!.symId
x { ringId, symRingId }
protected def NonCommRingM.getRing : NonCommRingM Ring := do
let s get'
let ringId := ( read).ringId
let s getArithState
let ringId := ( read).symRingId
if h : ringId < s.ncRings.size then
return s.ncRings[ringId]
else
throwError "`grind` internal error, invalid ringId"
protected def NonCommRingM.modifyRing (f : Ring Ring) : NonCommRingM Unit := do
let ringId := ( read).ringId
modify' fun s => { s with ncRings := s.ncRings.modify ringId f }
let ringId := ( read).symRingId
modifyArithState fun s => { s with ncRings := s.ncRings.modify ringId f }
instance : MonadRing NonCommRingM where
getRing := NonCommRingM.getRing
@@ -49,7 +49,37 @@ def setTermNonCommRingId (e : Expr) : NonCommRingM Unit := do
return ()
modify' fun s => { s with exprToNCRingId := s.exprToNCRingId.insert { expr := e } ringId }
instance : MonadSetTermId NonCommRingM where
setTermId e := setTermNonCommRingId e
def getRingEntry : NonCommRingM RingEntry := do
let s get'
let ringId := ( read).ringId
if h : ringId < s.ncRings.size then
return s.ncRings[ringId]
else
throwError "`grind` internal error, invalid ringId"
def modifyRingEntry (f : RingEntry RingEntry) : NonCommRingM Unit := do
let ringId := ( read).ringId
modify' fun s => { s with ncRings := s.ncRings.modify ringId f }
def mkNonCommVar (e : Expr) (gen : Nat) : NonCommRingM Var := do
unless ( alreadyInternalized e) do
internalize e gen
let s getRingEntry
if let some var := s.varMap.find? { expr := e } then
return var
let var : Var := s.vars.size
modifyRingEntry fun s => { s with
vars := s.vars.push e
varMap := s.varMap.insert { expr := e } var
}
setTermNonCommRingId e
ringExt.markTerm e
return var
instance : MonadMkVar NonCommRingM where
mkVar := mkNonCommVar
instance : MonadGetVar NonCommRingM where
getVar x := return ( getRingEntry).vars[x]!
end Lean.Meta.Grind.Arith.CommRing

View File

@@ -8,20 +8,36 @@ prelude
public import Lean.Meta.Tactic.Grind.Arith.CommRing.SemiringM
public section
namespace Lean.Meta.Grind.Arith.CommRing
open Sym.Arith
structure NonCommSemiringM.Context where
semiringId : Nat
symSemiringId : Nat
abbrev NonCommSemiringM := ReaderT NonCommSemiringM.Context GoalM
abbrev NonCommSemiringM.run (semiringId : Nat) (x : NonCommSemiringM α) : GoalM α :=
x { semiringId }
instance : MonadCanon NonCommSemiringM where
canonExpr e := do shareCommon ( canon e)
synthInstance? e := Grind.synthInstance? e
abbrev NonCommSemiringM.run (semiringId : Nat) (x : NonCommSemiringM α) : GoalM α := do
let s get'
let symSemiringId := s.ncSemirings[semiringId]!.symId
x { semiringId, symSemiringId }
protected def NonCommSemiringM.getSemiring : NonCommSemiringM Semiring := do
let s getArithState
let semiringId := ( read).symSemiringId
if h : semiringId < s.ncSemirings.size then
return s.ncSemirings[semiringId]
else
throwError "`grind` internal error, invalid semiringId"
protected def NonCommSemiringM.modifySemiring (f : Semiring Semiring) : NonCommSemiringM Unit := do
let semiringId := ( read).symSemiringId
modifyArithState fun s => { s with ncSemirings := s.ncSemirings.modify semiringId f }
instance : MonadSemiring NonCommSemiringM where
getSemiring := NonCommSemiringM.getSemiring
modifySemiring := NonCommSemiringM.modifySemiring
def getSemiringEntry : NonCommSemiringM SemiringEntry := do
let s get'
let semiringId := ( read).semiringId
if h : semiringId < s.ncSemirings.size then
@@ -29,14 +45,10 @@ protected def NonCommSemiringM.getSemiring : NonCommSemiringM Semiring := do
else
throwError "`grind` internal error, invalid semiringId"
protected def NonCommSemiringM.modifySemiring (f : Semiring Semiring) : NonCommSemiringM Unit := do
def modifySemiringEntry (f : SemiringEntry SemiringEntry) : NonCommSemiringM Unit := do
let semiringId := ( read).semiringId
modify' fun s => { s with ncSemirings := s.ncSemirings.modify semiringId f }
instance : MonadSemiring NonCommSemiringM where
getSemiring := NonCommSemiringM.getSemiring
modifySemiring := NonCommSemiringM.modifySemiring
def getTermNonCommSemiringId? (e : Expr) : GoalM (Option Nat) := do
return ( get').exprToNCSemiringId.find? { expr := e }
@@ -48,7 +60,25 @@ def setTermNonCommSemiringId (e : Expr) : NonCommSemiringM Unit := do
return ()
modify' fun s => { s with exprToNCSemiringId := s.exprToNCSemiringId.insert { expr := e } semiringId }
instance : MonadSetTermId NonCommSemiringM where
setTermId e := setTermNonCommSemiringId e
def mkNonCommSVar (e : Expr) (gen : Nat) : NonCommSemiringM Var := do
unless ( alreadyInternalized e) do
internalize e gen
let s getSemiringEntry
if let some var := s.varMap.find? { expr := e } then
return var
let var : Var := s.vars.size
modifySemiringEntry fun s => { s with
vars := s.vars.push e
varMap := s.varMap.insert { expr := e } var
}
setTermNonCommSemiringId e
ringExt.markTerm e
return var
instance : MonadMkVar NonCommSemiringM where
mkVar := mkNonCommSVar
instance : MonadGetVar NonCommSemiringM where
getVar x := return ( getSemiringEntry).vars[x]!
end Lean.Meta.Grind.Arith.CommRing

View File

@@ -6,20 +6,26 @@ Authors: Leonardo de Moura
module
prelude
public import Lean.Meta.Tactic.Grind.Types
import Lean.Meta.Tactic.Grind.Arith.CommRing.DenoteExpr
public import Lean.Meta.Tactic.Grind.Arith.CommRing.Types
public import Lean.Meta.Tactic.Grind.Arith.CommRing.Denote
import Lean.Meta.Sym.Arith.DenoteExpr
import Init.Omega
public section
namespace Lean.Meta.Grind.Arith.CommRing
open Sym.Arith
private abbrev M := StateT CommRing MetaM
private abbrev M := StateT (CommRing × CommRingEntry) MetaM
private instance : MonadCanon M where
canonExpr e := return e
synthInstance? e := Meta.synthInstance? e none
private instance : MonadCommRing M where
getCommRing := get
modifyCommRing := modify
getCommRing := return ( get).1
modifyCommRing f := modify fun (r, e) => (f r, e)
private instance : MonadGetVar M where
getVar x := return ( get).2.vars[x]!
private def toOption (cls : Name) (header : Thunk MessageData) (msgs : Array MessageData) : Option MessageData :=
if msgs.isEmpty then
@@ -30,15 +36,18 @@ private def toOption (cls : Name) (header : Thunk MessageData) (msgs : Array Mes
private def push (msgs : Array MessageData) (msg? : Option MessageData) : Array MessageData :=
if let some msg := msg? then msgs.push msg else msgs
private def getCommRingEntry : M CommRingEntry :=
return ( get).2
private def ppBasis? : M (Option MessageData) := do
let mut basis := #[]
for c in ( getCommRing).basis do
for c in ( getCommRingEntry).basis do
basis := basis.push (toTraceElem ( c.denoteExpr))
return toOption `basis "Basis" basis
private def ppDiseqs? : M (Option MessageData) := do
let mut diseqs := #[]
for d in ( getCommRing).diseqs do
for d in ( getCommRingEntry).diseqs do
diseqs := diseqs.push (toTraceElem ( d.denoteExpr))
return toOption `diseqs "Disequalities" diseqs
@@ -50,9 +59,12 @@ private def ppRing? : M (Option MessageData) := do
def pp? (goal : Goal) : MetaM (Option MessageData) := do
let mut msgs := #[]
for ring in ( ringExt.getStateCore goal).rings do
let some msg ppRing? |>.run' ring | pure ()
msgs := msgs.push msg
for ringEntry in ( ringExt.getStateCore goal).rings do
-- let ring ← getCommRingOfId ringEntry.symId
-- let some msg ← ppRing? |>.run' (_, ringEntry) | pure ()
-- msgs := msgs.push msg
-- TODO: fix
pure ()
if msgs.isEmpty then
return none
else if h : msgs.size = 1 then

View File

@@ -8,16 +8,19 @@ prelude
public import Lean.Meta.Tactic.Grind.Arith.CommRing.RingId
public import Lean.Meta.Tactic.Grind.Arith.CommRing.NonCommRingM
public import Lean.Meta.Tactic.Grind.Arith.CommRing.NonCommSemiringM
public import Lean.Meta.Tactic.Grind.Arith.CommRing.Denote
import Lean.Data.RArray
import Lean.Meta.Tactic.Grind.Diseq
import Lean.Meta.Tactic.Grind.ProofUtil
import Lean.Meta.Tactic.Grind.Arith.CommRing.DenoteExpr
import Lean.Meta.Tactic.Grind.Arith.CommRing.ToExpr
import Lean.Meta.Tactic.Grind.Arith.CommRing.VarRename
import Lean.Meta.Sym.Arith.DenoteExpr
import Lean.Meta.Sym.Arith.ToExpr
import Lean.Meta.Sym.Arith.VarRename
import Init.Data.Nat.Order
import Init.Data.Order.Lemmas
public section
namespace Lean.Meta.Grind.Arith.CommRing
open Sym.Arith
/--
Returns a context of type `RArray α` containing the variables `vars` where
`α` is the type of the ring.
@@ -55,7 +58,7 @@ private def throwNoNatZeroDivisors : RingM α := do
private def getPolyConst (p : Poly) : RingM Int := do
let .num k := p
| throwError "`grind` internal error, constant polynomial expected {indentExpr (← p.denoteExpr)}"
| throwError "`grind` internal error, constant polynomial expected {indentExpr (← denotePoly p)}"
return k
structure ProofM.State where
@@ -133,6 +136,9 @@ private def getSemiringIdOf : RingM Nat := do
private def getSemiringOf : RingM CommSemiring := do
SemiringM.run ( getSemiringIdOf) do getCommSemiring
private def getSemiringEntryOf : RingM CommSemiringEntry := do
SemiringM.run ( getSemiringIdOf) do getCommSemiringEntry
private def mkSemiringPrefix (declName : Name) : ProofM Expr := do
let sctx getSContext
let semiring getSemiringOf
@@ -141,7 +147,7 @@ private def mkSemiringPrefix (declName : Name) : ProofM Expr := do
private def mkSemiringAddRightCancelPrefix (declName : Name) : ProofM Expr := do
let sctx getSContext
let semiring getSemiringOf
let some addRightCancelInst SemiringM.run ( getSemiringIdOf) do getAddRightCancelInst?
let some addRightCancelInst SemiringM.run ( getSemiringIdOf) do return ( getCommSemiring).addRightCancelInst?
| throwError "`grind` internal error, `AddRightCancel` instance is not available"
return mkApp4 (mkConst declName [semiring.u]) semiring.type semiring.semiringInst addRightCancelInst sctx
@@ -239,7 +245,7 @@ private def mkContext (h : Expr) : ProofM Expr := do
collectMapVars ( get).exprDecls (·.collectVars) <| {}
let vars' := usedVars.toArray
let varRename := mkVarRename vars'
let vars := ( getRing).vars
let vars := ( getCommRingEntry).vars
let vars := vars'.map fun x => vars[x]!
let h := mkLetOfMap ( get).polyDecls h `p (mkConst ``Grind.CommRing.Poly) fun p => toExpr <| p.renameVars varRename
let h := mkLetOfMap ( get).monDecls h `m (mkConst ``Grind.CommRing.Mon) fun m => toExpr <| m.renameVars varRename
@@ -256,11 +262,12 @@ private def mkContext (h : Expr) : ProofM Expr := do
private def mkSemiringContext (h : Expr) : ProofM Expr := do
let some sctx := ( read).sctx? | return h
let some semiringId := ( getCommRing).semiringId? | return h
let semiring getSemiringOf
let semiring getSemiringOf
let semiringEntry getSemiringEntryOf
let usedVars := collectMapVars ( get).sexprDecls (·.collectVars) {}
let vars' := usedVars.toArray
let varRename := mkVarRename vars'
let vars := vars'.map fun x => semiring.vars[x]!
let vars := vars'.map fun x => semiringEntry.vars[x]!
let h := mkLetOfMap ( get).sexprDecls h `s (mkConst ``Grind.CommRing.Expr) fun s => toExpr <| s.renameVars varRename
let h := h.abstract #[sctx]
if h.hasLooseBVars then
@@ -325,7 +332,7 @@ def setSemiringDiseqUnsat (a b : Expr) (sa sb : SemiringExpr) : SemiringM Unit :
let usedVars := sa.collectVars >> sb.collectVars <| {}
let vars' := usedVars.toArray
let varRename := mkVarRename vars'
let vars := ( getSemiring).vars
let vars := ( getCommSemiringEntry).vars
let vars := vars'.map fun x => vars[x]!
let sa := sa.renameVars varRename
let sb := sb.renameVars varRename
@@ -340,11 +347,12 @@ terms s.t. `ra.toPoly_nc == rb.toPoly_nc`, close the goal.
-/
def setNonCommRingDiseqUnsat (a b : Expr) (ra rb : RingExpr) : NonCommRingM Unit := do
let ring getRing
let ringEntry getRingEntry
let hne mkDiseqProof a b
let usedVars := ra.collectVars >> rb.collectVars <| {}
let vars' := usedVars.toArray
let varRename := mkVarRename vars'
let vars := ring.vars
let vars := ringEntry.vars
let vars := vars'.map fun x => vars[x]!
let ra := ra.renameVars varRename
let rb := rb.renameVars varRename
@@ -362,11 +370,12 @@ terms s.t. `sa.toPolyS_nc == sb.toPolyS_nc`, close the goal.
-/
def setNonCommSemiringDiseqUnsat (a b : Expr) (sa sb : SemiringExpr) : NonCommSemiringM Unit := do
let semiring getSemiring
let semiringEntry getSemiringEntry
let hne mkDiseqProof a b
let usedVars := sa.collectVars >> sb.collectVars <| {}
let vars' := usedVars.toArray
let varRename := mkVarRename vars'
let vars := semiring.vars
let vars := semiringEntry.vars
let vars := vars'.map fun x => vars[x]!
let sa := sa.renameVars varRename
let sb := sb.renameVars varRename
@@ -395,7 +404,8 @@ private def norm (vars : PArray Expr) (lhs rhs lhs' rhs' : RingExpr) : NormResul
def mkLeIffProof (leInst ltInst isPreorderInst orderedRingInst : Expr) (lhs rhs lhs' rhs' : RingExpr) : RingM Expr := do
let ring getCommRing
let { lhs, rhs, lhs', rhs', vars } := norm ring.vars lhs rhs lhs' rhs'
let ringEntry getCommRingEntry
let { lhs, rhs, lhs', rhs', vars } := norm ringEntry.vars lhs rhs lhs' rhs'
let ctx toContextExpr vars
let h := mkApp6 (mkConst ``Grind.CommRing.le_norm_expr [ring.u]) ring.type ring.commRingInst leInst ltInst isPreorderInst orderedRingInst
let h := mkApp6 h ctx (toExpr lhs) (toExpr rhs) (toExpr lhs') (toExpr rhs') eagerReflBoolTrue
@@ -407,7 +417,8 @@ def mkLeIffProof (leInst ltInst isPreorderInst orderedRingInst : Expr) (lhs rhs
def mkLtIffProof (leInst ltInst lawfulOrdLtInst isPreorderInst orderedRingInst : Expr) (lhs rhs lhs' rhs' : RingExpr) : RingM Expr := do
let ring getCommRing
let { lhs, rhs, lhs', rhs', vars } := norm ring.vars lhs rhs lhs' rhs'
let ringEntry getCommRingEntry
let { lhs, rhs, lhs', rhs', vars } := norm ringEntry.vars lhs rhs lhs' rhs'
let ctx toContextExpr vars
let h := mkApp7 (mkConst ``Grind.CommRing.lt_norm_expr [ring.u]) ring.type ring.commRingInst leInst ltInst lawfulOrdLtInst isPreorderInst orderedRingInst
let h := mkApp6 h ctx (toExpr lhs) (toExpr rhs) (toExpr lhs') (toExpr rhs') eagerReflBoolTrue
@@ -419,7 +430,8 @@ def mkLtIffProof (leInst ltInst lawfulOrdLtInst isPreorderInst orderedRingInst :
def mkEqIffProof (lhs rhs lhs' rhs' : RingExpr) : RingM Expr := do
let ring getCommRing
let { lhs, rhs, lhs', rhs', vars } := norm ring.vars lhs rhs lhs' rhs'
let ringEntry getCommRingEntry
let { lhs, rhs, lhs', rhs', vars } := norm ringEntry.vars lhs rhs lhs' rhs'
let ctx toContextExpr vars
let h := mkApp2 (mkConst ``Grind.CommRing.eq_norm_expr [ring.u]) ring.type ring.commRingInst
let h := mkApp6 h ctx (toExpr lhs) (toExpr rhs) (toExpr lhs') (toExpr rhs') eagerReflBoolTrue
@@ -434,7 +446,8 @@ Given `e` and `e'` s.t. `e.toPoly == e'.toPoly`, returns a proof that `e.denote
-/
def mkTermEqProof (e e' : RingExpr) : RingM Expr := do
let ring getCommRing
let { lhs, lhs', vars, .. } := norm ring.vars e (.num 0) e' (.num 0)
let ringEntry getCommRingEntry
let { lhs, lhs', vars, .. } := norm ringEntry.vars e (.num 0) e' (.num 0)
let ctx toContextExpr vars
let h := mkApp2 (mkConst ``Grind.CommRing.Expr.eq_of_toPoly_eq [ring.u]) ring.type ring.commRingInst
let h := mkApp4 h ctx (toExpr lhs) (toExpr lhs') eagerReflBoolTrue
@@ -444,7 +457,8 @@ def mkTermEqProof (e e' : RingExpr) : RingM Expr := do
def mkNonCommLeIffProof (leInst ltInst isPreorderInst orderedRingInst : Expr) (lhs rhs lhs' rhs' : RingExpr) : NonCommRingM Expr := do
let ring getRing
let { lhs, rhs, lhs', rhs', vars } := norm ring.vars lhs rhs lhs' rhs'
let ringEntry getRingEntry
let { lhs, rhs, lhs', rhs', vars } := norm ringEntry.vars lhs rhs lhs' rhs'
let ctx toContextExpr vars
let h := mkApp6 (mkConst ``Grind.CommRing.le_norm_expr_nc [ring.u]) ring.type ring.ringInst leInst ltInst isPreorderInst orderedRingInst
let h := mkApp6 h ctx (toExpr lhs) (toExpr rhs) (toExpr lhs') (toExpr rhs') eagerReflBoolTrue
@@ -456,7 +470,8 @@ def mkNonCommLeIffProof (leInst ltInst isPreorderInst orderedRingInst : Expr) (l
def mkNonCommLtIffProof (leInst ltInst lawfulOrdLtInst isPreorderInst orderedRingInst : Expr) (lhs rhs lhs' rhs' : RingExpr) : NonCommRingM Expr := do
let ring getRing
let { lhs, rhs, lhs', rhs', vars } := norm ring.vars lhs rhs lhs' rhs'
let ringEntry getRingEntry
let { lhs, rhs, lhs', rhs', vars } := norm ringEntry.vars lhs rhs lhs' rhs'
let ctx toContextExpr vars
let h := mkApp7 (mkConst ``Grind.CommRing.lt_norm_expr_nc [ring.u]) ring.type ring.ringInst leInst ltInst lawfulOrdLtInst isPreorderInst orderedRingInst
let h := mkApp6 h ctx (toExpr lhs) (toExpr rhs) (toExpr lhs') (toExpr rhs') eagerReflBoolTrue
@@ -468,7 +483,8 @@ def mkNonCommLtIffProof (leInst ltInst lawfulOrdLtInst isPreorderInst orderedRin
def mkNonCommEqIffProof (lhs rhs lhs' rhs' : RingExpr) : NonCommRingM Expr := do
let ring getRing
let { lhs, rhs, lhs', rhs', vars } := norm ring.vars lhs rhs lhs' rhs'
let ringEntry getRingEntry
let { lhs, rhs, lhs', rhs', vars } := norm ringEntry.vars lhs rhs lhs' rhs'
let ctx toContextExpr vars
let h := mkApp2 (mkConst ``Grind.CommRing.eq_norm_expr_nc [ring.u]) ring.type ring.ringInst
let h := mkApp6 h ctx (toExpr lhs) (toExpr rhs) (toExpr lhs') (toExpr rhs') eagerReflBoolTrue
@@ -483,7 +499,8 @@ Given `e` and `e'` s.t. `e.toPoly_nc == e'.toPoly_nc`, returns a proof that `e.d
-/
def mkNonCommTermEqProof (e e' : RingExpr) : NonCommRingM Expr := do
let ring getRing
let { lhs, lhs', vars, .. } := norm ring.vars e (.num 0) e' (.num 0)
let ringEntry getRingEntry
let { lhs, lhs', vars, .. } := norm ringEntry.vars e (.num 0) e' (.num 0)
let ctx toContextExpr vars
let h := mkApp2 (mkConst ``Grind.CommRing.Expr.eq_of_toPoly_nc_eq [ring.u]) ring.type ring.ringInst
let h := mkApp4 h ctx (toExpr lhs) (toExpr lhs') eagerReflBoolTrue

Some files were not shown because too many files have changed in this diff Show More