chore: use double-backtick

Mathport/Binary/Number.lean

@@ -7,15 +7,16 @@ These functions mimic the ones in lean3/src/library/num.cpp
 -/
 import Lean
 import Mathport.Binary.Basic
+import Mathlib.Init.ZeroOne
 
 namespace Mathport.Binary
 
 open Lean
 
 partial def isConcreteNat? (e : Expr) : Option Nat := do
-  if e.isConstOf `Nat.zero then
+  if e.isConstOf ``Nat.zero then
     some 0
-  else if e.isAppOfArity `Nat.succ 1 then
+  else if e.isAppOfArity ``Nat.succ 1 then
     let n ← isConcreteNat? e.appArg!
     some (n+1)
   else
@@ -31,13 +32,13 @@ structure NumInfo where
   deriving Inhabited
 
 partial def isNumber? (e : Expr) : Option NumInfo := do
-  if e.isAppOfArity `Zero.zero 2 then pure {
+  if e.isAppOfArity ``Zero.zero 2 then pure {
     number   := 0,
     level    := e.getAppFn.constLevels!.head!,
     type     := e.getArg! 0
     hasZero? := e.getArg! 1
   }
-  else if e.isAppOfArity `One.one 2 then pure {
+  else if e.isAppOfArity ``One.one 2 then pure {
     number  := 1,
     level   := e.getAppFn.constLevels!.head!,
     type    := e.getArg! 0,
@@ -67,7 +68,7 @@ def translateNumbers (e : Expr) : MetaM Expr := Meta.transform e (pre := core) w
       | some { number := n, level, type, .. } =>
         -- TODO: we will want to avoid wrapping "normal" Nat numbers
         -- (current workaround is for the OfNat instances to `no_index` the numbers)
-        let inst := mkAppN (mkConst `OfNat.mk [level]) #[type, mkRawNatLit n, e]
-        TransformStep.done $ mkAppN (mkConst `OfNat.ofNat [level]) #[type, mkRawNatLit n, inst]
+        let inst := mkAppN (mkConst ``OfNat.mk [level]) #[type, mkRawNatLit n, e]
+        TransformStep.done $ mkAppN (mkConst ``OfNat.ofNat [level]) #[type, mkRawNatLit n, inst]
 
 end Mathport.Binary