hi folks,

I'm trying to write a Parse::RecDescent chunk that will split up a line of input in to <tags> and non-tags such that

"<this>this is tagged</this> <bye>bye!</bye>";

becomes a list suchlike -

[
   '<this>',
   'this is tagged',
   '</this>',
   ' ',
   '<bye>',
   'bye!',
   '</bye>'
]
[download]

the below works, but only if TOKEN is specified as 'TAG | LITERAL' .. i'm sure there are a billion things i don't understand about Parse::RecDescent

$Parser = Parse::RecDescent->new(
   q(
      TAG       : <skip:''> /\\<.*?\\>/
      LITERAL   : <skip:''> /[^<]*/
      TOKEN     : TAG | LITERAL

      startrule:
         TOKEN(s) {
            @main::PARSED = @{$item[1]};
         }
   )
);
[download]

can anyone shed any light on why TOKEN would not evaluate in an OR | context properly if the arguments are reversed? (any other comments on what i may be doing improperly are more than welcome)

thanks!

Edit kudra, 2002-10-27 Replaced code tags around entire message with markup