Documentation

Init.Data.Array.Attach

@[implemented_by _private.Init.Data.Array.Attach.0.Array.attachWithImpl]
def Array.attachWith {α : Type u_1} (xs : Array α) (P : αProp) (H : ∀ (x : α), x xsP x) :
Array { x : α // P x }

O(1). "Attach" a proof P x that holds for all the elements of xs to produce a new array with the same elements but in the type {x // P x}.

Equations
  • xs.attachWith P H = { toList := xs.toList.attachWith P }
Instances For
    @[inline]
    def Array.attach {α : Type u_1} (xs : Array α) :
    Array { x : α // x xs }

    O(1). "Attach" the proof that the elements of xs are in xs to produce a new array with the same elements but in the type {x // x ∈ xs}.

    Equations
    Instances For
      @[simp]
      theorem List.attachWith_toArray {α : Type u_1} {l : List α} {P : αProp} {H : ∀ (x : α), x l.toArrayP x} :
      l.toArray.attachWith P H = (l.attachWith P ).toArray
      @[simp]
      theorem List.attach_toArray {α : Type u_1} {l : List α} :
      l.toArray.attach = (l.attachWith (fun (x : α) => x l.toArray) ).toArray
      @[simp]
      theorem Array.toList_attachWith {α : Type u_1} {l : Array α} {P : αProp} {H : ∀ (x : α), x lP x} :
      (l.attachWith P H).toList = l.toList.attachWith P
      @[simp]
      theorem Array.toList_attach {α : Type u_1} {l : Array α} :
      l.attach.toList = l.toList.attachWith (fun (x : α) => x l)

      unattach #

      Array.unattach is the (one-sided) inverse of Array.attach. It is a synonym for Array.map Subtype.val.

      We use it by providing a simp lemma l.attach.unattach = l, and simp lemmas which recognize higher order functions applied to l : Array { x // p x } which only depend on the value, not the predicate, and rewrite these in terms of a simpler function applied to l.unattach.

      Further, we provide simp lemmas that push unattach inwards.

      def Array.unattach {α : Type u_1} {p : αProp} (l : Array { x : α // p x }) :

      A synonym for l.map (·.val). Mostly this should not be needed by users. It is introduced as in intermediate step by lemmas such as map_subtype, and is ideally subsequently simplified away by unattach_attach.

      If not, usually the right approach is simp [Array.unattach, -Array.map_subtype] to unfold.

      Equations
      • l.unattach = Array.map (fun (x : { x : α // p x }) => x.val) l
      Instances For
        @[simp]
        theorem Array.unattach_nil {α : Type u_1} {p : αProp} :
        #[].unattach = #[]
        @[simp]
        theorem Array.unattach_push {α : Type u_1} {p : αProp} {a : { x : α // p x }} {l : Array { x : α // p x }} :
        (l.push a).unattach = l.unattach.push a.val
        @[simp]
        theorem Array.size_unattach {α : Type u_1} {p : αProp} {l : Array { x : α // p x }} :
        l.unattach.size = l.size
        @[simp]
        theorem List.unattach_toArray {α : Type u_1} {p : αProp} {l : List { x : α // p x }} :
        l.toArray.unattach = l.unattach.toArray
        @[simp]
        theorem Array.toList_unattach {α : Type u_1} {p : αProp} {l : Array { x : α // p x }} :
        l.unattach.toList = l.toList.unattach
        @[simp]
        theorem Array.unattach_attach {α : Type u_1} {l : Array α} :
        l.attach.unattach = l
        @[simp]
        theorem Array.unattach_attachWith {α : Type u_1} {p : αProp} {l : Array α} {H : ∀ (a : α), a lp a} :
        (l.attachWith p H).unattach = l

        Recognizing higher order functions using a function that only depends on the value. #

        theorem Array.foldl_subtype {α : Type u_1} {β : Type u_2} {p : αProp} {l : Array { x : α // p x }} {f : β{ x : α // p x }β} {g : βαβ} {x : β} {hf : ∀ (b : β) (x : α) (h : p x), f b x, h = g b x} :
        Array.foldl f x l = Array.foldl g x l.unattach

        This lemma identifies folds over arrays of subtypes, where the function only depends on the value, not the proposition, and simplifies these to the function directly taking the value.

        @[simp]
        theorem Array.foldl_subtype' {α : Type u_1} {β : Type u_2} {stop : Nat} {p : αProp} {l : Array { x : α // p x }} {f : β{ x : α // p x }β} {g : βαβ} {x : β} {hf : ∀ (b : β) (x : α) (h : p x), f b x, h = g b x} (h : stop = l.size) :
        Array.foldl f x l 0 stop = Array.foldl g x l.unattach

        Variant of foldl_subtype with side condition to check stop = l.size.

        theorem Array.foldr_subtype {α : Type u_1} {β : Type u_2} {p : αProp} {l : Array { x : α // p x }} {f : { x : α // p x }ββ} {g : αββ} {x : β} {hf : ∀ (x : α) (h : p x) (b : β), f x, h b = g x b} :
        Array.foldr f x l = Array.foldr g x l.unattach

        This lemma identifies folds over arrays of subtypes, where the function only depends on the value, not the proposition, and simplifies these to the function directly taking the value.

        @[simp]
        theorem Array.foldr_subtype' {α : Type u_1} {β : Type u_2} {start : Nat} {p : αProp} {l : Array { x : α // p x }} {f : { x : α // p x }ββ} {g : αββ} {x : β} {hf : ∀ (x : α) (h : p x) (b : β), f x, h b = g x b} (h : start = l.size) :
        Array.foldr f x l start = Array.foldr g x l.unattach

        Variant of foldr_subtype with side condition to check stop = l.size.

        @[simp]
        theorem Array.map_subtype {α : Type u_1} {β : Type u_2} {p : αProp} {l : Array { x : α // p x }} {f : { x : α // p x }β} {g : αβ} {hf : ∀ (x : α) (h : p x), f x, h = g x} :
        Array.map f l = Array.map g l.unattach

        This lemma identifies maps over arrays of subtypes, where the function only depends on the value, not the proposition, and simplifies these to the function directly taking the value.

        @[simp]
        theorem Array.filterMap_subtype {α : Type u_1} {β : Type u_2} {p : αProp} {l : Array { x : α // p x }} {f : { x : α // p x }Option β} {g : αOption β} {hf : ∀ (x : α) (h : p x), f x, h = g x} :
        @[simp]
        theorem Array.unattach_filter {α : Type u_1} {p : αProp} {l : Array { x : α // p x }} {f : { x : α // p x }Bool} {g : αBool} {hf : ∀ (x : α) (h : p x), f x, h = g x} :
        (Array.filter f l).unattach = Array.filter g l.unattach

        Simp lemmas pushing unattach inwards. #

        @[simp]
        theorem Array.unattach_reverse {α : Type u_1} {p : αProp} {l : Array { x : α // p x }} :
        l.reverse.unattach = l.unattach.reverse
        @[simp]
        theorem Array.unattach_append {α : Type u_1} {p : αProp} {l₁ : Array { x : α // p x }} {l₂ : Array { x : α // p x }} :
        (l₁ ++ l₂).unattach = l₁.unattach ++ l₂.unattach