[
Lists Home |
Date Index |
Thread Index
]
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
/ "Simon St.Laurent" <simonstl@simonstl.com> was heard to say:
| 4) QNames in content seems to be an ever-expanding mess, with no means
| in sight for a normalization method that makes them context-independent.
| (That's largely because there's no simple mapping between QNames and
| URIs - hashes vs. slashes keeps that complicated.)
That's interesting. I usually put the "lexical form" problem before
the "URI/QName" mapping problem. Do we need to solve them both? (Yes,
I suppose we do.)
The URI/QName mapping problem is over-constrained in the general case.
If you want to start with a QName and end with a single string and you
want the operation to be reversible, you're really saying that you
want a structured string value. If you say that you want the resulting
string to be a valid URI, you're effectively saying that concatenation
with a separator is the only choice. And since there are no available
markup characters in URIs to use as a separator, you lose.
Unless you adopt a convention for the URIs that you use as namespace
names, I don't think there's any way to make the mapping reversible.
And no matter what convention you choose, someone will rebel against
it.
Be seeing you,
norm
- --
Norman.Walsh@Sun.COM | One of the great misfortunes of mankind is
XML Standards Architect | that even his good qualities are sometimes
Web Tech. and Standards | useless to him, and that the art of employing
Sun Microsystems, Inc. | and well directing them is often the latest
| fruit of his experience.--Chamfort
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.2 (GNU/Linux)
Comment: Processed by Mailcrypt 3.5.7 <http://mailcrypt.sourceforge.net/>
iD8DBQE/KRssOyltUcwYWjsRAq+ZAKChSazBS94+RNwxiH474lpN3Xe7wwCcC8cx
f8UHuBgxmAnSwRI613lBL+U=
=GFLt
-----END PGP SIGNATURE-----
|
