REP

Hello all, I'm going to submit the following REP to feedback: http://www.rebol.it/REPs/PARSE.html however, before doing that, I'd like to be sure that compile-rules.r works correctly and I'd like to have comments from you. Please let me know. Regards, Gabriele. -- Gabriele Santilli <[g--santilli--tiscalinet--it]> -- REBOL Programmer Amigan -- AGI L'Aquila -- REB: http://web.tiscali.it/rebol/index.r

Hi Gabriele, GS> I'm going to submit the following REP to feedback: GS> http://www.rebol.it/REPs/PARSE.html GS> however, before doing that, I'd like to be sure that GS> compile-rules.r works correctly and I'd like to have comments from GS> you. Please let me know. I haven't tried it out yet, but after reading your document it seems like a good idea. I haven't given any thought about deep implications, but I'm not sure EVALUATE is the best word since it doesn't indicate that the following word will be set to a value. What about adding an /EVAL refinement to SET? -- Gregg

Hi Gregg, On Saturday, January 11, 2003, 8:33:28 PM, you wrote: GI> but I'm not sure EVALUATE is the best word since it doesn't indicate GI> that the following word will be set to a value. Hmm, does COPY indicate that the following word will be set to a value? :-) Seriously, if you have any better names (THROW could be confused with the THROW function, so a different name could be useful too), just let me know. GI> What about adding an GI> /EVAL refinement to SET? That could be an option. What do others think? Regards, Gabriele. -- Gabriele Santilli <[g--santilli--tiscalinet--it]> -- REBOL Programmer Amigan -- AGI L'Aquila -- REB: http://web.tiscali.it/rebol/index.r

Why does it need to be a refinement? That would lead one to think that SET is a function. Parse has its own dialect, and SET is just a word in that dialect, so it could just be written as: set evaluate .... But then, why introduce new words when we already have REDUCE and DO, whose functions are well known. We could write just: set reduce ... or set do ... It means less learning of new words is necessary. Anton.

Hi Anton, On Sunday, January 12, 2003, 6:29:29 AM, you wrote: A> a word in that dialect, so it could just be A> written as: A> set evaluate .... That creates a compatibility issue. If you had a PARSE rule that was setting the word 'EVALUATE, it would change its meaning... Regards, Gabriele. -- Gabriele Santilli <[g--santilli--tiscalinet--it]> -- REBOL Programmer Amigan -- AGI L'Aquila -- REB: http://web.tiscali.it/rebol/index.r

Hi Gabriele, beyond your proposal, compile-rule is a wonderfull tool, i'm already using it to build custom new rules like not rule to-first ["a" "b" "c"] and so on

>but I'm not sure EVALUATE is the best word since it doesn't indicate >that the following word will be set to a value. What about adding an >/EVAL refinement to SET?

Of course, you are right. Anton.