Announcing a Preview of XString, an XML technology - XML as a String

[wgilreath]

Announcing a Preview of XString, an XML technology - XML as a String.

I've created a means to efficiently encode XML as a flat string,
which I term XString. I've written a paper, put some examples, and
have created an online demonstration that can process XML into XString.

My goal in annoucing/previewing the technology is to get comments and
feedback from others to inspire and guide future efforts.

Online at www.williamgilreath.com/xstring.

Synopsis:

XML requires a parser to process and tree structure to represent the
markup. XML by its very nature creates a verbose, bloated document.
Such a big, unwieldy structure is complex to handle.

XString is an efficient encoding for XML as a long string, but not as
large, complex, data structure. XString compresses XML into a more
compact representation, and one that is manipulated as a string. The
syntax style is similar to XPath, so familiar and not completely
unusual in representation.

The five major points about XString:

XString is XML just encoded differently.
XString is similar in syntax to XPath so familiar.
XString is more compact (compressed) in representation so less data
bloat.
XString is useable where ever XML is used, such as SOAP or XHTML.
XString allows XML to be embedded in XML.

Quick-and-dirty example of XML and XString:

XML:

<?xml version="1.0" encoding="ISO-8859-1"?>
<duh:ROOT ATTRIBUTE="VALUE" xmlns:duh="http://www.w3.org/noduh">
<duh:CHILD>text</duh:CHILD>
<duh:EMPTY EMPTY_ATTRIBUTE/>
<!--comment-->
<?processing instruction?>
<![CDATA[function nop(){}]]>
<duh:NIL></duh:NIL>
</duh:ROOT>

XString:

/duh:ROOT+7@ATTRIBUTE=VALUE@xmlns:duh=http://www.w3.org/noduh/duh:CHILD+1'text/duh:EMPTY+0@EMPTY_ATTRIBUTE-comment?processing
instruction[function nop(){}/duh:NIL+0


Cheers! My best,

Will Gilreath
(e-mail address removed)
(e-mail address removed)

 

[Joe Kesselman]

Great; yet another pretokenized-XML syntax. This has been done
repeatedly, and in general turns out to be much less useful that it
initially appears -- partially because some off XML's value is tied up
in the fact that it is fairly human-readable, partly because XML parsing
is actually faster than most folks think it is, partially because you
can get better size reduction by simply putting standard XML through
standard compression algorithms.

Which problem are you trying to solve, and have you run the numbers to
confirm that your solution is actually a good one?