[XML-DEV Mailing List Archive Home] [By Thread] [By Date] [Recent Entries] [Reply To This Message]

Re: Create XML

• To: <xml-dev@l...>
• Subject: Re: Create XML
• From: Arjun Ray <aray@n...>
• Date: Fri, 27 Jun 2003 23:58:07 +0000
• In-reply-to: <p04330100bb21da0deaee@[192.168.254.4]>
• References: <20030620034910.1E6EB13383E@b...> <3EF29241.8050508@j...> <3EF34E1E.2090805@t...> <07c901c337bb$9c124180$4bc8a8c0@A...> <p04330100bb21da0deaee@[192.168.254.4]>

Elliotte Rusty Harold <elharo@m...> wrote:

| But since most attributes either are CDATA or are treated as CDATA, 
| that only works as long as there are no runs of more than one white 
| space character, and no leading or trailing white space. Attribute 
| value normalization by the parser rarely goes as far as XPath's 
| normalize-space does.

True, and arguably normalize-space doesn't go far enough either (though
enough for XPath only).  The basic distinction between CDATA and the other
declared value types has to do with (white-space delimited) tokenization.
XML 1.0 missed the boat here, in that tokenization still needs information
from the DTD.  We might have dedicated the LITA delimiters (single quotes)
for this as a purely syntactic distinction, for instance. 

• References:
  • Re: Create XML
    • From: Uche Ogbuji <uche.ogbuji@f...>
  • Re: Create XML
    • From: James Clark <jjc@j...>
  • Re: Create XML
    • From: Tim Bray <tbray@t...>
  • Re: Create XML
    • From: "Rick Jelliffe" <ricko@a...>
  • Re: Create XML
    • From: Elliotte Rusty Harold <elharo@m...>

• Prev by Date: Re: XML-based Automation (Was: Zen or Games?)
• Next by Date: RE: OFFTOPIC: Job Opportunity
• Previous by thread: Re: Create XML
• Next by thread: RE: Create XML
• Index(es):
  • Date
  • Thread