BLS wrote:
> Matti Niemenmaa schrieb:
> 
>> BLS wrote:
>>
>>> Yep. Porting Walter's language/grammar description into EBNF is a
>>> challenge. You may have recognized that I am working on Netbeans IDE
>>> support for D/MiniD... in order to establish code folding I have to
>>> create an NBS Schliemann file which is (mostly) similar to EBNF.
>>> Instead of {}  I have to use ()*, in fact a few minor things.
>>> Probabely we can exchange some information ?
>>
>>
>> A lot of people appear to be working on something which needs to parse D.
>>
>> It might be an idea to set up a page on Wiki4D called "CorrectCompleteDGrammar"
>> or something, writing a complete EBNF (or using Walter's syntax, which is also
>> fine) for D which is actually correct.
>>
> 
> I like this idea.
> 
> // Using this EBNF grammar  ???
> S = ModuleDeclaration {Statement} ";" .
> ModuleDeclaration = "module" identifier {"." identifier} ";" .
>  and so on ...
> 
> /* It will make live easier, when we have  a least our "Standard" EBNF defination.
> */
> Bjoern
> 
> 

What might be fun would be to have one "standard" version in EBNF and then automagically derive others like pure BNF, right recursive, or whatever.

Jascha, many, many thanks.
Just one question :
What is the meaning of : epsilon ?
Bjoern

Jascha Wetzel schrieb:
> Matti Niemenmaa wrote:
>> BLS wrote:
>>> Yep. Porting Walter's language/grammar description into EBNF is a
>>> challenge. You may have recognized that I am working on Netbeans IDE
>>> support for D/MiniD... in order to establish code folding I have to
>>> create an NBS Schliemann file which is (mostly) similar to EBNF.
>>> Instead of {}  I have to use ()*, in fact a few minor things.
>>> Probabely we can exchange some information ?
>>
>> A lot of people appear to be working on something which needs to parse D.
>>
>> It might be an idea to set up a page on Wiki4D called "CorrectCompleteDGrammar"
>> or something, writing a complete EBNF (or using Walter's syntax, which is also
>> fine) for D which is actually correct.
>>
> 
> i exported the current of the D grammar that i use for SEATD:
> http://seatd.mainia.de/grammar.html
> http://seatd.mainia.de/grammar.xml

BLS wrote:
> What is the meaning of : epsilon ?

Usually this means nothing, erm, "nothing" in such a context.

Something like "a <- b / epsilon" is equivalent to "a <- b?".

Regards, Frank

0ffh wrote:
> BLS wrote:
>> What is the meaning of : epsilon ?
> 
> Usually this means nothing, erm, "nothing" in such a context.
> 
> Something like "a <- b / epsilon" is equivalent to "a <- b?".
> 
> Regards, Frank

yep, epsilon means, the production rule may produce nothing. in the seatd grammar, in most cases, non-terminals that have an epsilon rule, also have a name with suffix "Opt" for "optional".
note that epsilon rules are just a convenient shortcut:
EnumDeclaration -> "enum" EnumBaseTypeOpt EnumBody
may be rewritten as:
EnumDeclaration -> "enum" EnumBaseType EnumBody
EnumDeclaration -> "enum" EnumBody