How can I get a YACC-type parser to stop on the shortest matching input?

Question

Here's the context: I have a file (or stream), and I want to process its header, stop, and then deal with the rest of the data differently. That is, suppose I have a stream like BEGIN GOOD GOOD GOOD GOOD GOOD DONE JUNK JUNK JUNK JUNK .... The first part (BEGIN GOOD+ DONE) matches my grammar, and I want to stop as soon as I see DONE.

AFAIK, YACC and friends want to keep consuming tokens until there are no more, so I have to write more grammar to match all the JUNK. Is there a way around this?

I'm using PLY (a Lex and Yacc implementation for Python), but I'm open to general answers here.

Answer 1

In normal yacc, you can use the yacc macro YYACCEPT in an action to immediately return a successful parse without reading any more tokens. I don't know if PLY supports that, however.