How can I use a regex to replace a character that is inside two other specific characters?

Question

I'm working on implementing a way to allow HTML in a field that I have with PHP. The characters are saved to the database as they are inputted, so when I print them to the screen later on, I want the HTML to be respected (security is not an issue here, so please don't make it one). What I need right now is a way to change any " instances to the proper " character, but only when it falls inside an HTML declaration (i.e., < and > characters). When it's not inside these characters, it shouldn't be changed.

What is the proper way to do something like this with regular expression?

Answer 1

Something like this to find tags? X is the character, but may be replaced by a string.

<[^>]*X[^>]*>

A practical example, which finds links that start with <a href:

<a href="([^"]*)"[^>]*>.

Answer 2

Try this:

$str = preg_replace('/&quot;(?=[^<>]*+>)/', '"', $str);

From your description I gather you mean "inside an HTML tag" (declaration doesn't really mean anything in this context). I'm also assuming there can't be any '<' or '>' characters inside the tag (because they are legal in quoted attribute values).

The lookahead (?=[^<>]*+>) asserts that there is a right angle bracket up ahead, and no other angle brackets between here and there. Assuming every angle bracket is part of a balanced pair, that should mean the current position is inside an HTML tag.