[
Lists Home |
Date Index |
Thread Index
]
juanrgonzaleza@canonicalscience.com said:
> Rick Jelliffe wrote:
>>
>> Because XML parsers are available in every language built-in, while LISP
>> and scheme parser are not ubiquitous, have less good
>> internationalization,
>> encourage rather than discourage the addition of processing, and does
>> not
>> have validation languages.
>
> Whereas, i am not claiming that both SXML and Scheme were last words in
> datuments manipulation i would recommend a view to
>
> [http://okmij.org/ftp/Scheme/SXML.html]
>
> and specially to
>
> [http://okmij.org/ftp/Scheme/xml.html]
>
> The thread about XML syntax for XPath begins to be a bit hot ;-)
>
> I would note that in the Scheme approach, SXPath is a SXML structure can
> be analised, modified, and even automatically generated in the way than a
> hypotetical XPathX would but with the advantages of a full programming
> environment (Scheme).
They are good pages indeed.
The connection between S-expressions and XML is well understood. There
used to be a great publishing system called Interleaf in the 80s and 90s
which was LISP-based and did amazing SGML publishing: we used it for
military publishing and it was great (inserts in different size paper,
etc.) About 17 years ago, I co-wrote a custom version of LISP with a
built-in mini-SGML parser for document processing, called RISP, and a few
years later commented on com.text.sgml that LISP was workable (e.g. for
ISO DSSSL, which is a Scheme dialect.)
For a different view on LISP, also see my "LISP is better than XML, but
worse is better" at
http://www.oreillynet.com/digitalmedia/blog/2004/12/lisp_is_better_than_xml_but_wo.html
which includes a bit of a dummy-spit by me at the end.
Cheers
Rick Jelliffe
|
