14877 – Regexp replace halts on unicode char

RESOLVED DUPLICATE of bug 10370 14877

Regexp replace halts on unicode char

https://bugs.webkit.org/show_bug.cgi?id=14877

Summary Regexp replace halts on unicode char

Derk-Jan Hartman

Reported 2007-08-03 14:21:54 PDT

2 testcases attached

Attachments
testcase by bdash (146 bytes, text/html) 2007-08-03 14:23 PDT, Derk-Jan Hartman	no flags	Details
Original case (1.17 KB, text/html) 2007-08-03 14:26 PDT, Derk-Jan Hartman	no flags	Details
Improved test case (284 bytes, text/html) 2007-08-03 14:54 PDT, Mark Rowe (bdash)	no flags	Details
Show Obsolete (1) View All Add attachment proposed patch, testcase, etc.

Derk-Jan Hartman

Comment 1 2007-08-03 14:23:58 PDT

Created attachment 15828 [details] testcase by bdash

Derk-Jan Hartman

Comment 2 2007-08-03 14:26:16 PDT

Created attachment 15830 [details] Original case

Derk-Jan Hartman

Comment 3 2007-08-03 14:37:32 PDT

Issue is also present in Safari 2.0.4 (419.3)

Mark Rowe (bdash)

Comment 4 2007-08-03 14:53:01 PDT

The issue is probably narrower in scope than the title suggests. In this specific example the character class [\s\S] should match any single character (\s is the set of all whitespace characters, \S is the set of non-whitespace characters). For some reason this is not correctly matching LEFTWARDS ARROW. It would be interesting to know whether other character classes exhibit this issue, and whether other Unicode characters do too.

Mark Rowe (bdash)

Comment 5 2007-08-03 14:54:00 PDT

Created attachment 15833 [details] Improved test case

Mark Rowe (bdash)

Comment 6 2007-08-03 14:55:10 PDT

<rdar://problem/5385211>

Alexey Proskuryakov

Comment 7 2007-08-03 23:37:42 PDT