Re: an infinate loop

[Ted Dennison<dennison@telepath.com>]

In article <3B56C6B2.559EE5F3@bsdger.org>, Matthias Kretschmer says...
>Another way of implementing a state machine would be to make an array of 
>subprogram access types to subprograms handling the different states. This 
>could be faster than using case with using the current state as an index to 
>the array. This should be nearly as fast as using goto, just it doesn't use 
>fixed jump-points. But if you just implement features because of speed 
>enhancements of special problems, than you should implement everything 
>including an assembler. Goto doesn't make it look worse, but make it may be 
>more complicated. Using for each state a seperate subprogram would be probably 
>make the whole stuff more appealing to the eyes of other people.

To pull this off, you'd still need a loop around everything so you could iterate
through the array. This is essentially the same as a case statement in a loop,
except that you are replacing the "case" with an array. Another difference is
that you now have subprogram call overhead for each state change.

Often jump-table type code can be replaced with dynamic distpatch too, and that
allows users to arbitrarily modify the routines in the jump-table without
modifying the existing (debugged) sources. If this line of reasoning in lexical
analysis interests you, you really ought to look into OpenToken, which
implements it. See http://www.telepath.com/dennison/Ted/OpenToken/OpenToken.html