MoinMoin: BisonGen

BisonGen creates LALR parsers from a formal grammar input file, as C python extension using Gnu Bison and in pure python.

• http://www.gnu.org/software/bison/manual/html_mono/bison.html

• ftp://ftp.fourthought.com/pub/BisonGen/

• http://docs.python.org/ext/building.html

• irc://irc.freenode.net/4suite

• http://meta.wikipedia.org/wiki/Wikipedia lexer

• http://en.wikipedia.org/wiki/Wikipedia:How_to_edit_a_page

Some usage notes

Lexer states

In the <lexer> section it's possible to define states which determine which token patterns the lexer is looking for while scanning:

{{{ <states>

• <exclusive>OPERATOR</exclusive>

• </states>

}}} The default state is INITIAL, this adds an additional OPERATOR state. <exclusive> tells the lexer to only apply OPERATOR rules when in this state, anything else like <inclusive> will extend the INITIAL (toplevel) rules if the state is active. Two states can be active at the same time, INITIAL and one inclusive state. Rules for a state are placed inside a scope block:

    <scope state='OPERATOR'>
      <pattern expression='or'>
        <begin>INITIAL</begin>
        <token>OR</token>
      </pattern>
<!-- more rules -->
    </scope>

The begin element determines the next state to go to when the rule matches, in this case it would go back to INITIAL. The same is used in the top-level INITIAL rules that aren't placed inside a special scope block:

     <pattern expression='\)|\]'>
      <begin>OPERATOR</begin>
      <token>@ASCII@</token>
    </pattern>

This one would switch to the OPERATOR state..