[stringtemplate-interest] Idea for passing parameters to another template...

[Mark Wharton]

Thanks Terence,

OK, that makes sense. I suppose the order of arguments requirement could become difficult to manage at some point. I have implemented both to see how it goes, but will most probably end up sticking with the ST way for compatibility reasons.

Cheers,

Mark

On Sat, 15 Mar 2008 18:33:05 -0700, "Terence Parr" <parrt at cs.usfca.edu> said:
> Hi Mark,  seems ok, but I've kinda assumed attribute lists aren't  
> ordered as they are named with <foo(a=x, b=y, ...)>.  No reason one  
> couldn't design a system with ordered args.
> Ter
> On Mar 15, 2008, at 2:44 AM, Mark Wharton wrote:
> 
> > Hello,
> >
> > I'm looking for feedback on an idea for passing parameters to  
> > another template...
> >
> > In "Passing Parameters To Another Template" section of http:// 
> > www.antlr.org/wiki/display/ST/Expressions it states:
> >
> > "If you are using StringTemplate groups, then you have formal  
> > parameters and for those templates with a sole formal argument, you  
> > can pass just an expression instead of doing an assignment to the  
> > argument name. For example, if you do $bold(name)$ and bold has one  
> > formal argument called item, then item gets the value of name just  
> > as if you had said {$bold(item=name)$}."
> >
> > Could this be extended to support multiple arguments, where just  
> > expressions can be passed for each of the formal arguments in a  
> > template?
> >
> > Take the following for example:
> >
> >   group sample;
> >
> >   main(a, bitems, c) ::= "<bitems:item(a, it, c)>"
> >
> >   item(a, b, c) ::= "a = <a>, b = <b>, c = <c>"
> >
> > In the above example, the template application in main is passing  
> > (a, it, c) where 'it' is eventually assigned to 'b'. It fits well  
> > in the context of formal arguments and interfaces etc. It really  
> > makes sense in template application, for regular template includes  
> > I don't believe it makes so much sense, but it's basically OK. For  
> > me, it's clearer than the equivalent "<bitems:item(b=it, ...)>",  
> > which is how it works in the current StringTemplate implementation.
> >
> > Does the idea seem reasonable, or does it violate the basic rules  
> > of StringTemplate? My reason for asking - I'm thinking about  
> > supporting this feature in a lightweight implementation of  
> > StringTemplate which I'm working on, but I don't want to completely  
> > mess up on the compatibility! If StringTemplate could support this  
> > one day then great, otherwise, I'll most probably drop the idea and  
> > just support assignment to the argument name, as in (b=it, ...).
> >
> > Thanks,
> >
> > Mark
> >
> > Mark Wharton
> > + 8190 9834 2559
> > contact at moonbase.com.au
> >
> > _______________________________________________
> > stringtemplate-interest mailing list
> > stringtemplate-interest at antlr.org
> > http://www.antlr.org:8080/mailman/listinfo/stringtemplate-interest
> 
Mark Wharton
+ 8190 9834 2559
contact at moonbase.com.au