Report: IgnoringInactive: a bug?

Question

Bug introduced in 10.0.0 and fixed in 10.0.1

---

Just to report.

IgnoringInactiveshould work with Cases. In agreement with the things written at the Documentation Center this:

 Cases[{Cos[a], Inactive[Cos][b]}, IgnoringInactive[Cos[x_]]]

works correctly yielding

(*   {Cos[a], Inactive[Cos][b]}   *)

exactly as expected.

However, this

Cases[{Cos[a], Inactive[Cos][b], Tan[c]}, IgnoringInactive[Cos[x_]]]

works (as I think) incorrectly returning

(*  {Cos[a], Inactive[Cos][b], Tan[c]}  *)

while I would expect again {Cos[a], Inactive[Cos][b]}.

Answer 1

Since IgnoringInactive is presently bugged we may wish to have a workaround for this operation.
Here is my first attempt at implementing this. Please test it and give me feedback.

patternHead = 
  Pattern | Blank | Repeated | HoldPattern | Verbatim | Inactive | _Alternatives;

ignInac[pat_] := 
 Replace[pat, (h : Except[patternHead])[x___] :> (h | Inactive[h])[x], {0, -1}]

Test for an example from the documentation:

diff[ignInac[int[f_, x_]], x_] := f

diff[int[Sin[x], x], x]
diff[Inactive[int][Sin[x], x], x]

Sin[x]

Sin[x]

And demonstrating correct function with the examples from the question and comments:

Cases[{Cos[a], Inactive[Cos][b], Tan[c]}, ignInac @ Cos[x_]]

{Cos[a], Inactive[Cos][b]}

MatchQ[1, ignInac[0]]
MatchQ["a", ignInac["b"]]

False
False

Notes:

• The function ignInac directly transforms patterns so definitions do not look as clean as with the built-in, however for the cases I've tried it appears equivalent.

• I did not attempt to handle the case of HoldPattern outside of IgnoringInactive (or ignInac) but I do not see an example of that in the documentation.