Re: CASL Concrete Syntax - Call for Proposals

[pdmosses@brics.dk]

Magne Haveraaen <Magne.Haveraaen@ii.uib.no> asks:

> >Note that concrete syntax concerns not only the form in which a text
> >is to be input as a linear string of characters, but also that in
> >which it is to be displayed on a 2-dimensional page or screen.  Let us
> >refer to these two aspects of concrete syntax as "input syntax" and
> >"display syntax", respectively.
> 
> Should there also be a requirement that display syntax is a subset of
> input syntax, or, more generally, that there is an injection function from
> display syntax to input syntax?
> 
> Thus it will at least be possible to use an OCR system to scan a printed
> specification and then parse it into the abstract form again.
> 
> A fancy display format should then be accompanied by the definition of
> the corresponding "injection function".
> 
> Magne

The idea of scanning printed CASL specifications, to input and parse
them, admittedly hadn't occurred to us in the Tools (or Language
Design) task groups.  We imagined that documents would always be
transferred electronically, in an interchange format.

So no such requirement of mapping display syntax back to input syntax
has been imposed: all that is required of the display syntax is that
it should be easy for *humans* to read and interpret.  It may well
exploit various visual typographical devices that are not easily
represented as linear sequences of characters, being rendered
essentially at the pixel or outline level (e.g., the frames used in Z).

Of course, use of familiar typographical elements is anticipated: CASL
readers shouldn't be presented with a whole new vocabulary of strange
symbols!  Moreover, it would be an advantage if one could always give
a reasonable approximation of the preferred display syntax in HTML, so
that we might exploit WWW straightforwardly for browsing CASL
libraries.  But such features have not been imposed as requirements -
partly to avoid undue prejudice of the style of display syntax.

Thanks to Magne for raising this issue, which was indeed unclear in
the call for proposals.

Peter

 ----   --------------------------------------------
 \  /  | Peter D Mosses         <pdmosses@brics.dk> |
 CoFI  | Common Framework Initiative  - Coordinator |
 /  \  | WWW URL: http://www.brics.dk/Projects/CoFI |
 ----   --------------------------------------------