Regular expression - PCRE (PHP) - word boundary (\b) and accent characters

Question

Why does the letter é count as a word boundary matching \b in the following example?

Pattern: /\b(cum)\b/i

Text: écumé

Matches 'cum' which is not desired.

Is it possible to overcome this?

stema · Accepted Answer · 2014-02-27T12:57:12.137

10

It will work, when you add the u modifier to your regex

/\b(cum)\b/iu

edited Feb 27 '14 at 12:57

answered Feb 27 '14 at 12:39

stema

Thanks! Silly me, should have investigated modifiers a bit more. – marekful Feb 27 '14 at 12:41
Adding the Unicode modifier makes the regex treat input string as Unicode. PS: underrated question & solution. – JayKandari Dec 16 '20 at 08:21

score 0 · Answer 2 · answered Feb 27 '14 at 12:40

0

To deal with unicode, replace \b with

/(?<=^|\PL)(cum)(?=\PL|$)/i

answered Feb 27 '14 at 12:40

Toto

1

Thanks, but it seems too complex. stema's answer addresses the issue more straightforward, as the problem is that the text has unicode characters but the pattern is not made aware of this. – marekful Feb 27 '14 at 12:44

2 Answers2