Julian Reschke wrote:
>
> > From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]]
> > Sent: Tuesday, May 29, 2001 3:01 PM
> > To: [EMAIL PROTECTED]
> > Cc: [EMAIL PROTECTED]
> > Subject: RE: javax.xml.transform: prohibiting Java extensions
> >
> >
> >
> > > So the very presence of a scheme name makes "xalan:..." absolute...
> >
> > I'm afraid not. The presence of a scheme name means you have to look at
> > that scheme's syntax definition to determine whether the URI is
> > relative of
> > absolute. For example, "http:my/local.file" is a relative URI reference
> > even though it specifies the http: scheme.
>
> Nope. See RFC2396:
>
> An absolute URI contains the name of the scheme being used (<scheme>)
> followed by a colon (":") and then a string (the <scheme-specific-
> part>) whose interpretation depends on the scheme.
>
> and
>
> Relative URI references are distinguished from absolute URI in that
> they do not begin with a scheme name. Instead, the scheme is
> inherited from the base URI, as described in Section 5.2.
Julian --
While you are technically correct here, this doesn't really help all
that much. According to RFC 2396 section 5.2, Joe's URI reference
"http:my/local.file" is not a legal URI reference any more. However, I
think that Joe was referring to relative-path references versus
absolute-path references discussed in section 5.
In any event, to get back to where we started, your original message on
this thread said:
"By the way, quoting from
<http://xml.apache.org/xalan-j/extensions.html>:
"Although the namespace declarations for the class and package formats
are
shown with the xalan:// prefix, the current implementation for those
formats
will simply use the string to the right of the rightmost forward slash
as
the Java class name. This format is shown in order to comply with W3C
recommendations for namespace declarations."
This statement is completely misleading. The XML namespace spec says
that
the namespace name should be a URI ref. If you choose to "invent" a URI
scheme as "xalan:", you have complete control of scheme-specific-part,
there
is no explicit requirement for the leading "//". Besides, if "xalan:" is
going to be supported in the future (instead of the XSLT 1.1 draft
method),
at least the URI scheme should be officially registered..."
What should we do so that the extensions.html page is not "completely
misleading". Can I just delete the last sentence of the quoted
paragraph?
Gary
