Re: Some seeming context sensitivity, and shift/reduce errors

[Hans Aberg]

On 9 Dec 2007, at 09:50, Jack T Mudge III wrote:

> > [Please keep the cc to the Help-Bison list, as other may want to join
> > in.]
>
> I thought it was; must have forgotten it. Apologies, Kontact doesn't
> automatically copy the CC line (yet).

Don't know about your mailer, but often there is a special key  
combination or menu entry "reply all".

> > One can push the context sensitivities into the actions. This way,
> > one can even handle language like C++. An alternate approach is to
> > use the GLR parser (see the manual).
>
> Some way to tell what tokens appeared in a context?

One can tweak the lexer, so that it gives different tokens depending  
context. A common from of conetxt sensitivity is a definition: before  
the definition, it is a name, but it has a different syntax  
afterwards. Then let the lexer look up names in a look-up table, and  
based on that, return token value to the parser. The parser can then  
change that. Also check out "start conditions" in the Flex manual.

> > Look around for some similar languages, which helps avoiding
> > pitfalls, and may give ideas on grammar design.
>
> The syntax of this language is based mostly on that used by  
> didiwiki. I would
> perfer something like a simplified TeX, myself, but didiwiki's  
> syntax is
> friendly to my mom's ignorance and my dad's need for voice recognition
> software. Since DidiWiki seems intent on removing itself from the  
> web, I
> figured there's no time like the present to replace it (and perhaps  
> improve
> upon it, but I'll save that for after replacing it).
>
> The source is all but inaccessible at this point, but I assume the  
> syntax
> *bold*, /italic/, _underline_, -strike through-, [link], etc. are  
> common
> among at least earlier wikis. I guess the source for some other
> implementations will be my next stop while I wait.

Perhaps somebody else can give suggestions. TeX has stuff in it,  
making it hard to treat with a Flex/Bison combination.

  Hans Aberg