• From: Richard Lanyon <rgl@d...>
• To: xml-dev@l...
• Date: Tue, 28 Nov 2000 10:34:08 +0000 (GMT)

On Tue, 28 Nov 2000, Eric van der Vlist wrote:

> By contrast, to solve this generic issue, our parsers are pretty
> helpless and XPath is rather verbose even though one can use
> "ancestor-or-self::*[@xml:lang][1]/@xml:lang" to find the language of a
> node.

Indeed - currently available versions of XML Script use their own
XPath-like syntax (XML Script was initially developed before XPath
became a Recommendation) one of whose features was easy addressing of
"inherited" attributes. If you imagine a similar technique shoehorned
into XPath, then instead of doing

@bar

to get the bar attribute of the current node, you'd do

!bar

to get the "inherited" bar attribute - i.e. the bar attribute of the
current node, or, if there is no such attribute, the bar attribute of
the least distant ancestor.

In my opinion this convenience notation is the only serious omission
from the XPath spec.

-- 
Richard Lanyon (Software Engineer) |     "The medium is the message"
XML Script development,            |             - Marshall McLuhan
DecisionSoft Ltd.                  |

• References:
  • What about the scope of the XML nodes ?
    • From: Eric van der Vlist <vdv@d...>

• Prev by Date: No Subject
• Next by Date: SVG ?
• Previous by thread: What about the scope of the XML nodes ?
• Next by thread: 2 queries - image in XML and XSL Date
• Index(es):
  • Date
  • Thread