On Fri, 5 Sep 2003, Peter Gibbs wrote:
> On Fri, 5 Sep 2003, Dan Sugalski wrote:
>
> > Go for it. There is the possibility that identifying a digit is more
> > involved than we might otherwise want for some chartypes, however--the
>
> Yeah - I am hoping to handle the simpler cases generically, so that we only
> need to write specific code for the less simple ones.
> The current methods for both digit handling and transcoding are
> context-free,
> which I suspect may become a problem later; if so, some form of iterator
> with
> context information will be required.
Which should be just *so* much fun... :)
Since you're modifying the struct, make sure there are entries for
*functions* that do all the things you're putting in pointers to data
members for. We can NULL them out for now, but it'll mean that when
someone throws the Shift-JIS chartype code in we won't have to change the
struct and rebuild the world.
Dan
