Hi, these are just some spelling corrections.
There is one I wasn't completely sure about in
src/Init/Data/List/Lemmas.lean:
> See also
> ...
> Also
> \* \`Init.Data.List.Monadic\` for **addiation** _(additional?)_ lemmas
about \`List.mapM\` and \`List.forM\`
This PR continues the homogenization between matchers and splitters,
following up on #11256. In particular it removes the ambiguity whether
`numParams` includes the `discrEqns` or not.
This PR adds a `Unit` assumption to alternatives of the splitter that
would otherwise not have arguments. This fixes#11211.
In practice these argument-less alternatives did not cause wrong
behavior, as the motive when used with `split` is always a function
type. But it is better to be safe here (maybe someone uses splitters in
other ways), it may increase the effectiveness of #10184 and simplifies
#11220.
The perf impact is insignificant in the grand scheme of things on
stdlib, but the change is effective:
```
~/lean4 $ build/release/stage1/bin/lean tests/lean/run/matchSplitStats.lean
969 splitters found
455 splitters are const defs
~/lean4 $ build/release/stage2/bin/lean tests/lean/run/matchSplitStats.lean
969 splitters found
829 splitters are const defs
```
This PR lets the `split` tactic generalize discriminants that are not
free variables and proofs using `generalize`. If the only
non-fvar-discriminants are proofs, then this avoids the more elaborate
generalization strategy of `split`, which can fail with dependent
motives, thus mitigating issue #10424.
This PR adjusts the import graph, primarily of `Lean`, such that the
worst case rebuild time of core (`lean` only) is below 3 minutes on the
speedcenter machine (not captured by benchmark yet).
This PR fixes a performance issue that occurs when generating equation
lemmas for functions that use match-expressions containing several
literals. This issue was exposed by #9322 and arises from a combination
of factors:
1. Literal values are compiled into a chain of dependent if-then-else
expressions.
2. Dependent if-then-else expressions are significantly more expensive
to simplify than regular ones.
3. The `split` tactic selects a target, splits it, and then invokes
`simp` on the resulting subgoals. Moreover, `simp` traverses the entire
goal bottom-up and does not stop after reaching the target.
This PR addresses the issue by introducing a custom simproc that avoids
recursively simplifying nested if-then-else expressions. It does **not**
alter the user-facing behavior of the `split` tactic because such a
change would be highly disruptive. Instead, the PR adds a new flag,
`backward.split` to control the behavior of the user-facing `split`
tactic. It is currently set to `true`, i.e., the old behavior is still
the default one. In a future PR, we should set this flag to `false` by
default and begin repairing all affected proofs.
closes#9322
This PR migrates usages of `Std.Range` to the new polymorphic ranges.
This PR unfortunately increases the transitive imports for
frequently-used parts of `Init` because the ranges now rely on iterators
in order to provide their functionality for types other than `Nat`.
However, iteration over ranges in compiled code is as efficient as
before in the examples I checked. This is because of a special
`IteratorLoop` implementation provided in the PR for this purpose.
There were two issues that were uncovered during migration:
* In `IndPredBelow.lean`, migrating the last remaining range causes
`compilerTest1.lean` to break. I have minimized the issue and came to
the conclusion it's a compiler bug. Therefore, I have not replaced said
old range usage yet (see #9186).
* In `BRecOn.lean`, we are publicly importing the ranges. Making this
import private should theoretically work, but there seems to be a
problem with the module system, causing the build to panic later in
`Init.Data.Grind.Poly` (see #9185).
* In `FuzzyMatching.lean`, inlining fails with the new ranges, which
would have led to significant slowdown. Therefore, I have not migrated
this file either.
This PR replaces all usages of `[:]` slice notation in `src` with the
new `[...]` notation in production code, tests and comments. The
underlying implementation of the `Subarray` functions stays the same.
Notation cheat sheet:
* `*...*` is the doubly-unbounded range.
* `*...a` or `*...<a` contains all elements that are less than `a`.
* `*...=a` contains all elements that are less than or equal to `a`.
* `a...*` contains all elements that are greater than or equal to `a`.
* `a...b` or `a...<b` contains all elements that are greater than or
equal to `a` and less than `b`.
* `a...=b` contains all elements that are greater than or equal to `a`
and less than or equal to `b`.
* `a<...*` contains all elements that are greater than `a`.
* `a<...b` or `a<...<b` contains all elements that are greater than `a`
and less than `b`.
* `a<...=b` contains all elements that are greater than `a` and less
than or equal to `b`.
Benchmarks have shown that importing the iterator-backed parts of the
polymorphic slice library in `Init` impacts build performance. This PR
avoids this problem by separating those parts of the library that do not
rely on iterators from those those that do. Whereever the new slice
notation is used, only the iterator-independent files are imported.
This PR adds the following features to `simp`:
- A routine for simplifying `have` telescopes in a way that avoids
quadratic complexity arising from locally nameless expression
representations, like what #6220 did for `letFun` telescopes.
Furthermore, simp converts `letFun`s into `have`s (nondependent lets),
and we remove the #6220 routine since we are moving away from `letFun`
encodings of nondependent lets.
- A `+letToHave` configuration option (enabled by default) that converts
lets into haves when possible, when `-zeta` is set. Previously Lean
would need to do a full typecheck of the bodies of `let`s, but the
`letToHave` procedure can skip checking some subexpressions, and it
modifies the `let`s in an entire expression at once rather than one at a
time.
- A `+zetaHave` configuration option, to turn off zeta reduction of
`have`s specifically. The motivation is that dependent `let`s can only
be dsimped by let, so zeta reducing just the dependent lets is a
reasonable way to make progress. The `+zetaHave` option is also added to
the meta configuration.
- When `simp` is zeta reducing, it now uses an algorithm that avoids
complexity quadratic in the depth of the let telescope.
- Additionally, the zeta reduction routines in `simp`, `whnf`, and
`isDefEq` now all are consistent with how they apply the `zeta`,
`zetaHave`, and `zetaUnused` configurations.
The `letToFun` option is addressing a TODO in `getSimpLetCase` ("handle
a block of nested let decls in a single pass if this becomes a
performance problem").
Performance should be compared to before #8804, which temporarily
disabled the #6220 optimizations for `letFun` telescopes.
Good kernel performance depends on carefully handling the `have`
encoding. Due to the way the kernel instantiates bvars (it does *not*
beta reduce when instantiating), we cannot use congruence theorems of
the form `(have x := v; f x) = (have x ;= v'; f' x)`, since the bodies
of the `have`s will not be syntactically equal, which triggers zeta
reduction in the kernel in `is_def_eq`. Instead, we work with `f v = f'
v'`, where `f` and `f'` are lambda expressions. There is still zeta
reduction, but only when converting between these two forms at the
outset of the generated proof.
This PR fixes `split` in the presence of metavariables in the target.
The fix consists of replacing an internal use of `apply` for
instantiating match splitters by a new, simpler variant `applyN`. This
new `applyN` is not prone to #8436, which is the ultimate cause for
`split` failing on targets containing metavariables.
---------
Co-authored-by: Sebastian Graf <sg@lean-fro.org>
Co-authored-by: Joachim Breitner <mail@joachim-breitner.de>
This PR changes the `simpMatch` function, used inside the equation
generator for WF-rec functions, to not do eta-expansion.
This makes the process a bit more robust and disciplined, and avoids
removing match-statements (and introduce projections and dependencies)
that we'd rather split instead.
Also adds more tracing to the equational theorem generator.
Extracted from #6898.
This PR implements `Simp.Config.implicitDefEqsProofs`. When `true`
(default: `true`), `simp` will **not** create a proof term for a
rewriting rule associated with an `rfl`-theorem. Rewriting rules are
provided by users by annotating theorems with the attribute `@[simp]`.
If the proof of the theorem is just `rfl` (reflexivity), and
`implicitDefEqProofs := true`, `simp` will **not** create a proof term
which is an application of the annotated theorem.
The default setting does change the existing behavior. Users can use
`simp -implicitDefEqProofs` to force `simp` to create a proof term for
`rfl`-theorems. This can positively impact proof checking time in the
kernel.
This PR also fixes an issue in the `split` tactic that has been exposed
by this feature. It was looking for `split` candidates in proofs and
implicit arguments. See new test for issue exposed by the previous
feature.
---------
Co-authored-by: Kim Morrison <kim@tqft.net>
This PR avoids runtime array bounds checks in places where it can
trivially be done at compile time.
None of these changes are of particular consequence: I mostly wanted to
learn how much we do this, and what the obstacles are to doing it less.
PR #3432 will introduce more operations on `MatcherApp`, including somet
that have more dependencies.
This change prepares by introducing `Lean.Meta.Match.MatcherApp.Basic`
for the basic definition, and `Lean.Meta.MatcherApp.Transform` for the
transformations, currently `addArg` and `refineThrough`, but more to
come.
Before this commit, `Simproc`s were defined as `Expr -> SimpM (Option Step)`, where `Step` is inductively defined as follows:
```
inductive Step where
| visit : Result → Step
| done : Result → Step
```
Here, `Result` is a structure containing the resulting expression and a proof demonstrating its equality to the input. Notably, the proof is optional; in its absence, `simp` assumes reflexivity.
A simproc can:
- Fail by returning `none`, indicating its inapplicability. In this case, the next suitable simproc is attempted, along with other simp extensions.
- Succeed and invoke further simplifications using the `.visit`
constructor. This action returns control to the beginning of the
simplification loop.
- Succeed and indicate that the result should not undergo further
simplifications. However, I find the current approach unsatisfactory, as it does not align with the methodology employed in `Transform.lean`, where we have the type:
```
inductive TransformStep where
/-- Return expression without visiting any subexpressions. -/
| done (e : Expr)
/--
Visit expression (which should be different from current expression) instead.
The new expression `e` is passed to `pre` again.
-/
| visit (e : Expr)
/--
Continue transformation with the given expression (defaults to current expression).
For `pre`, this means visiting the children of the expression.
For `post`, this is equivalent to returning `done`. -/
| continue (e? : Option Expr := none)
```
This type makes it clearer what is going on. The new `Simp.Step` type is similar but use `Result` instead of `Expr` because we need a proof.
