BNFs are suited to describing the "context-free portions" of a given
grammer..   even a BNF for C is not quite complete as a 'label' which
has been assigned as a function identifier is syntactically distinct
from a label which has been assigned to a variable.

J is very context-sensitive..  any given name may represent a noun, a
verb, and conjuction, etc which all have different syntactic
requirements...   the BNF alone cannot provide the whole story.

 - michael dykman

On Thu, Sep 3, 2009 at 4:50 PM, Sean Stickle<[email protected]> wrote:
> Very nice!
> Regarding the warning at the top, that BNF is "not the most concise, nor
> precise, way of specifying J's syntax". Verbosity is no worry to me, but is
> the implication that a BNF spec of J is incomplete or inaccurate in some
> important ways? If so, knowing those would be very helpful.
>
> Thanks for putting this together,
>
> Sean Stickle
>
> On Thu, Sep 3, 2009 at 16:37, Don Guinn <[email protected]> wrote:
>
>> The URL is messed up. Should be
>> http://www.jsoftware.com/jwiki/Backus–Naur%20Form<
>> http://www.jsoftware.com/jwiki/Backus%E2%80%93Naur%20Form>
>>
>> On Thu, Sep 3, 2009 at 11:20 AM, Raul Miller <[email protected]>
>> wrote:
>>
>> > I have put up a draft copy of a BNF representation of J's syntax on the
>> > jwki.
>> >
>> > http://www.jsoftware.com/jwiki/Backus–Naur%20Form
>> >
>> > If anyone has time, could you look it over, for silly obvious type
>> errors?
>> >
>> > Thanks,
>> >
>> > --
>> > Raul
>>
> ----------------------------------------------------------------------
> For information about J forums see http://www.jsoftware.com/forums.htm



-- 
 - michael dykman
 - [email protected]

Don’t worry about people stealing your ideas. If they’re any good,
you’ll have to ram them down their throats!

   Howard Aiken
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm

Reply via email to