Cart
[Home] [By Thread] [By Date] [Recent Entries]
At 10:58 AM 19/03/01 -0500, John Cowan wrote: >>Note (although no processor other than Lark ever did this as >>far as I know) that if you want to build a DFA-based XML >>processor, you can use the trick of recognizing all the syntax >>characters with a 7-bit state table and a remarkably small amount of clever sidestepping is required to deal with all >>the non-ASCII characters. > >Fair enough. But presumably you need a separate recognizer for >XML names. Yeah, but there's not much to it. You need two 64k bitmaps, one for name-start-char and for name-char, and to instrument your state table so you can tell what class of char you're looking for... like I said, not much sidestepping. -Tim
|
