Typeclass for a type F
with an injective map to A → B
#
This typeclass is primarily for use by homomorphisms like MonoidHom
and LinearMap
.
There is the "D"ependent version DFunLike
and the non-dependent version FunLike
.
Basic usage of DFunLike
and FunLike
#
A typical type of morphisms should be declared as:
structure MyHom (A B : Type*) [MyClass A] [MyClass B] where
(toFun : A → B)
(map_op' : ∀ (x y : A), toFun (MyClass.op x y) = MyClass.op (toFun x) (toFun y))
namespace MyHom
variable (A B : Type*) [MyClass A] [MyClass B]
instance : FunLike (MyHom A B) A B where
coe := MyHom.toFun
coe_injective' := fun f g h => by cases f; cases g; congr
@[ext] theorem ext {f g : MyHom A B} (h : ∀ x, f x = g x) : f = g := DFunLike.ext f g h
/-- Copy of a `MyHom` with a new `toFun` equal to the old one. Useful to fix definitional
equalities. -/
protected def copy (f : MyHom A B) (f' : A → B) (h : f' = ⇑f) : MyHom A B where
toFun := f'
map_op' := h.symm ▸ f.map_op'
end MyHom
This file will then provide a CoeFun
instance and various
extensionality and simp lemmas.
Morphism classes extending DFunLike
and FunLike
#
The FunLike
design provides further benefits if you put in a bit more work.
The first step is to extend FunLike
to create a class of those types satisfying
the axioms of your new type of morphisms.
Continuing the example above:
/-- `MyHomClass F A B` states that `F` is a type of `MyClass.op`-preserving morphisms.
You should extend this class when you extend `MyHom`. -/
class MyHomClass (F : Type*) (A B : outParam Type*) [MyClass A] [MyClass B]
[FunLike F A B] : Prop :=
(map_op : ∀ (f : F) (x y : A), f (MyClass.op x y) = MyClass.op (f x) (f y))
@[simp]
lemma map_op {F A B : Type*} [MyClass A] [MyClass B] [FunLike F A B] [MyHomClass F A B]
(f : F) (x y : A) :
f (MyClass.op x y) = MyClass.op (f x) (f y) :=
MyHomClass.map_op _ _ _
-- You can add the below instance next to `MyHomClass.instFunLike`:
instance : MyHomClass (MyHom A B) A B where
map_op := MyHom.map_op'
-- [Insert `ext` and `copy` here]
Note that A B
are marked as outParam
even though they are not purely required to be so
due to the FunLike
parameter already filling them in. This is required to see through
type synonyms, which is important in the category theory library. Also, it appears having them as
outParam
is slightly faster.
The second step is to add instances of your new MyHomClass
for all types extending MyHom
.
Typically, you can just declare a new class analogous to MyHomClass
:
structure CoolerHom (A B : Type*) [CoolClass A] [CoolClass B] extends MyHom A B where
(map_cool' : toFun CoolClass.cool = CoolClass.cool)
class CoolerHomClass (F : Type*) (A B : outParam Type*) [CoolClass A] [CoolClass B]
[FunLike F A B] extends MyHomClass F A B :=
(map_cool : ∀ (f : F), f CoolClass.cool = CoolClass.cool)
@[simp] lemma map_cool {F A B : Type*} [CoolClass A] [CoolClass B] [FunLike F A B]
[CoolerHomClass F A B] (f : F) : f CoolClass.cool = CoolClass.cool :=
CoolerHomClass.map_cool _
variable {A B : Type*} [CoolClass A] [CoolClass B]
instance : FunLike (CoolerHom A B) A B where
coe f := f.toFun
coe_injective' := fun f g h ↦ by cases f; cases g; congr; apply DFunLike.coe_injective; congr
instance : CoolerHomClass (CoolerHom A B) A B where
map_op f := f.map_op'
map_cool f := f.map_cool'
-- [Insert `ext` and `copy` here]
Then any declaration taking a specific type of morphisms as parameter can instead take the class you just defined:
-- Compare with: lemma do_something (f : MyHom A B) : sorry := sorry
lemma do_something {F : Type*} [FunLike F A B] [MyHomClass F A B] (f : F) : sorry :=
sorry
This means anything set up for MyHom
s will automatically work for CoolerHomClass
es,
and defining CoolerHomClass
only takes a constant amount of effort,
instead of linearly increasing the work per MyHom
-related declaration.
Design rationale #
The current form of FunLike was set up in pull request https://github.com/leanprover-community/mathlib4/pull/8386:
https://github.com/leanprover-community/mathlib4/pull/8386
We made FunLike
unbundled: child classes don't extend FunLike
, they take a [FunLike F A B]
parameter instead. This suits the instance synthesis algorithm better: it's easy to verify a type
does not have a FunLike
instance by checking the discrimination tree once instead of searching
the entire extends
hierarchy.
The class DFunLike F α β
expresses that terms of type F
have an
injective coercion to (dependent) functions from α
to β
.
For non-dependent functions you can also use the abbreviation FunLike
.
This typeclass is used in the definition of the homomorphism typeclasses,
such as ZeroHomClass
, MulHomClass
, MonoidHomClass
, ....
- coe : F → (a : α) → β a
The coercion from
F
to a function. - coe_injective' : Function.Injective DFunLike.coe
The coercion to functions must be injective.
Instances
The class FunLike F α β
(Fun
ction-Like
) expresses that terms of type F
have an injective coercion to functions from α
to β
.
FunLike
is the non-dependent version of DFunLike
.
This typeclass is used in the definition of the homomorphism typeclasses,
such as ZeroHomClass
, MulHomClass
, MonoidHomClass
, ....
Instances For
This is not an instance to avoid slowing down every single Subsingleton
typeclass search.