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

Re: xml:base and fragments

• From: Peter Flynn <peter@silmaril.ie>
• To: xml-dev@lists.xml.org
• Date: Mon, 8 May 2017 19:51:22 +0100

On 05/05/17 17:48, Liam R. E. Quin wrote:
[...]
> There's no expectation in HTML at least that following a link to
> #apple will involve fetching a different resource - rather, base,
> and xml:base, tell us where we really are when we do that.

We can now muddy the water by asking where the target 'apple' links to
in the example below, if we are processing valid XML and the schema/dtd
declares 'target' as IDREF.

<div xml:base="http://www.dictionary.com/a.html">
  <list>
    <item xml:id="apple">
      <label>Apple</label>
      <p>A fruit, and a <link uri="http://www.apple.com/">make
         of computer</a></p>
    </item>
    <item xml:id="pear">
      <label>Pear</label>
      <p>A fruit, certainly not an <ref target="apple">Apple</ref>.</p>
    </item>
  </list>
</div>

The 'uri' attribute has an absolute value, so it overrides any value
given in xml:base. The 'target' value is no longer a #fragment, and so I
assume it is regarded as out of scope for resolution wrt the xml:base.

I am clearly missing something here: why are people using #fragment
identifiers in XML (non-XHTML/HTML5) documents (eg TEI)?

///Peter

• Follow-Ups:
  • Re: xml:base and fragments
    • From: "Simon St.Laurent" <simonstl@simonstl.com>

• References:
  • xml:base and fragments
    • From: "John P. McCaskey" <mailbox@johnmccaskey.com>
  • Re: xml:base and fragments
    • From: "Liam R. E. Quin" <liam@w3.org>