>because I use up time just testing the string length before I get to the
>table lookup.
>
>I'm willing to sacrifice memory, inflexible and ugly code in order to
>test the idea of getting nearly instantaneous jumps but none of the 
>ideas I have are fast.
>
>Any ideas?
>
Wave,

I have a number of ideas, but it's hard to imagine what could need faster 
than a mere 40-element select. I'm betting any improvement will save only 
a few second over many millions of iterations.

The first thing to consider is, you have 2 to 5 data: lenByte, char1, 
char2, char3, char4. The lenByte is an important determinant, so you want 
to use it. Could any of the others be dropped? (i.e., if you see a 4-char 
string beginning with "SI", is it safe to assume it will be "SIMM"? How 
about "SIM"? The fewer determinants you have to use, the faster. If the 
list is only 40 long, send it to me and I'll tell you how many you need, 
(assuming any entry will be one of those 40).

If you tell me exactly what it is you're doing, I may be able to tailor a 
solution that will save a bit, but I doubt it will save much. If you're 
short on speed, there's a good chance you can find bigger culprits in 
places other than your branching code.

 e-e
 =J= a  y
  "