TODO

2.0

Ensure there is no way to say a specific character in regex which does not fall in UTF-16 range (0 - 10FFFF).
Optimize PureInterpretor::CharMap, only 0 - 127 is in an array, others are in a range map.
Put Regex and RegexLexer to different cpp files.
A RegexReplacer, which could access anonymius or named captures:
- Use them to build a nee string
- Join list with delimiter
- Take anonymous and named captures as source and pairwide them to be a new list
Enhance RegexTokenizer. It could handle C++ multiline string literal by:
- Define the prefix as a token
  - It will allow capturing (but not others) and remove capturing to make it DFA-eligible
  - When a prefix is found, run the regex with capturing.
  - The above step will be skipped if no capturing definition is found in the regex.
- RegexReplacer gives a postfix
  - The postfix will be scanned to ensure all capturings are defined and no capturing is unused.
  - If there is no capturing in prefix, it will be a plain text, but escaping still in use.
- Search all the way to find the first postfix
- Make it a whole token, using the ID of the prefix token
  - Since the postfix is not used in tokenizing it will not occupy an ID.