Curling quotes, spelling, grammar, and architecture

[DarkTyger]

Consider using an event-based system for emitting tokens when parsing the Markdown document. There are a number of related features that could execute in parallel to speed up document processing. Consider the following Markdown document:

# Chaptar 1

"Hallo, world!" she didn't say ... he did

When building the AST, having a quote curler, spell checker, and grammar engine receive emitted tokens would greatly improve overall performance because they could do so in their own processing threads. The given sample document could emit, for example, the following tokens:

Token<Hashtag>, Token<Space>, Token<Word>, Token<Space>, Token<Number>,
Token<EOL>, Token<EOL>, Token<StraightDoubleQuote>, Token<Word>, Token<Comma>,
Token<Space>, Token<Word>, Token<Exclamation>, Token<StraightDoubleQuote>, Token<Space>,
Token<Word>, Token<Space>, Token<Word>, Token<StraightSingleQuote>, Token<Word>,
Token<Space>, Token<Word>, Token<Space>, Token<Ellipses>, Token<Space>, Token<Word>,
Token<Space>, Token<Word>, Token<EOF>

This is enough information to build the Markdown AST, curl quotation marks, check spelling, and verify grammar.