Because of

(a) the XML Illegal characters - e.g., NUL (char code 0) is perfectly legal in 
DFDL's infoset, but not allowed in the XML Infoset even if expressed as a 
character entity.

(See https://daffodil.apache.org/infoset/#xml-illegal-characters)

This requires us to have a notation for expressing NUL which does not involve 
use of an actual NUL character in a DFDL schema, nor use of � which is ALSO 
not allowed in XML documents of any kind, including XML Schemas. XML version 
1.0 actually disallows many perfectly good character codes that appear all the 
time in data.


(b) The lack of many usefully named entities e.g., %CR; in XML (There are lots 
in HTML, but only a handful in XML, and they are mostly about escaping XML's 
own notation, not convenience for expressing data or things to be found in data 
such as delimiters.)


(c) the fact that XML standards could invent new notations based on & as the 
introductory character, (such as add a whole bunch of named character entities) 
which would put DFDL in conflict with them.


I suppose that's enough reasons.


-mike beckerle

Tresys




________________________________
From: Costello, Roger L. <[email protected]>
Sent: Monday, March 18, 2019 9:24:23 AM
To: [email protected]
Subject: Why is the syntax used to denote a DFDL entity different than the 
syntax used to denote an XML entity?

Hello DFDL community,

As you know, an entity in DFDL is identified using this syntax: %NL;

An entity in XML is identified using this syntax: &apos;

Why does DFDL use a different syntax than XML?

/Roger

Reply via email to