#33437 [Opn->Bgs]: comments throwing preg_replace script

ID: 33437 Updated by: tony2001@php.net Reported By: php at karsites dot net -Status: Open +Status: Bogus Bug Type: PCRE related Operating System: SuSE Linux 9.2 pro PHP Version: 4.3.10 New Comment: So, no bug here -> bogus. Previous Comments: ------------------------------------------------------------------------ [2005-06-22 18:01:12] php at karsites dot net Woops! I've just noticed the ?> in the regular expression. Obviously, this has been recognised as the 'back to html mode' php tag! Keith ------------------------------------------------------------------------ [2005-06-22 17:52:29] tony2001@php.net Please try using this CVS snapshot: http://snaps.php.net/php4-STABLE-latest.tar.gz For Windows: http://snaps.php.net/win32/php4-win32-STABLE-latest.zip ------------------------------------------------------------------------ [2005-06-22 17:47:26] php at karsites dot net Description: ------------ I'm actually on php4-4.3.8-8.7 - stuck with that at the moment,as that ships with my Linux distro - SuSE 9.2 pro I did search the database for 'coments' but found nothing there regarding this possible bug. The code below trashes the script with the commented line in the script // original regex $regex = "!(</?)(\w+)([^>]*?>)!e"; Surely, everything to the right of // should be ignored as a comment? With the comment in place, the script produces the following trashed output: )!e"; $regex = "!(]*?>)!e"; $replacement = "'\\1'.strtolower('\\2').'\\3'"; $subject = " A rose"; echo "Before replacement \$subject is: $subject "; $new_sub = preg_replace($regex, $replacement, $subject); echo "After replacement \$subject is: $new_sub "; ?> Is this a bug, because it is not expected behaviour. If I remove the offending comment, the script produces the correct output. Regards - Keith Roberts Reproduce code: --------------- <?php // original regex $regex = "!(</?)(\w+)([^>]*?>)!e"; $regex = "!(</?)(\w+)([^>]*?>)!e"; $replacement = "'\\1'.strtolower('\\2').'\\3'"; $subject = "<BIGTEXT> A <BIGGEST>rose</BIGGEST> <PARA></BIGTEXT>"; echo "Before replacement \$subject is: $subject <br />"; $new_sub = preg_replace($regex, $replacement, $subject); echo "After replacement \$subject is: $new_sub <br />"; ?> Expected result: ---------------- Before replacement $subject is: A rose After replacement $subject is: A rose (the browser ignores the dummy tags, but the correct result can be seen with 'view source code' - ie lowercased tags) Actual result: -------------- )!e"; $regex = "!(]*?>)!e"; $replacement = "'\\1'.strtolower('\\2').'\\3'"; $subject = " A rose"; echo "Before replacement \$subject is: $subject "; $new_sub = preg_replace($regex, $replacement, $subject); echo "After replacement \$subject is: $new_sub "; ?> ------------------------------------------------------------------------ -- Edit this bug report at http://bugs.php.net/?id=33437&edit=1