Thank you very much. But, as far as I understood, even though rfc3986 makes obsolete to rfc2396, xs:anyURI must conform to rfc2396. I could not find any refence that xs:anyURI must comply either with rfc3986 or rfc3987.
I did some search and found many people were in trouble with xs:anyURI because of ambiguity of rfc2396: http://www.stylusstudio.com/xmldev/200404/post30160.html I also found a technical note from OASIS titled "Handling of anyURI datatypes": http://www.oasis-open.org/committees/uddi-spec/doc/tn/uddi-spec-tc-tn-anyurihandling-20040921.htm#_ftn1. It has interesting footnote that says: --------------------------------------------- "Discussions within the OASIS UDDI TC have shown that different TC members had different interpretations of what the allowed character set for anyURI based XML elements and attributes is. Clarifications with W3C representatives have shown that the allowed character set is unambiguously defined, as discussed in this Technical Note." --------------------------------------------- What I have as a result: - Even though it is not expected, someone can send strange MessageIDs to my services and Axis2 allows those message to successfully pass into my services. - Since my services can be confronted with any string as MessageID, I will have no chance to know its size and format. Thus, there is no way to predict a *suitable* table field to store it. I'm really confused... Ali Sadik Kumlali --- "Flores, Raul" <[EMAIL PROTECTED]> wrote: > http://www.ietf.org/rfc/rfc3987.txt > > > Even though it may not conform, you still have to deal with it! > > Raul Flores > > > > > -----Original Message----- > From: Ali Sadik Kumlali [mailto:[EMAIL PROTECTED] > Sent: Thursday, April 13, 2006 9:31 AM > To: axis-user@ws.apache.org > Subject: Re: [Axis2] What values are allowed for <wsa:MessageID>? > > I've tested it. Axis2 happily sends and receives MessageID like "abc > 123 4as arw4343 -?;,[EMAIL PROTECTED] $". I was also able to send 1000 > characters > long MessageID without any problem. Does it conform to IRI standard? > > Thanks a lot. > > Ali Sadik Kumlali > > --- Ali Sadik Kumlali <[EMAIL PROTECTED]> wrote: > > > Thank you very much for the quick answer. > > > > Please see my inline comments. > > > > --- David Illsley <[EMAIL PROTECTED]> wrote: > > > > According to http://www.w3.org/2005/08/addressing > > > > schema, it can be any value. > > > > > > That's not entirely correct, the definition is broad but there > are > > > some requirements. The specification text requires [message id] > be > > > a valid absolute IRI while the schema requires an xs:anyURI > > You seem right. But, I was able to put any text in any size into > > xs:anyURI qualified field. You can test it with the following > schema > > and instance. > > > > ------------------------------------------------------------ > > anyURITest.xsd > > ------------------------------------------------------------ > > <?xml version="1.0" encoding="UTF-8"?> <xs:schema > > xmlns:xs="http://www.w3.org/2001/XMLSchema"; > > xmlns:wsa="http://www.w3.org/2005/08/addressing"; > > elementFormDefault="qualified" > attributeFormDefault="unqualified"> > > > > <xs:import > > namespace="http://www.w3.org/2005/08/addressing"; > > > > schemaLocation="http://www.w3.org/2005/08/addressing/ws-addr.xsd"/> > > > > <xs:element name="URITest"> > > <xs:complexType> > > <xs:sequence> > > <xs:element name="xsAnyURI" type="xs:anyURI"/> > > <xs:element name="wsaAttributedURIType" > > type="wsa:AttributedURIType"/> > > <xs:element ref="wsa:MessageID"/> > > </xs:sequence> > > </xs:complexType> > > </xs:element> > > > > </xs:schema> > > > > ------------------------------------------------------------ > > Xml instance which is valid against the anyURITest.xsd > > ------------------------------------------------------------ > > <?xml version="1.0" encoding="UTF-8"?> <URITest > > xmlns:wsa="http://www.w3.org/2005/08/addressing"; > > xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"; > > xsi:noNamespaceSchemaLocation="anyURITest.xsd"> > > <xsAnyURI>abc 123 4as arw4343 -?;,[EMAIL PROTECTED] $</xsAnyURI> > > <wsaAttributedURIType>abc 123 4as arw4343 -?;,[EMAIL PROTECTED] > > $</wsaAttributedURIType> > > <wsa:MessageID>abc 123 4as arw4343 -?;,[EMAIL PROTECTED] > > $</wsa:MessageID> > > </URITest> > > > > > > > <snip> > > > > > > > > 1) I need to write MessageID value to the database and > therefore > > > > need to know exact size and format of it. > > > > Is there a *sender neutral* way of this? > > > > > > No. All you can know is that a conformant message id is an IRI > > If I'm not missing something, according to the my previous comment, > > > IRI can be *anything*. > > > > > > > > > > 2) Although http://www.w3.org/2005/08/addressing schema does > not > > > > mandate for a special format, does Axis2 expect a special > pattern? > > > > > > The spec requires an IRI so if it did expect something I'd be > more > > > than a little surprised given the interop results. > > Is there any interop test that sets MessageID to the nonconformant > IRI > > > value? I mean, if someone has sent > > a value like "abc 123 4as arw4343 -?;,[EMAIL PROTECTED] $" ? > > (I'm going to try it with Axis2 :-) > > > > > > 3) Must I make sure that I can safely write MessageID without > > > > contradicting database schema constraints, if it has passed > > > > through Axis2? > > > > > > A quick look at the code suggests no :-( I can't see any > validation > > > of the MessageID e.g. that it is actually a valid URI/IRI > > > > > > Hope this helps, > > > David > > Thanks again. > > > > Ali Sadik Kumlali > > > > __________________________________________________ > > Do You Yahoo!? > > Tired of spam? Yahoo! Mail has the best spam protection around > > http://mail.yahoo.com > > > > > __________________________________________________ > Do You Yahoo!? > Tired of spam? Yahoo! Mail has the best spam protection around > http://mail.yahoo.com > __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com
