- From: Rogers, Tony <Tony.Rogers@ca.com>
- Date: Mon, 24 Jul 2006 18:45:31 +1000
- To: "Youenn Fablet" <youenn.fablet@crf.canon.fr>, <www-ws-desc@w3.org>
- Cc: "Arthur Ryman" <ryman@ca.ibm.com>
- Message-ID: <BEE2BD647C052D4FA59B42F5E2D946B317B5DA@AUSYMS12.ca.com>
Without referring to any authoritative source, I would have thought that the terminal slash was an error. (I fully expect to be beaten to death with a 600 page standard for saying that, of course)
Given that we now have examples in our test suite showing both versions (even if one is officially incorrect), I wonder if we should mandate a bit of canonicalisation of URIs (perhaps guaranteeing the presence or absence of a trailing slash, lowercase for the prefix, simple stuff like that). I can see this coming up over and over again.
Tony Rogers
CA, Inc
Senior Architect, Development
tony.rogers@ca.com
co-chair UDDI TC at OASIS
co-chair WS-Desc WG at W3C
________________________________
From: www-ws-desc-request@w3.org on behalf of Youenn Fablet
Sent: Mon 24-Jul-06 18:24
To: www-ws-desc@w3.org
Cc: Arthur Ryman
Subject: URI comparison
Hi all,
I dug into some of the WSDL documents for which our wsdl parser does not
match with the baseline.
Some differences are due to the use in some documents (Echo-1G for
instance) of the following binding URI:
wsoap:protocol="http://www.w3.org/2003/05/soap/bindings/HTTP"
Our parser recognizes the use of the SOAP1.2 HTTP binding with the
following URI:
wsoap:protocol="http://www.w3.org/2003/05/soap/bindings/HTTP/"
The sole difference is the last character '/'.
We currently use a simple character-by-character comparison to match the
URI against the SOAP1.2 binding URI, as defined IIRC somewhere in part 1.
Arthur, do you know how Woden is handling URI processing and comparison ?
Regards,
Youenn
Received on Monday, 24 July 2006 08:50:25 UTC