css3.pm

use v6;
grammar CSS {
    # http://www.w3.org/TR/css3-syntax/#tokenization
    rule ident      { '-'? <nmstart> <nmchar>* }
    rule name       { <nmchar>+ }
    rule nmstart    { <[a..zA..Z_]> | <nonascii> | <escape> }
    rule nonascii   { <[\x80..\xd7ff\xe000..\xfffd\x10000..\x10ffff]> }
    rule unicode    { '\\' <xdigit>**1..6 <wc>? }
    rule escape     { '\\' <[\x20..\x7e]> | <unicode> | <nonascii> }
    rule nmchar     { <nmstart> | <digit> | '-' }
    rule num        { <digit>* '.' <digit>+ | <digit>+ }
    rule string     { '"' ~ '"' [ <stringchar> | "'" ]*
                    | "'" ~ "'" [ <stringchar> | '"' ]* }
    rule stringchar { <urlchar> | ' ' | '\\' <nl> }
    rule urlchar    { <nmstart> | <[\x9\x21\x23..\x26\x27..\x7e]> }
    rule nl         { "\n" | "\r\n" | "\r" | "\xc" }
    rule w          { <wc>* }
    rule wc         { "\t" | "\n" | "\xc" | "\xd" | ' ' }

    rule IDENT          { <ident> }
    rule ATKEYWORD      { '@' <ident> }
    rule STRING         { <string> }
    rule HASH           { '#' <name> }
    rule NUMBER         { <num> }
    rule PERCENTAGE     { <num> '%' }
    rule DIMENSION      { <num> <ident> }
    rule URI            { 'url(' <w> [ <string> | <urlchar> ]* <w> ')' }
    rule UNICODE_RANGE  { 'U+' <[0..9A..F]>**1..6 [ '-' <[0..9A..F]>**1..6 ]? }
    rule CDO            { '<!--' }
    rule CDC            { '-->' }
    rule S              { <wc>+ }
    rule COMMENT        { '/*' <[^*]>* '*'+ [ <[^/]> <[^*]>* '*'+ ]* '/' }
    rule FUNCTION       { <ident> '(' }
    rule INCLUDES       { '~=' }
    rule DASHMATCH      { '|=' }
    rule PREFIXMATCH    { '^=' }
    rule SUFFIXMATCH    { '$=' }
    rule SUBSTRINGMATCH { '*=' }
    rule CHAR           { <[^'"]> }

    # http://www.w3.org/TR/css3-syntax/#style
    token TOP           { [ <CDO> | <CDC> | <S> | <statement> ]* }
    token statement     { <ruleset> | <at_rule> }
    token at_rule       { <ATKEYWORD> <S>* <css_any>* [ <block> | ';' <S>* ] }
    token block         { '{' ~ [ '}' <S>* ] <S>* [ <css_any> | <block> | <ATKEYWORD> <S>* | ';' <S>* ]* }
    token ruleset       { <selector>? '{' <S>* <declaration>? [ ';' <S>* <declaration>? ]* '}' <S>* }
    token selector      { <css_any>+ }
    token declaration   { <property> ':' <S>* <value> }
    token property      { <IDENT> <S>* }
    token value         { [ <css_any> | <block> | <ATKEYWORD> <S>* ]+ }
    token css_any       { [ <IDENT> | <NUMBER> | <PERCENTAGE> | <DIMENSION> | <STRING> | <CHAR>
                          | <URI> | <HASH> | <UNICODE_RANGE> | <INCLUDES>
                          | <FUNCTION> <S>* <css_any>* ')' | <DASHMATCH>
                          | '(' <S>* <css_any>* ')' | '[' <S>* <css_any>* ']'
                          ] <S>* }
}

sub MAIN {
    # Should all parse
    my @tests = (
        abc => 'ident',
        ab  => 'ident',
        a   => 'ident',
        abc => 'name',
        ab  => 'name',
        a   => 'name',
    );

    for @tests {
        say "$_ \t", CSS.parse($_.key, :rule($_.value)).Bool;
    }
}