Re: Draft - Fixup or Full XML Parser

On 2/21/2012 9:19 AM, Anne van Kesteren wrote:
> On Tue, 21 Feb 2012 14:16:48 +0100, David Lee 
> <David.Lee@marklogic.com> wrote:
>> My personal opinion is that the XML ER should be speced as the fixup 
>> parser only and not presume that it is a full XML parser.  I think 
>> this will save us a lot of work, and provide more value.
>> Comments ?  Objections ? Am I passed left field ?
>
> How would you envision this "fixup" to work? What you describe sounds 
> like 1. determine whether it needs fixup; 2. fixup; 3. parse. The 
> alternate approach is just parse, which seems somewhat more 
> straightforward.
>

I envision this (in one type of environment) as an element of a toolchain:

XML Parser ---> Tools that deal with well formed XML
        |                               ^
        V                              |
XML-ER Parser ---------

So in my world I would try to use a traditional (and therefore speedy) 
XML parser.  If that failed, failover to XML-ER where it would fix-up 
the input and what comes out would be tool-chain ready.

Maybe I am now guilty of presupposing a solution...

-- 
Shane McCarron
Managing Director, Applied Testing and Technology, Inc.
+1 763 786 8160 x120

Received on Tuesday, 21 February 2012 15:48:46 UTC