I\'ve problem with my regex code, the problem is: I want to replace all \"foo\" values but only inside \"()\". Example:
try foo foo-foo (try foo inside , foo
This finds all occurences of foo inside () and replace them with str_replace. From my experience I can say, that this is faster than preg_replace.
while (preg_match_all('/^(.*)(\([^\)]*foo[^\)]*\))(.*)$/', $string, $pat)) {
$string = $pat[1] . str_replace('foo', 'boo', $pat[2]) . $pat[3];
}
Note the three groups: The string before the ( ($pat[1]), the string in the () which contains foo ($pat[2]) and the string behind the ) ($pat[3]).
If you can live with using a loop and replacing multiple times, this solution will work:
do {
$string = preg_replace('/(\([^)]*)foo([^)]*\))/U', '\1boo\2', $string, -1, $count);
}while($count != 0);
echo $string;
Assuming parentheses are "valid", you can replace all foos that are followed by ), without a ( in between:
$str = preg_replace('/foo(?=[^()]*\))/', 'bar', $str);
Working example: http://ideone.com/xezU7
Another option, also assuming no nested pairs, is to match "(...)" tokens and use a callback, but it may be use of excessive force:
Callback method:
function replace_foo($matches) {
return str_replace('foo', 'bar', $matches[0]);
}
use:
$str = preg_replace_callback("/\([^()]*\)/","replace_foo", $str);
Working example: http://ideone.com/ZzBem