Unicode (utf-8)

[booniepepper]

• Choose a utf8 String implementation. (Promising: zigstr and utf8proc)
• Implement it as the "String" type in dt.
  • Note: the dictionary type will have to move away from StringHashMap(Command) which not the worst thing ever, but also not trivial.
• Expand tokenization to consider more kinds of whitespace (e.g. CJK full-width space)
• Add runes and remove chars
• Add more tests on string-sensitive commands like upcase, downcase, split, join, len, or eq
• Add tests that include utf-8 source code (Like demos/世界を挨拶.dt)