grammar dreme::datum::Grammar;

token TOP {
    <datum>
    [ $ || <.panic: "Syntax error"> ]
}

## Lexer items

# This <ws> rule treats # as "comment to eol".
token ws {
    <!ww>
    <comment>*
}

token comment {
    | <line_comment>
    | <nested_comment>
}

token line_comment {
    ';' \N* \n?
}

token nested_comment {
    '#|' ~ '|#' .
}

## Data

rule datum {
    | <simple_datum>
    | <compound_datum>
}

rule simple_datum {
    | <boolean>
    | <character>
    | <number>
    | <string>
    | <symbol>
    | <bytevector>
}

rule compound_datum {
    | <list>
    | <vector>
}

rule symbol { <identifier> }

rule list {
    | '(' ~ ')' <list_data>
    | '[' ~ ']' <list_data> # R6RS
}

rule list_data {
    | <datum>*
    | <datum>+ '.' <datum>
    | <datum> '.' <datum> '.' <datum> # Racket
}

rule abbreviation {
    | "'" <datum>
    | '`' <datum>
    | ',' <datum>
    | ',@' <datum>
}

rule vector {
    '#' '(' ~ ')' <datum>*
}

token bytevector_hash {
    | '#vu8' # R6RS
    | '#u8'  # R7RS
}

rule bytevector {
    <bytevector_hash> '(' ~ ')' <u8>
}

token u8 {
    \d**1..3 <?{$/ < 256}>
}