Tokens, from the spec
=====================

Tokens generated by the tokenizer pass, only

* <whitespace>
* <number>
* <string>
* <bad-string>
* <CDC>
* <CDO>
* <EOF>
* <colon>
* <comma>
* <semicolon>
* <rbrace>
* <rbrack>
* <rparen>
* <lbrace>
* <lbrack>
* <lparen>
* <column>
* <dash-match>
* <prefix-match>
* <substring-match>
* <include-match>
* <suffix-match>
* <delim>
* <at-keyword>
* <hash>
* <bad-url>
* <dimension>
* <function>
* <ident>
* <percentage>
* <url>
* <unicode-range>