MPE Home Metamath Proof Explorer < Previous   Next >
Nearby theorems
Mirrors  >  Home  >  MPE Home  >  Th. List  >  df-s1 Structured version   Visualization version   GIF version

Definition df-s1 14744
Description: Define the canonical injection from symbols to words. Although not required, 𝐴 should usually be a set. Otherwise, the singleton word ⟨“𝐴”⟩ would be the singleton word consisting of the empty set, see s1prc 14752, and not, as maybe expected, the empty word, see also s1nz 14755. (Contributed by Stefan O'Rear, 15-Aug-2015.) (Revised by Mario Carneiro, 26-Feb-2016.)
Assertion
Ref Expression
df-s1 ⟨“𝐴”⟩ = {⟨0, ( I ‘𝐴)⟩}

Detailed syntax breakdown of Definition df-s1
StepHypRef Expression
1 cA . . 3 class 𝐴
21cs1 14743 . 2 class ⟨“𝐴”⟩
3 cc0 11201 . . . 4 class 0
4 cid 5545 . . . . 5 class I
51, 4cfv 6539 . . . 4 class ( I ‘𝐴)
63, 5cop 4590 . . 3 class ⟨0, ( I ‘𝐴)⟩
76csn 4584 . 2 class {⟨0, ( I ‘𝐴)⟩}
82, 7wceq 1570 1 wff ⟨“𝐴”⟩ = {⟨0, ( I ‘𝐴)⟩}
Colors of variables:    wff setvar class
This definition is used by:  ids1  14745  s1val  14746  s1eq  14748  s1len  14754  s1nz  14755  funcnvs1  15064
  Copyright terms: Public domain W3C validator