On Fri, May 9, 2008 7:01 pm, chromatic wrote: > Per the TODO in the most recent summary, here's a very preliminary > implementation of the ... operator. I have only the barest test case, and > submit this patch only in the hopes that if it's a very wrong approach, > someone will offer a better way. > > This patch needs a regen_perly.pl. Yay. But I have the feeling that leaves bison disambiguating, when we'd rather have toke.c do it, by returning a different token for unary ... when expecting a term.Thread Previous | Thread Next