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
```
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>
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>
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.
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>
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>
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>
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>
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>
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>
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>
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.
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>
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>
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>
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>
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>
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>
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>
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>
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>
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>
This PR verifies the `String.dropWhile` and `String.takeWhile`
functions.
It also includes a refactor of the `PatternModel` class so that the
`not_matches_empty` condition is moved into a separate typeclass
`StrictPatternModel`. This allows string patterns to implement
`LawfulForwardPatternModel` unconditionally, which means that more of
the general theory about patterns directly applies to string patterns
without having to do a case distinction for empty strings.
This PR also includes a study of the `PatternModel` machinery given to
slices `s` and `t` such that `s.copy = t.copy`. From these results, we
deduce statements like `s.copy.startsWith pat = s.startsWith pat` (which
is far from obvious!).
This PR changes the counter-example accumulator in the match compiler
from
a `List` (built with cons, producing reverse order) to an `Array` (built
with push, preserving declaration order). Missing cases are now reported
in
the order constructors appear in the inductive type definition.
For example, given `inductive Enum | a | b | c | d`, missing cases `c`
and
`d` were previously shown as `d, c` and are now shown as `c, d`.
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
privateConstantMap capacity was inflated by IR extraConstNames that are
only inserted into const2ModIdx. const2ModIdx capacity included
numPublicConsts even though public constants are never inserted into it.
This PR adds a `match.maxCounterExamples` option (default 5) to bound
the
number of "missing cases" counter-examples the match compiler generates.
When the match compiler runs out of alternatives for a variable, it
case-splits to explore missing cases. Previously, this would recursively
split all inductive-typed fields of each constructor, leading to
O(K^fields)
counter-examples for types with K constructors per field. For nested
incomplete matches on types like `Op w` (20 constructors with `Operand
w`
fields having 8 constructors each), this produced thousands of
counter-examples and took several seconds.
The fix checks the counter-example count in `isConstructorTransition`:
once
the limit is reached and there are no remaining alternatives, the match
compiler stops exploring further case splits. The error message notes
when
output has been truncated and names the option. The existing protection
against infinite recursion on recursive types is preserved.
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
This PR adds a `Core.checkInterrupted` call in `checkInferTypeCache` on
cache miss, allowing cancellation to be detected during large type
inference traversals. Previously, `inferTypeImp` could run for >100ms
without any interruption check when processing large expressions (e.g.
BVDecide proof terms), making IDE cancellation unresponsive.
The check is only on cache miss (actual computation), so cache hits have
zero overhead. `checkInterrupted` is used instead of the heavier
`checkSystem` to minimize performance impact — local benchmarks show no
measurable regression on `tests/elab/5664.lean`.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
This PR changes elaboration of structure instance notation when used in
patterns (e.g. `s matches { x := 1, y := [] }`) so that the structure's
default values are not used to elaborate the pattern. The motivation is
that default values frequently lead to surprisingly over-specific
patterns. It will now report "field missing" errors. The error can be
suppressed using `{ x := 1, .. }` ellipsis notation, which has the same
behavior as before. The pretty printer is also modified to stay in sync
with this feature. **Breaking change:** patterns using structure
instance notation may need missing fields or a `..` added, as
appropriate.
The rationale for the previous behavior is that with `..` you could
opt-in to not using default values, and now with this PR's behavior you
cannot opt-in. However, default values in structure instance patterns
are very likely to silently cause bugs. There are a couple examples in
this PR of unintentional default values in patterns in core Lean
(luckily these were not triggering bugs). With the new behavior, you can
now tell for sure whether every explicit field in a structure is being
matched explicitly or not, by the absence or presence of `..`.
Closes#10753
This PR adds support for marking options as deprecated. When a
deprecated option is used via `set_option`, a warning is emitted
(controlled by `linter.deprecated.options`).
An `OptionDeprecation` structure with a required `since` field and an
optional `text?` field is added to `OptionDecl`. Each `set_option`
elaborator (command, term, tactic, grind) calls `checkDeprecatedOption`
to emit warnings. The C++ `register_option` is updated to account for
the new field.
As a first use, `backward.eqns.nonrecursive` and
`backward.eqns.deepRecursiveSplit` are marked deprecated. Continues
earlier work done in #11096.
This PR adds test infrastructure and tests for tactic completion in
empty `by` blocks.
**Test runner improvements (`src/Lean/Server/Test/Runner.lean`):**
- Add `--⬑` marker variant that targets the column of `--` itself,
enabling column 0 tests (which `--^` cannot reach since `^` is always at
column 2+).
**New test file (`tests/server_interactive/completionEmptyBy.lean`):**
- Tests tactic completion in empty `by` blocks for both top-level `by`
and nested `by` (inside `id <| have := by`).
- Tests at various column positions on the line below `by`: indented
past `by`, at column 2, and at column 0.
- Tests on the `by` token itself (no completions expected).
- All positions below `by` currently offer tactic completions.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
---------
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
This PR adds support for let configuration options (`(eq := h)`,
`+nondep`, `+usedOnly`, `+zeta`) in `do` block `let` and `have`
declarations, matching the behavior available in term-level
`let`/`have`. Configuration options are rejected with `let mut` since
they are incompatible with mutable bindings. `+postponeValue` and
`+generalize` are also rejected in `do` blocks.
Follow-up to #13250 which added the parser support. Now that stage0 is
updated, this PR replaces the backward-compat index helpers with proper
quotation patterns and implements the actual `letConfig` elaboration.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
This PR extends the `doLet`, `doLetElse`, `doLetArrow`, and `doHave`
parsers to accept `letConfig` (e.g. `(eq := h)`, `+nondep`, `+usedOnly`,
`+zeta`), matching the syntax of term-level `let`/`have`. The
elaborators are adjusted to handle the shifted syntax indices but do not
yet process the configuration; that will be done in a follow-up PR after
stage0 is updated, allowing the use of proper quotation patterns.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
---------
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
This PR removes a FIXME in Sym-based mvcgen' concerning a hardcoded step
limit for grind simplification. I tested that this is no longer
necessary even at highest setting for the GetThrowSetGrind benchmark.
This PR extends Lean syntax for dotted function notation (`.f`) to add
support for explicit mode (`@.f`), explicit universes (`.f.{u,v}`), and
both simultaneously (`@.f.{u,v}`). This also includes a fix for a bug
involving overloaded functions, where it used to give erroneous
deprecation warnings about declarations that the function did not
elaborate to.
Closes#10984
This PR improves pretty printing of level metavariables: they now print
with a per-definition index rather than their per-module internal
identifiers. Furthermore, `+` is printed uniformly in level expressions
with surrounding spaces. **Breaking metaprogramming change:** level
pretty printing should use `delabLevel` or `MessageData.ofLevel`;
functions such as `format` or `toString` do not have access to the
indices, since they are stored in the current metacontext. Absent index
information, metavariables print with the raw internal identifier as
`?_mvar.nnn`. **Note:** The heartbeat counter also increases quicker due
to counting allocations that record level metavariable indices. In some
tests we needed to increase `maxHeartbeats` by 20–50% to compensate,
without a corresponding slowdown.
