I've separated the library into CleanRegex and SafeRegex. The SafeRegex is an exact copy of all preg_*() methods (like preg_match(), preg_replace()). They're used like this:

• preg_match() -> preg::match()
• preg_replace() -> preg::replace()

I didn't want to use namespaces (like \SafeRegex\preg_match) because that would mean the programmers reading the code would have to check the imported namespaces every time they look at the code. With double colon :: it's obvious. Read on, for more details.

preg::*() methods analyse warnings, return types of methods and result of preg_last_error(), and if they find that the pattern or subject failed in some way, they throw an exception, instead of a warning or returning null. I want developers to stop worrying about what might cause a bug, when they use regular expressions, and just focus on the creative part. I want them to feel: "This code has to work, it would throw an exception otherwise".

And the CleanRegex is what it was - the API part, that's using SafeRegex never to care about warnings or returning null/false/[]/''.

That's it for today! :)