Re^2: "Inverse" Pattern Matching

[hdb]

The length of the pattern can be hundreds. The main source of inefficiency when looping is IMHO that every 'x' constitutes a branch in the pattern. If there is only one 'x' it could stand for one character and one has to pick n-1 other characters to create a pattern of the same length as my short string. However, if 'x' stands for two characters one has to pick only n-2 characters. And if every 'x' stands for two choices, the one has 2**k variants if there are k 'x's.