World: r3wp

I've also had a look inside xml-parse, it seems to be really like 
SAX - ready to use. But nobody is maintaining it, I think. As far 
as I understand, somebody could create a Handler to get the desired 
block structure (for instance a Handler for RebXML or any other model). 
I have to learn about this in REBOL.


A question: how can I measure memory for a block or an object tree 
in REBOL?

RebXML: I did some testing with rebxml, the documents I used can 
be found here:
 http://www.simplix.de/rebol/resources/xml/xmltests.zip

There is also a simple script that reads the XML docs in and writes 
them back.

Some problems I found:
- empty attributes, I have fixed this in the zip

- entities in content: all should be escaped, because they can be 
found there, otherwise a " gets "
- comments after last element missed
- comments before first element - missing line feed
- missing PIs in output


Another question: encoding - it seems that all output files will 
be written in iso-8859-1 ?

I have no idea about comparision of XML documents (input and output 
of rebxml for instance ) to ensure correctness, but it seems to be 
difficult.

About memory for block or object, If you mean in bytes internally 
in REBOL, I don't know. But you could save the block or object to 
a file and see a size that way. You can of course see the length 
of a serie with: length?

About encoding in RebXML, rebxml2xml let you produce utf-8 by specifying 
the /utf-8 refinement:
rebxml2xml/utf-8 <some rebxml data>

With length? i need some recursion, otherwise I get only the first 
level of the block if it is nested? How to serialize an object tree 
in REBOL - is there some function available?

Carsten, a recursive function to count length of blocks with nested 
blocks:

total-length?: func [b [block!] /local n] [
	n: 0
	foreach e b [if block? e [n: n + total-length? e] n: n + 1]
]

total-length? will count elements, and another block is also an element.

John: Thank you, I'll play with it.

I found this python tool - maybe some interessting ideas there:
http://uche.ogbuji.net/uche.ogbuji.net/tech/4suite/amara/quickref

He uses objects but I like the idea for accessing xml - replacing 
the dots with slashes it looks for me like REBOL:
doc/a/nodeName
doc/a/b/1
...
doc/xml

Catching up a little.  Be interesting to summarise this thread as 
there are many different ideas expressed.  rebxml looks interesting 
for loading, saving and likely extracting xml, but still perhaps 
difficult to manipulate.

note: this group isn't showing on the web site, is this due to [web 
public] instead of [web-public] ?

I've also noticed a tendency to kick the DOM (no doubt for good reason) 
-- though worth noting that it is a complete api to xml and it is 
a standard api, I wouldn't underestimate the value of the latter, 
particularly when it comes to Rebol advocacy...

RebXML is meant for conversion to/from the RebXML format and other 
formats (incl. XML). I use the RebXML format with NicomDoc, which 
makes it a lot easier to handle document formats. Let's say, you've 
got an XML file, and want to convert it to a format easily read by 
some application, then you first use xml2rebxml to get the XML file 
to RebXML format. Then make a converter from RebXML to the final 
format by renaming the rebxml2xml script and change it to do the 
output, that is wanted. rebxml2xml holds the structure of the RebXML 
format, so it's easier to start with that script. Search for "output" 
in rebxml2xml.

Maybe I should make a converter from RebXML to some format very easily 
manipulated directly within REBOL, like the python tool, Carsten 
found.

But this is the issue here with Rebol and XML, there are solutions 
that suit one XML operation or another.  Aiming for loosely implementing 
DOM gives us loading, extraction, modification, and saving without 
affecting the integrity of the data structure.  Examples: changing 
the title of an HTML page, adding an entry to an RSS file, etc.

Using DOM methods, you can do this albeit clumsily, but completely. 
 All through a set of standard functions, with no need to manipulate 
the structure directly.

hmm, couldn't we just somehow mix the aproach, so to have some streamed 
dom? :-) I don't like the idea of having 10MB XML interchange file 
to load into memory ....

Any less than you'd want a 10mb Rebol interchange file?  What % of 
cases would this be an issue?

xml is used to store word-files, rebol not? :)

in the moment i play a little bit with xml-parse.r, it has a lot 
of things done, some are still open (like  <!ENTITY ...> parsing) 
and it is like SAX - I try to implement some handlers to learn REBOL, 
but it's still in progess. A benefit of xml-parse is, that there 
would be only one parser and some kind of standard API and the handler 
could then generate rebxml or some other desired format

DOM: in java APIs there were allways problems with dom - big amount 
of memory, not optimized for a language, so there was a need for 
optimized tools like JDOM, XOM or DOM4J, they all prefer SAX for 
parsing and have their own internal model - of course the API is 
special for all these tools and no standard like DOM

I guess in rebol we have fewer problems than java, as rebol is dynamic 
and java has to emulate that? So it cant map its own classes because 
the format is not known at compile-time? While we can. And then xml 
in memory should be in the order of rebol-blocks?

out of the blue, can anyone point me to the (or one) official XML 
spec ? (if there are many, it should be the one most used on windows 
and in things like PHP)

thanks!

http://www.w3.org/TR/REC-xml/

thanks Chris !

will be reading top to bottom ...  not that this is any fun...  ;-)

Has somebody already give a try to a SAX implementation ?

http://tech.motion-twin.com/xmllight.html

xml is such bloat.. I am parsing xml these days and for two characters 
of data, I often have a 100+ characters of nested stupidity.

an empiric test (subjective to the xml structure and tag names obviously, 
but this IS a real world xml file)

693 kb in xml form   ==>  90 kb  in nested rebol blocks

I left the tabs at 2 spaces in the rebol output, so that the comparison 
is fair.

no need to convince us :-)

Agreed.  So, write a Rebol block ML that does everything as well 
as XML, and we will support it.

manual resync...

my god reading the w3c spec for XML is insane.

XML overcomplicates soooo many things.  its like the standard, for 
people who can't make up their minds:


You can do this, or that or this too, but only when this and that 
or this occur outside and inside that other thing.

XML was intended to be a simplification of SGML.

But they forgot to ask first "why is SGML apparently some complicated?"

So they ended up adding back in most of the complications in an ad 
hoc way.

XML was a simple 2 page spec originally.

I think that might be why the microformats are taking off. They  
use XML in its simplest, intended form.

I'm on a list discussing, inter alia, CCR .. which stands for continuity 
of care record. It's XML, and so guys are saying it's taken them 
50,000 lines to write the parsing code etc.

Possibly an exaggeration on their part.

I think not, Graham .... we have such a problem ... big corporation, 
we try to define xml formats. The trouble is, big products do wrap 
it for you, but what about smaller companies?

not to mention browser incompatibilities, because in the case of 
XML, browser is your "preview" interface ...

If you need a simple XML spec, don't forget my RebXML: http://home.tiscali.dk/john.niclasen/rebxml/
(Only a couple of pages.)

It's an easy way to work with XML inside REBOL, and on the same page 
you'll find scripts for converting between XML and RebXML.

I think that current situation of XML world shows, that it is rather 
absurd situation .... the idea of automatic exchange of docs, their 
specs, without involvement of ppl, is naive ....

I remember talk a few years ago, that MS would make their .doc format 
XML based. And people thought, that would mean, it would become an 
'open' format, which could now be read and written by any wordprocessor. 
If you have somehing like:
#{78797A7138373837}
in binary, and choose to make it into XML:
<xyzq>8787</xyzq>

does that make you know, what it mean? No, of course not. Some MS 
employee later told in an interview, that MS of course would guard 
their IP - 'intellectual property'. I don't know, where the story 
ended, and I don't care much, as I keep away from MS formats.

If you choose to have your doc formats in something like the XML-based 
OpenDocument, where every tag is explained in an open specification, 
then it can be used. (Only problem then is XML itself with it's problems.)

saved out a 15 cell spread sheet in microsoft xml yesterday...  58kb 
of data HAHAHHAHAHAHAHA

geomol, does RebXML handle namespaces and unicode?

Geomol, just looked over rebxml...  I've build a similar engine, 
even simpler actually.  but it might become a little bit smarter 
in a few weeks... maybe supporting more of the XML 1.0 specs like 
&chars conversion and such.