Feature request: Improve implementation of the format string parser #155

Describe the problem

The format string parser that extracts expressions (i.e {{ expression... }} blocks) from the string for processing is a pretty naive character-by-character inspection of the input string. It works but it's not very performant. A better implementation is desirable.

Proposed Solution

A solution that leverages the performant parts of the Python runtime would be better. For example, something that efficiently leverages the regular expression parser.

The implementation of the pygments lexer (BSD 2-clause license) could be a solid source of inspiration for this.

Example Use Cases

N/A