This PR adds the following tactics to the `grind` interactive mode:
- `focus <grind_tac_seq>`
- `next => <grind_tac_seq>`
- `any_goals <grind_tac_seq>`
- `all_goals <grind_tac_seq>`
- `grind_tac <;> grind_tac`
- `cases <anchor>`
- `tactic => <tac_seq>`
Example:
```lean
def g (as : List Nat) :=
match as with
| [] => 1
| [_] => 2
| _::_::_ => 3
example : g bs = 1 → g as ≠ 0 := by
grind [g.eq_def] =>
instantiate
cases #ec88
next => instantiate
next => finish
tactic =>
rw [h_2] at h_1
simp [g] at h_1
```
This PR implements *anchors* (also known as stable hash codes) for
referencing terms occurring in a `grind` goal. It also introduces the
commands `show_splits` and `show_state`. The former displays the anchors
for candidate case splits in the current `grind` goal.
This PR implements the basic tactics for the new `grind` interactive
mode. While many additional `grind` tactics will be added later, the
foundational framework is already operational. The following `grind`
tactics are currently implemented: `skip`, `done`, `finish`, `lia`, and
`ring`.
This PR also removes the notion of `grind` fallback procedure since it
is subsumed by the new framework. Examples:
```lean
example (x y : Nat) : x ≥ y + 1 → x > 0 := by
grind => skip; lia; done
open Lean Grind
example [CommRing α] (a b c : α)
: a + b + c = 3 →
a^2 + b^2 + c^2 = 5 →
a^3 + b^3 + c^3 = 7 →
a^4 + b^4 + c^4 = 9 := by
grind => ring
```
This PR adds infrastructure for the upcoming `grind` tactic mode, which
will be similar to the `conv` mode. The goal is to extend `grind` from a
terminal tactic into an interactive mode: `grind => …`.
It will serve as the foundation for `ungrind`, the process of converting
an expensive (and potentially fragile) `grind` proof into a robust
script. This mode will include tactics for expensive reasoning steps
such as cutsat model-based search, Gröbner basis computation,
E-matching, case splits, and more.
It will also provide robust, succinct references to facts and terms:
labels, structural matches, and anchors (e.g., `#abcd`).
This PR implements support for negative constraints in `grind order`.
Examples:
```lean
open Lean Grind
example [LE α] [LT α] [Std.LawfulOrderLT α] [Std.IsLinearPreorder α]
(a b c d : α) : a ≤ b → ¬ (c ≤ b) → ¬ (d ≤ c) → d < a → False := by
grind -linarith (splits := 0)
example [LE α] [Std.IsLinearPreorder α]
(a b c d : α) : a ≤ b → ¬ (c ≤ b) → ¬ (d ≤ c) → ¬ (a ≤ d) → False := by
grind -linarith (splits := 0)
example [LE α] [LT α] [Std.LawfulOrderLT α] [Std.IsLinearPreorder α] [CommRing α] [OrderedRing α]
(a b c d : α) : a - b ≤ 5 → ¬ (c ≤ b) → ¬ (d ≤ c + 2) → d ≤ a - 8 → False := by
grind -linarith (splits := 0)
```
This PR implements support for positive constraints in `grind order`.
The new module can already solve problems such as:
```lean
example [LE α] [LT α] [Std.LawfulOrderLT α] [Std.IsPreorder α]
(a b c : α) : a ≤ b → b ≤ c → c < a → False := by
grind
example [LE α] [LT α] [Std.LawfulOrderLT α] [Std.IsPreorder α]
(a b c d : α) : a ≤ b → b ≤ c → c < d → d ≤ a → False := by
grind
example [LE α] [Std.IsPreorder α]
(a b c : α) : a ≤ b → b ≤ c → a ≤ c := by
grind
example [LE α] [Std.IsPreorder α]
(a b c d : α) : a ≤ b → b ≤ c → c ≤ d → a ≤ d := by
grind
```
It also generalizes support for offset constraints in `grind` to rings.
The new module implements theory propagation and reduces the number of
case splits required to solve problems:
```lean
example [LE α] [LT α] [Std.LawfulOrderLT α] [Std.IsPreorder α] [Ring α] [OrderedRing α]
(a b : α) : a ≤ 5 → b ≤ 8 → a > 6 ∨ b > 10 → False := by
grind -linarith (splits := 0)
example [LE α] [LT α] [Std.LawfulOrderLT α] [Std.IsPreorder α] [CommRing α] [OrderedRing α]
(a b c : α) : a + b*c + 2*c ≤ 5 → a + c > 5 - c - c*b → False := by
grind -linarith (splits := 0)
example (a b : Int) (h : a + b > 5) : (if a + b ≤ 0 then b else a) = a := by
grind -linarith -cutsat (splits := 0)
```
We still need to implement support for negated constraints.
This PR implements the function for adding new edges to the graph used
by `grind order`. The graph maintains the transitive closure of all
asserted constraints.
This PR simplifies the `grind order` module, and internalizes the order
constraints. It removes the `Offset` type class because it introduced
too much complexity. We now cover the same use cases with a simpler
approach:
- Any type that implements at least `Std.IsPreorder`
- Arbitrary ordered rings.
- `Nat` by the `Nat.ToInt` adapter.
This PR adds a code action for `grind` parameters. We need to use
`set_option grind.param.codeAction true` to enable the option. The PR
also adds a modifier to instruct `grind` to use the "default" pattern
inference strategy.
This PR refines and clarifies the `meta` phase distinction in the module
system.
* `meta import A` without `public` now has the clarified meaning of
"enable compile-time evaluation of declarations in or above `A` in the
current module, but not downstream". This is now checked statically by
enforcing that public meta defs, which therefore may be referenced from
outside, can only use public meta imports, and that global evaluating
attributes such as `@[term_parser]` can only be applied to public meta
defs.
* `meta def`s may no longer reference non-meta defs even when in the
same module. This clarifies the meta distinction as well as improves
locality of (new) error messages.
* parser references in `syntax` are now also properly tracked as meta
references.
* A `meta import` of an `import` now properly loads only the `.ir` of
the nested module for the purposes of execution instead of also making
its declarations available for general elaboration.
* `initialize` is now no longer being run on import under the module
system, which is now covered by `meta initialize`.
This PR ensures users can select the "minimal indexable subexpression"
condition in `grind` parameters. Example, they can now write `grind [!
-> thmName]`. `grind?` will include the `!` modifier whenever users had
used `@[grind!]`. This PR also fixes a missing case in the new pattern
inference procedure.
It also adjusts some `grind` annotations and tests in preparation for
setting the new pattern inference heuristic as the new default.
This PR implements the new E-matching pattern inference heuristic for
`grind`. It is not enabled yet. You can activate the new behavior using
`set_option backward.grind.inferPattern false`. Here is a summary of the
new behavior.
* `[grind =]`, `[grind =_]`, `[grind _=_]`, `[grind <-=]`: no changes;
we keep the current behavior.
* `[grind ->]`, `[grind <-]`, `[grind =>]`, `[grind <=]`: we stop using
the *minimal indexable subexpression* and instead use the first
indexable one.
* `[grind! <mod>]`: behaves like `[grind <mod>]` but uses the minimal
indexable subexpression restriction. We generate an error if the user
writes `[grind! =]`, `[grind! =_]`, `[grind! _=_]`, or `[grind! <-=]`,
since there is no pattern search in these cases.
* `[grind]`: it tries `=`, `=_`, `<-`, `->`, `<=`, `=>` with and without
the minimal indexable subexpression restriction. For the ones that work,
we generate a code action to encourage users to select the one they
prefer.
* `[grind!]`: it tries `<-`, `->`, `<=`, `=>` using the minimal
indexable subexpression restriction. For the ones that work, we generate
a code action to encourage users to select the one they prefer.
* `[grind? <mod>]`: where `<mod>` is one of the modifiers above, it
behaves like `[grind <mod>]` but also displays the pattern.
Example:
```lean
/--
info: Try these:
• [grind =] for pattern: [f (g #0)]
• [grind =_] for pattern: [r #0#0]
• [grind! ←] for pattern: [g #0]
-/
#guard_msgs in
@[grind] axiom fg₇ : f (g x) = r x x
```
This PR adds the helper theorem `eq_normS_nc` for normalizing
non-commutative semirings. We will use this theorem to justify
normalization steps in the `grind ring` module.
This PR adds support for non-commutative ring normalization in `grind`.
The new normalizer also accounts for the `IsCharP` type class. Examples:
```lean
open Lean Grind
variable (R : Type u) [Ring R]
example (a b : R) : (a + 2 * b)^2 = a^2 + 2 * a * b + 2 * b * a + 4 * b^2 := by grind
example (a b : R) : (a + 2 * b)^2 = a^2 + 2 * a * b + -b * (-4) * a - 2*b*a + 4 * b^2 := by grind
variable [IsCharP R 4]
example (a b : R) : (a - b)^2 = a^2 - a * b - b * 5 * a + b^2 := by grind
example (a b : R) : (a - b)^2 = 13*a^2 - a * b - b * 5 * a + b*3*b*3 := by grind
```
This PR introduces limited functionality frontends `cutsat` and
`grobner` for `grind`. We disable theorem instantiation (and case
splitting for `grobner`), and turn off all other solvers. Both still
allow `grind` configuration options, so for example one can use `cutsat
+ring` (or `grobner +cutsat`) to solve problems that require both.
For `cutsat`, it is helpful to instantiate a limited set of theorems
(e.g. `Nat.max_def`). Currently this isn't supported, but we intend to
add this later.
This PR adds missing `grind` normalization rules for `natCast` and
`intCast` Examples:
```
open Lean.Grind
variable (R : Type) (a b : R)
section CommSemiring
variable [CommSemiring R]
example (m n : Nat) : (m + n) • a = m • a + n • a := by grind
example (m n : Nat) : (m * n) • a = m • (n • a) := by grind
end CommSemiring
section CommRing
variable [CommRing R]
example (m n : Nat) : (m + n) • a = m • a + n • a := by grind
example (m n : Nat) : (m * n) • a = m • (n • a) := by grind
example (m n : Int) : (m * n) • (a * b) = (m • a) * (n • b) := by grind
end CommRing
```
This PR adds the auxiliary theorem `Lean.Grind.Linarith.eq_normN` for
normalizing `NatModule` equations when the instance `AddRightCancel` is
not available.
This PR implements the infrastructure for supporting `NatModule` in
`grind linarith` and uses it to handle disequalities. Another PR will
add support for equalities and inequalities. Example:
```lean
open Lean Grind
variable (M : Type) [NatModule M] [AddRightCancel M]
example (x y : M) : 2 • x + 3 • y + x = 3 • (x + y) := by
grind
```
This PR implements equality propagation from the new AC module into the
`grind` core. Examples:
```lean
example {α β : Sort u} (f : α → β) (op : α → α → α) [Std.Associative op] [Std.Commutative op]
(a b c d : α) : op a (op b b) = op d c → f (op (op b a) (op b c)) = f (op c (op d c)) := by
grind only
example (a b c : Nat) : min a (max b (max c 0)) = min (max c b) a := by
grind -cutsat only
example {α β : Sort u} (bar : α → β) (op : α → α → α) [Std.Associative op] [Std.IdempotentOp op]
(a b c d e f x y w : α) :
op d (op x c) = op a b →
op e (op f (op y w)) = op (op d a) (op b c) →
bar (op d (op x c)) = bar (op e (op f (op y w))) := by
grind only
```
This PR adds the extra critical pairs to ensure the `grind ac` procedure
is complete when the operator is associative and idempotent, but not
commutative. Example:
```lean
example {α : Sort u} (op : α → α → α) [Std.Associative op] [Std.IdempotentOp op] (a b c d e f x y w : α)
: op d (op x c) = op a b →
op e (op f (op y w)) = op a (op b c) →
op d (op x c) = op e (op f (op y w)) := by
grind only
example {α : Sort u} (op : α → α → α) [Std.Associative op] [Std.IdempotentOp op] (a b c d e f x y w : α)
: op a (op d x) = op b c →
op e (op f (op y w)) = op a (op b c) →
op a (op d x) = op e (op f (op y w)) := by
grind only
```
This PR adds the extra critical pairs to ensure the `grind ac` procedure
is complete when the operator is AC and idempotent. Example:
```lean
example {α : Sort u} (op : α → α → α) [Std.Associative op] [Std.Commutative op] [Std.IdempotentOp op]
(a b c d : α) : op a (op b b) = op d c → op (op b a) (op b c) = op c (op d c) := by
grind only
```
This PR adds the inverse of a dyadic rational, at a given precision, and
characterising lemmas. Also cleans up various parts of the `Int.DivMod`
and `Rat` APIs, and proves some characterising lemmas about
`Rat.toDyadic`.
---------
Co-authored-by: Rob23oba <152706811+Rob23oba@users.noreply.github.com>
This PR adds superposition for associative (but non-commutative)
operators in `grind ac`. Examples:
```lean
example {α} (op : α → α → α) [Std.Associative op] (a b c d : α)
: op a b = c →
op b a = d →
op (op c a) (op b c) = op (op a d) (op d b) := by
grind
example {α} (a b c d : List α)
: a ++ b = c →
b ++ a = d →
c ++ a ++ b ++ c = a ++ d ++ d ++ b := by
grind only
```
This PR implements the proof terms for the new `grind ac` module.
Examples:
```lean
example {α : Sort u} (op : α → α → α) [Std.Associative op] (a b c d : α)
: op a (op b b) = op c d → op c (op d c) = op (op a b) (op b c) := by
grind only
example {α : Sort u} (op : α → α → α) [Std.Associative op] [Std.Commutative op] (a b c d : α)
: op a (op b b) = op d c → op (op b a) (op b c) = op c (op d c) := by
grind only
example {α : Sort u} (op : α → α → α) [Std.Associative op] [Std.Commutative op]
(one : α) [Std.LawfulIdentity op one] (a b c d : α)
: op a (op (op b one) b) = op d c → op (op b a) (op (op b one) c) = op (op c one) (op d c) := by
grind only
```
The `grind ac` module is not complete yet, we still need to implement
critical pair computation and fix the support for idempotent operators.
