W3C home > Mailing lists > Public > public-xml-core-wg@w3.org > May 2006

RE: Canonicalization xml:base processing

From: Richard Tobin <richard@inf.ed.ac.uk>
Date: Tue, 30 May 2006 18:18:53 +0100 (BST)
To: "Grosso, Paul" <pgrosso@ptc.com>, "Richard Tobin" <richard@inf.ed.ac.uk>, "Konrad Lanz" <Konrad.Lanz@iaik.tugraz.at>
Cc: <public-xml-core-wg@w3.org>
Message-Id: <20060530171853.46D036D6D18@macintosh.inf.ed.ac.uk>

> Just to check my understanding of one small part,
> when you say the base URI at some point is some
> URI whose last segment is a file name and/or fragment,
> what does this mean?  I thought a base URI was 
> effectively always a directory.

That depends on what you mean by "effectively".  Let's ignore the
fragment case, because it works differently in RFC2396 and RFC 3986
(though the final result of merging is always the same).

> So, for example,
> when you say base-uri="file://three/four/a.file", do
> you really mean the [base URI] is "file://three/four/"?

The base URI is file://three/four/a.file, but any use of it to resolve
a relative URI will involve throwing away the "a.file" bit.  The first
step of the resolution algorithm throws away everything after the last
slash in the base URI.  So it has the same effect as if it were
file://three/four/ .

-- Richard
Received on Tuesday, 30 May 2006 17:19:07 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 8 January 2008 14:21:34 GMT