On: How I fixed a bug in Atom

[TazeTSchnitzel]

This could be solved by using regexes to tokenise instead, then detecting brackets. Tokenising is something regular expressions can actually do well. Directly parsing source code is not.