Jason,
If you just want a stand-alone database than using JavaCC will be fine, but
if you are trying to make something that will work with embedded devices as
Hypersonic was than using JavaCC is probably not something that you will
want to use.
Well being on Pheonix, we're not ordinarily targetting embedded devices.
However it is a good point. I think that we could support multiple
implementations of the SQLParser interface, some robost, some stripped
down.
If I remember correctly a JavaCC grammar with one rule will produce a
parsing mechanism that's over 100k. Velocity has a relatively small grammar
and the parsing mechanism weighs in at over 1mb. The grammar for SQL is by
far the largest I've seen and you will probably get something huge. Totally
guessing I would say 2-3mb. If you don't care about embedded devices than it
probably doesn't matter, but JavaCC makes robust parsers not small ones :-)
Mckoi's weighs in at 360K (in source form). It's probably not complete.
The well made point here is that we'd be victims of our own sucess.
Hypersonic is interesting in that it has no such well made parser but
still manages to do the job.
Multiple implementations is the key. Horses for courses et al.
Regards,
- Paul H
--
To unsubscribe, e-mail: <mailto:[EMAIL PROTECTED]>
For additional commands, e-mail: <mailto:[EMAIL PROTECTED]>
