0.1.0: ready for integration

First, a shout out for @aMOPel efforts to write queries to bring this
grammar into https://github.com/nvim-treesitter/nvim-treesitter.

Their work on this allowed the grammar to mature greatly with regards to
grammar node design.

Per tradition, here are the parse statistics against the compiler for
this version:

Total parses: 3987;
successful parses: 3911;
failed parses: 76;
success percentage: 98.09%

This is done from a full clone of the compiler source including extra
code from nimble and atlas.

The parser size grew a little to 47MiB in this release.

Below are notable changes of this cycle:

Breaking changes

• type_expression is no longer an anonymized node by @aMOPel (#29).
  This allows for an easier time matching against all the nitty gritty
  expressions that can happen within a type context.

Parsing improvements

• using blocks are now supported by @aMOPel (#31).
• Long string opening are now matchable tokens by @alaviss (#34).
• The parsing behavior for *: now matches the compiler by @alaviss (#36).

Regressions

• Unicode operators remains unsupported as adding them still inflates
  the parser size considerably.