This module contains: * `Token` - This record type is parameterized by some type `k`. It holds a value of type `k` representing the kind of token that was recognised, along with the text that was recognised for this particular `Token`. * `TokenKind` - Implementing this interface for some type `k` allows the user to specify how the recognised text from a token should convert into a value. There are two functions in the interface. * `tokType` - Determines the type that a recognised string should convert into for this kind of token. * `tokValue` - Given a kind and a string, convert the string into a value of the correct type. Using `Token k` as the token input to your grammar (and output of your lexer) allows for a couple advantages. 1. When a new kind of token is added, the compiler no longer accepts the program until it is accounted for in the `TokenKind` implementation. 2. The information about the recotnised text is kept for as long as possible. Since it's stored on the `Token` record, you don't lose information about the recognised text until a parse rule succeeds.