Given the following lines:

classes!("text-xl", "text-blue-500")
String::from("test")
classes!("text-sm")

I wish to extract text-xl, text-blue-500 and text-sm and potentially any other strings within classes!

What I have so far is:

• (?<=classes!).* - this gives me everything after classes! but not the exact values
• (?!^")".*?" - this gives me the values I want with but with the quotes
• I use the global and multi line flags

CodePudding user response：

If you are using Javascript, and the positive lookbehind is supported (you don't need the multiline flag, only the global flag)

(?<=classes!\((?:"[^"]*",\s*)*")[^"]*(?=")

• (?<= Positive lookbehind to assert to the left
  • classes!\( Match classes!(
  • (?:"[^"]*",\s*)* Match optional repetitions of "...",
  • " Match a double quote
• ) Close lookbehind
• [^"]* Match optional chars other than "
• (?=") Assert " to the right

Regex demo

const regex = /(?<=classes!\((?:"[^"]*",\s*)*")[^"]*(?=")/g;
const str = `classes!("text-xl", "text-blue-500")
String::from("test")
classes!("text-sm")
`;
console.log(str.match(regex))

CodePudding user response：

You could try a non fixed-width lookbehind like so:

(?<=^classes!\("(?:[^"]*"[^"()]*")*)[^"]*

See an online demo

• (?<= - Open a positive lookbehind;
  • ^classes!\(" - Start-line anchor and match literally 'classes!("';
  • (?:[^"]*"[^()"]*")* - A nested non-capture group to match (0 times) any char other than quotes (and paranthesis), a quote and that same sequence repeated.
  • ) - Close lookbehind;
• [^"]* - Match 0 (Greedy) characters other than quotes.