[R6RS] R6RS Unicode SRFI controversial issues

Anton van Straaten anton
Mon Jun 27 14:43:14 EDT 2005


Manuel Serrano wrote:
>>But as I said, I'm much more interested in having support for blocks of 
>>unescaped text, than I am in the exact syntax or features used to 
>>achieve that.
> 
> We could also consider the syntax we have designed for Skribe. It is
> available at (see section 2.1):
> 
> http://www-sop.inria.fr/mimosa/Manuel.Serrano/publi/jfp05/article.html
> 
> 
> I think that it is superior to the #<< syntax because it allows escapes
> inside texts.

I think it's important to have a way to quote text without escapes, as a 
low-level feature which tools can build on.  With a syntax for unescaped 
text, arbitrary escapes can be handled simply by wrapping quoted text in 
a procedure call which performs the desired escapes.

This would allow libraries to support different escape mechanisms, 
rather than imposing a single escape mechanism that may not be 
appropriate for some applications.

The Skribe syntax provides a good example of the issue: while I'm sure 
it's an excellent format for general documentation, it doesn't appear to 
be as good for embedding C-style syntax in a Scheme program.  The "]" 
character is common in C-like languages, and the ",(" sequence can 
easily appear in an argument list.  Both of these would need to be 
quoted in the Sk-expression syntax.  Finally, the quoting mechanism 
itself is likely to add the need for quotes of the quote sequence, in 
some cases.

The only general solution I can see here is to have a syntax that allows 
arbitrary text to be quoted, without having to worry about quoting 
undesired escape sequences.

Anton



More information about the R6RS mailing list