DGrammar - DMachine

Sjoerd van Leent svanleent at gmail.com
Thu Jun 15 09:21:43 PDT 2006


pragma schreef:
> In article <e6plrg$1usq$1 at digitaldaemon.com>, Sjoerd van Leent says...
>> Hello All,
>>
>> I stopped development at DGrammar months ago (perhaps more than a year 
>> ago). Nevertheless, I have come back and have made some decisions about 
>> what to do with it.
>>
>> 1 ) DGrammar was build using YACC, and thus C, C++, Flex and Bison where 
>> necessary to get the beast compiled, and it didn't even do it's job the 
>> way I would like it.
>>
>> 2 ) DGrammar was a bit of "make one messy file" architecture, which of 
>> course isn't about a professional environment.
>>
>> 3 ) I've began at first writing boiler plate Machines. The first one is 
>> a PDA Machine (Stack machine), and I like it to be tested. Simply run 
>> "build" over dmachine.d. And enter something like:
>>
>> "dmachine aabbbbbaa"
>>
>> This is just an example, try to make some machines with it, I try to 
>> make it working better with the lambda problem I'm still having (it is 
>> solved Q&R now, which isn't the way I like it, any solutions would be 
>> welcome)
> 
> Sjoerd, have you considered using Enki to help you get development restarted?  I
> know that the two projects share some basic similarities - heck, DGrammar is
> what inspired me to write it in the first place.
> 
> Anyway, the whole thing is BSD licensed, so feel free to use it as you see fit.
> 
> - EricAnderton at yahoo

Definitely consider Enki. The only difference is the way it is 
processed. At the end I want to let DGrammer use a "BDF" styled input 
file, and let it generate a parser.

When that's done, I'd like for it to be able (just as in the first 
trials) to run it through some kind of interface architecture, making it 
a two-step process: Parsing and Processing, just as SAX does.

Regards,
Sjoerd



More information about the Digitalmars-d mailing list