ansaurus

Question

Answer 1

+1 A:

([a-zA-Z]){20,}

Michał Niklas 2009-05-06 10:26:41

Be careful: this won't work with extended charsets...

Stefan Gehrig 2009-05-06 10:30:12

Of course it depends of author intentions. Sometimes \w is better, but remember that \w includes digits.

Michał Niklas 2009-05-06 10:59:43

That's right - it really depends on the author's needs.

Stefan Gehrig 2009-05-06 15:18:38

Answer 2

+3 A:

Try this regular expression:

\b\w{20,}

Or if the implementation you are using supports Unicode character properties:

\b\p{L}{20,}

Gumbo 2009-05-06 10:31:05

Shouldn't it be \w (lower-case W) to match any "word" character? \W will match any "non-word" character.

Stefan Gehrig 2009-05-06 10:33:35

Of course, sgehrig, you’re right. Thanks!

Gumbo 2009-05-06 10:38:19

Regex to parse long words