regex explained in english

Question

I have looked here and from what I understand the following regex simply means "any unicode character sequence". Can someone confirm this please?

Current Regex: /^(?>\P{M}\p{M}*)+$/u

Also if I read the manual it says

a) \P{M} = \PM

b) (?>\PM\pM*) = \X

So with these two things in hand, can I not simplify the regex to?:

Proposed Regex: /^\X+$/u

Which I still don't actually understand...

Ok, according to the anual you are right. But, what is the question? — Alberto De Caro, Jun 20 '12 at 14:14
[Me reading regex documentation](http://thejoysofcode.tumblr.com/post/35331265331/when-i-am-forced-to-read-regex-documentation) — Paul de Lange, Nov 17 '12 at 10:56

score 2 · Answer 1 · answered Jun 20 '12 at 14:17

2

^            # start of string followed by 
(?>          # an independent (non-backtracking) capturing group containing 
    \P{M}    # a single unicode character which is not in the `Mark` category
    \p{M}*   # 0 or more characters in the `Mark` category
)+           # with this capturing group repeated 1 or more times
$            # the end-of-line

Whereas ^\X+$ contains no capturing group; the \P{M}\p{M}* is otherwise equivalent.

answered Jun 20 '12 at 14:17

beerbajay

19,652
6
58
75

This was also a very helpful answer, but I can only accept one and the answer by Bart is more what I was looking for. – Paul de Lange Jun 20 '12 at 14:22

score 2 · Accepted Answer · answered Jun 20 '12 at 14:18

Yes, \P{M}\p{M}* could be simplified to \X, but not all languages support \X while (in my experience) \P{M} and \p{M} are supported more frequently.

For example, Java's and .NET's regex engines do not support \X (Perl does, of course...).

More info, see: http://www.regular-expressions.info/unicode.html

regex explained in english

2 Answers2