Hello,

those who are still interested in "lisp vs xml" and "xml is a poor copy
of s-expressions" issues, might find my recent writings interesting:

Towards s-expression based XPath/XSLT implementation
http://xmlhack.ru/protva/xquery/index.php/TowardsXslt

Abstract:

It's supposed that Lisp languages are ideal for implementing XML
standards. Then why we don't have an XSLT processor written in Common
Lisp or Scheme? I'm ignoring business issues and want to expose a
technical problem.

The popular representations of XML become showstoppers in implementing
hidden XML/XPath/XSLT features. In this paper, I try to list the
issues, both generic and specific to the representations:

* object oriented,
* DSSSL,
* SXML.

Also I record some low-level issues related to implementing XPath and
XSLT.

Finally, I introduce GSXML, a dialect of SXML.



--
Oleg Parashchenko olpa@ http://xmlhack.ru/ XML news in Russian
http://uucode.com/blog/ Generative Programming, XML, TeX, Scheme
XSieve at XTech 2006: http://xtech06.usefulinc.com/schedule/detail/44

Hello Stefan,

thanks for pointing to Unotal. I've read the documentation and found
the approach is quite interesting. But notice that there is a big
difference in our approaches. You define a superset of XML, and I
don't. I use SXML as an in-memory representation of XML, useful for my
developments.

I also have started a link list for "S-expression like markup languages".

Have you seen "XML Alternatives" by Paul Chistopolsky?
http://www.pault.com/xmlalternatives.html

By the way, Google Groups says about your message: "The author of this
message requested that it not be archived. This message will be removed
from Groups in 5 days". Maybe it's not what you want.


Stefan Ram wrote:

"Oleg Paraschenko" <olepar@gmail.comwrites:

>
I wrote a Reader for my S-expression-like notation "Unotal" in Java.
I decided to always represent elements as maps in memory,
which is a good representation for attributes. The "body" of
an element then is represented as the value of a special
"__body" attribute. So, while SXML has decided to represent
everything as a list, I took the other direction by
representing everything as a map in the first place (however,
the value of the "__body" attribute then is a list).
>
I have specified a variant of S-expressions named "Unotal",
which might be described as "S-expressions with attributes".
Here are some notes comparing it to plain S-expressions, which
also address attribute-representation with S-expressions:
>
http://www.purl.org/stefan_ram/pub/s...ared-to-unotal
>
And the general Unotal page:
>
http://www.purl.org/stefan_ram/pub/unotal_en
>
(A Java-Reader and -Writer for Unotal works, but lacks
documentation. It is available under the GPL on request.)
>
I also have started a link list for "S-expression like markup
languages".
>
http://purl.net/stefan_ram/garnoo/XO...KUPXGLANGUAGES