Help with regex: breaking strings down to 'words' and 'phrases'

Hi All, I'd very much appreciate some help building a regular expression for preg_match_all that can differentiate between 'words' and 'phrases'. For example, say I have a string that contains: 'this is an "example of a phrase"' I'd like to be able to break that down to: this is an example of a phrase My current preg_match_all regex: preg_match_all('([\w\-]+|[\(]|[\)])',"this is an \"example of a phrase\"',$arr); returns the following: Array ( [0] => Array ( [0] => this [1] => is [2] => an [3] => example [4] => of [5] => a [6] => phrase ) ) Note: I'm using this to break elements of a string down to build an sql string, which is why I'm looking for "(" and ")" characters (ie the "[\(]|[\)]" part of the regex) and maintaining them in the array. A real-world example of the the value being supplied to the regex might be "completed and "January 2005" and not (store or online)" etc. I already have the logic to handle "and", "or", "not" and "()" but haven't been able to figure out how to maintain substrings in quotes as a single value in the array. Any help appreciated! Much warmth, Murray