Loading ...
Sorry, an error occurred while loading the content.
 

multiple immediate children in doc/literal?

Expand Messages
  • Rich Salz
    Using doc/literal mode, does it make sense to have a soap message where the Body has more than one child node?
    Message 1 of 22 , Jul 30, 2002
      Using doc/literal mode, does it make sense to have a soap message where
      the Body has more than one child node?
      <soap-env:envelope>
      <soap-env:body>
      <tns:foo>...</tns:foo>
      <tns:bar>...</tns:bar>
      </soap-env:body>
      </soap-env:envelope>

      My gut feeling is that (a) I'm not sure this makes sense; and (b) it
      will cause interop problems, esp if tns:bar becomes
      some-other-namespace:bar.

      Comments?
      /r$
    • Matt Long
      Rich, I think it does make sense in the fact that two documents can be contained in the soap message and it is allowable under spec. I don t believe that the
      Message 2 of 22 , Jul 30, 2002
        Rich,

        I think it does make sense in the fact that two documents can be
        contained in the soap message and it is allowable under spec. I don't
        believe that the namespace issue is an issue. A validating parser (for
        literals) must understand all namespaces and their respective scopes.
        The prefixes are merely place holders for the resolving uri and must be
        resolved (validating parser or not).

        Thx,

        -Matt Long
        Phalanx Systems, LLC

        > -----Original Message-----
        > From: Rich Salz [mailto:r.salz@...]
        > Sent: Tuesday, July 30, 2002 10:11 AM
        > To: soapbuilders@yahoogroups.com
        > Subject: [soapbuilders] multiple immediate children in doc/literal?
        >
        > Using doc/literal mode, does it make sense to have a soap message
        where
        > the Body has more than one child node?
        > <soap-env:envelope>
        > <soap-env:body>
        > <tns:foo>...</tns:foo>
        > <tns:bar>...</tns:bar>
        > </soap-env:body>
        > </soap-env:envelope>
        >
        > My gut feeling is that (a) I'm not sure this makes sense; and (b) it
        > will cause interop problems, esp if tns:bar becomes
        > some-other-namespace:bar.
        >
        > Comments?
        > /r$
        >
        >
        > ------------------------ Yahoo! Groups Sponsor
        >
        > -----------------------------------------------------------------
        > This group is a forum for builders of SOAP implementations to discuss
        > implementation and interoperability issues. Please stay on-topic.
        >
        > To unsubscribe from this group, send an email to:
        > soapbuilders-unsubscribe@yahoogroups.com
        >
        >
        >
        > Your use of Yahoo! Groups is subject to
        http://docs.yahoo.com/info/terms/
        >
      • Rich Salz
        ... I was too terse. I meant how do the various SOAP stacks dispatch the message when top-level parts are in different namespaces? Or do they just look at
        Message 3 of 22 , Jul 30, 2002
          > I don't
          > believe that the namespace issue is an issue.

          I was too terse. I meant how do the various SOAP stacks dispatch the
          message when top-level parts are in different namespaces? Or do they
          just look at the first child and dispatch to its handler?
          /r$
        • Wes Moulder
          My guess would be either that, or decomposing it into two messages, a la boxcarring. Theoretically, it should match up against the message in the wsdl, and
          Message 4 of 22 , Jul 30, 2002

            My guess would be either that, or decomposing it into two messages, a la boxcarring.  Theoretically, it should match up against the message in the wsdl, and dispatch to the appropriate message that contains both parts though.

             

            --Wes

             

            -----Original Message-----
            From: Rich Salz [mailto:r.salz@...]
            Sent: Tuesday, July 30, 2002 12:27 PM
            To: Matt Long
            Cc: soapbuilders@yahoogroups.com
            Subject: Re: [soapbuilders] multiple immediate children in doc/literal?

             

            > I don't
            > believe that the namespace issue is an issue.

            I was too terse. I  meant how do the various SOAP stacks dispatch the
            message when top-level parts are in different namespaces?  Or do they
            just look at the first child and dispatch to its handler?
                  /r$



            -----------------------------------------------------------------
            This group is a forum for builders of SOAP implementations to discuss implementation and interoperability issues.  Please stay on-topic.

            To unsubscribe from this group, send an email to:
            soapbuilders-unsubscribe@yahoogroups.com



            Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service.
          • Bob Cunnings
            FWIW the toolkit here dispatches on the basis of the request URI, and the registered handler receives the entire contents of the body, which it may process as
            Message 5 of 22 , Jul 30, 2002
              FWIW the toolkit here dispatches on the basis of the request URI, and the
              registered handler receives the entire contents of the body, which it may
              process as it pleases. The quantity and namespaces of body children aren't
              relevant to the dispatch mechanism.

              RC

              > > I don't
              > > believe that the namespace issue is an issue.
              >
              > I was too terse. I meant how do the various SOAP stacks dispatch the
              > message when top-level parts are in different namespaces? Or do they
              > just look at the first child and dispatch to its handler?
              > /r$
              >
              >
            • Matt Long
              Hi Rich, Most use the first element name and/or + namespace much like what the tests where at Interop III F2F where configured to represent...of course I was
              Message 6 of 22 , Jul 30, 2002
                Hi Rich,

                Most use the first element name and/or + namespace much like what the
                tests where at Interop III F2F where configured to represent...of course
                I was having F2Tidy-Bowl man and was not able to address some of these
                issues on day 2.

                I suppose that method, i.e. first element, has been the natural
                progression especially since depreciation of SOAPAction in v1.2. The
                downside is that you cannot have a repeating first element name in the
                WSDL or import it. Some may use first element + namespace, some don't.

                Of course != SOAPAction across doc/lit messages also works, but it seems
                to be out-of-favor due to v1.2.


                Thx,

                -Matt Long
                Phalanx Systems, LLC



                > -----Original Message-----
                > From: Rich Salz [mailto:r.salz@...]
                > Sent: Tuesday, July 30, 2002 12:27 PM
                > To: Matt Long
                > Cc: soapbuilders@yahoogroups.com
                > Subject: Re: [soapbuilders] multiple immediate children in
                doc/literal?
                >
                > > I don't
                > > believe that the namespace issue is an issue.
                >
                > I was too terse. I meant how do the various SOAP stacks dispatch the
                > message when top-level parts are in different namespaces? Or do they
                > just look at the first child and dispatch to its handler?
                > /r$
                >
                >
                > ------------------------ Yahoo! Groups Sponsor
                >
                > -----------------------------------------------------------------
                > This group is a forum for builders of SOAP implementations to discuss
                > implementation and interoperability issues. Please stay on-topic.
                >
                > To unsubscribe from this group, send an email to:
                > soapbuilders-unsubscribe@yahoogroups.com
                >
                >
                >
                > Your use of Yahoo! Groups is subject to
                http://docs.yahoo.com/info/terms/
                >
              • Pete Hendry
                On the subject of using a validating parser, how can the parser know that the body should contain 2 elements, or what elements they should be? What I mean is
                Message 7 of 22 , Jul 30, 2002
                  On the subject of using a validating parser, how can the parser know that the body should contain 2 elements, or what elements they should be? What I mean is you can use a validating parser to validate that *each* element is valid but the parser can't tell what elements were supposed to be in the message in the first place. I find this a flaw in SOAP where a message cannot be validated as the correct message using a validating parser. However, I can't see how you could define a generic content holder like Body and solve this problem. My only suggestion would be that the Body could be defined as a restriction of env:Body defined in your target schema and defining the exact content of the body. However, this would not be allowed by the spec and would complicate things somewhat (although I would love to be able to fire the message at a validating parser and know it was valid).

                  Pete


                  Matt Long wrote:
                  Rich,

                  I think it does make sense in the fact that two documents can be
                  contained in the soap message and it is allowable under spec. I don't
                  believe that the namespace issue is an issue. A validating parser (for
                  literals) must understand all namespaces and their respective scopes.
                  The prefixes are merely place holders for the resolving uri and must be
                  resolved (validating parser or not).

                  Thx,

                  -Matt Long
                  Phalanx Systems, LLC

                  -----Original Message-----
                  From: Rich Salz [mailto:r.salz@...]
                  Sent: Tuesday, July 30, 2002 10:11 AM
                  To: soapbuilders@yahoogroups.com
                  Subject: [soapbuilders] multiple immediate children in doc/literal?

                  Using doc/literal mode, does it make sense to have a soap message
                  where
                  the Body has more than one child node?
                  <soap-env:envelope>
                  <soap-env:body>
                  <tns:foo>...</tns:foo>
                  <tns:bar>...</tns:bar>
                  </soap-env:body>
                  </soap-env:envelope>

                  My gut feeling is that (a) I'm not sure this makes sense; and (b) it
                  will cause interop problems, esp if tns:bar becomes
                  some-other-namespace:bar.

                  Comments?
                  /r$


                  ------------------------ Yahoo! Groups Sponsor

                  -----------------------------------------------------------------
                  This group is a forum for builders of SOAP implementations to discuss
                  implementation and interoperability issues. Please stay on-topic.

                  To unsubscribe from this group, send an email to:
                  soapbuilders-unsubscribe@yahoogroups.com



                  Your use of Yahoo! Groups is subject to
                  http://docs.yahoo.com/info/terms/



                  ------------------------ Yahoo! Groups Sponsor ---------------------~-->
                  Free $5 Love Reading
                  Risk Free!
                  http://us.click.yahoo.com/NsdPZD/PfREAA/Ey.GAA/W6uqlB/TM
                  ---------------------------------------------------------------------~->

                  -----------------------------------------------------------------
                  This group is a forum for builders of SOAP implementations to discuss implementation and interoperability issues. Please stay on-topic.

                  To unsubscribe from this group, send an email to:
                  soapbuilders-unsubscribe@yahoogroups.com



                  Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/



                • Matt Long
                  Pete, If the message is literal (rpc | document), each message part can be validated against the schema in WSDL. Likewise, literal headers can be validated
                  Message 8 of 22 , Jul 31, 2002

                    Pete,

                     

                    If the message is literal (rpc | document), each message part can be validated against the schema in WSDL.  Likewise, literal headers can be validated also.

                     

                    ..how can the parser know that the body should contain 2 elements”

                     

                    The wsdl:message will contain 2 wsdl:parts, _each_ part (not element) is then validated independent of the other (just like a literal header).  The part can be validated against the schema (after a check of the required namespaces).

                     

                     

                    Thx,

                     

                    -Matt Long

                    Phalanx Systems, LLC

                     

                     

                    -----Original Message-----
                    From: Pete Hendry [mailto:peter.hendry@...]
                    Sent: Wednesday, July 31, 2002 1:41 AM
                    To: soapbuilders@yahoogroups.com
                    Subject: Re: [soapbuilders] multiple immediate children in doc/literal?

                     

                    On the subject of using a validating parser, how can the parser know that the body should contain 2 elements, or what elements they should be? What I mean is you can use a validating parser to validate that *each* element is valid but the parser can't tell what elements were supposed to be in the message in the first place. I find this a flaw in SOAP where a message cannot be validated as the correct message using a validating parser. However, I can't see how you could define a generic content holder like Body and solve this problem. My only suggestion would be that the Body could be defined as a restriction of env:Body defined in your target schema and defining the exact content of the body. However, this would not be allowed by the spec and would complicate things somewhat (although I would love to be able to fire the message at a validating parser and know it was valid).

                    Pete


                    Matt Long wrote:

                    Rich,

                    I think it does make sense in the fact that two documents can be
                    contained in the soap message and it is allowable under spec.  I don't
                    believe that the namespace issue is an issue.  A validating parser (for
                    literals) must understand all namespaces and their respective scopes.
                    The prefixes are merely place holders for the resolving uri and must be
                    resolved (validating parser or not).

                    Thx,

                    -Matt Long
                    Phalanx Systems, LLC
                    -----Original Message-----
                    From: Rich Salz [mailto:r.salz@...]
                    Sent: Tuesday, July 30, 2002 10:11 AM
                    To: soapbuilders@yahoogroups.com
                    Subject: [soapbuilders] multiple immediate children in doc/literal?

                    Using doc/literal mode, does it make sense to have a soap message
                    where
                    the Body has more than one child node?
                    <soap-env:envelope>
                      <soap-env:body>
                        <tns:foo>...</tns:foo>
                        <tns:bar>...</tns:bar>
                      </soap-env:body>
                    </soap-env:envelope>

                    My gut feeling is that (a) I'm not sure this makes sense; and (b) it
                    will cause interop problems, esp if tns:bar becomes
                    some-other-namespace:bar.

                    Comments?
                    /r$


                    ------------------------ Yahoo! Groups Sponsor

                    -----------------------------------------------------------------
                    This group is a forum for builders of SOAP implementations to discuss
                    implementation and interoperability issues.  Please stay on-topic.

                    To unsubscribe from this group, send an email to:
                    soapbuilders-unsubscribe@yahoogroups.com



                    Your use of Yahoo! Groups is subject to
                    http://docs.yahoo.com/info/terms/



                    ------------------------ Yahoo! Groups Sponsor ---------------------~-->
                    Free $5 Love Reading
                    Risk Free!
                    http://us.click.yahoo.com/NsdPZD/PfREAA/Ey.GAA/W6uqlB/TM
                    ---------------------------------------------------------------------~->

                    -----------------------------------------------------------------
                    This group is a forum for builders of SOAP implementations to discuss implementation and interoperability issues.  Please stay on-topic.

                    To unsubscribe from this group, send an email to:
                    soapbuilders-unsubscribe@yahoogroups.com

                     

                    Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/




                    -----------------------------------------------------------------
                    This group is a forum for builders of SOAP implementations to discuss implementation and interoperability issues.  Please stay on-topic.

                    To unsubscribe from this group, send an email to:
                    soapbuilders-unsubscribe@yahoogroups.com



                    Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service.
                  • Pete Hendry
                    That s my point. Each body element can be validated as the parser can lookup the element name in the schema. But, the parser can t know what elements are
                    Message 9 of 22 , Jul 31, 2002
                      That's my point. Each body element can be validated as the parser can lookup the element name in the schema. But, the parser can't know what elements are supposed to be in the body and whether the elements that are there are valid against any valid message format.

                      Pete

                      Matt Long wrote:

                      Pete,

                       

                      If the message is literal (rpc | document), each message part can be validated against the schema in WSDL.   Likewise, literal headers can be validated also.

                       

                      .. how can the parser know that the body should contain 2 elements”

                       

                      The wsdl:message will contain 2 wsdl:parts, _ each_ part (not element) is then validated independent of the other (just like a literal header).  The part can be validated against the schema (after a check of the required namespaces).

                       

                       

                      Thx,

                       

                      -Matt Long

                      Phalanx Systems, LLC

                       

                       

                      -----Original Message-----
                      From: Pete Hendry [mailto:peter.hendry@...]
                      Sent: Wednesday, July 31, 2002 1:41 AM
                      To: soapbuilders@yahoogroups.com
                      Subject: Re: [soapbuilders] multiple immediate children in doc/literal?

                       

                      On the subject of using a validating parser, how can the parser know that the body should contain 2 elements, or what elements they should be? What I mean is you can use a validating parser to validate that *each* element is valid but the parser can't tell what elements were supposed to be in the message in the first place. I find this a flaw in SOAP where a message cannot be validated as the correct message using a validating parser. However, I can't see how you could define a generic content holder like Body and solve this problem. My only suggestion would be that the Body could be defined as a restriction of env:Body defined in your target schema and defining the exact content of the body. However, this would not be allowed by the spec and would complicate things somewhat (although I would love to be able to fire the message at a validating parser and know it was valid).

                      Pete


                      Matt Long wrote:

                      Rich,



                      I think it does make sense in the fact that two documents can be

                      contained in the soap message and it is allowable under spec.  I don't

                      believe that the namespace issue is an issue.  A validating parser (for

                      literals) must understand all namespaces and their respective scopes.

                      The prefixes are merely place holders for the resolving uri and must be

                      resolved (validating parser or not).



                      Thx,



                      -Matt Long

                      Phalanx Systems, LLC
                      -----Original Message-----

                      From: Rich Salz [mailto:r.salz@...]

                      Sent: Tuesday, July 30, 2002 10:11 AM

                      To: soapbuilders@yahoogroups.com

                      Subject: [soapbuilders] multiple immediate children in doc/literal?



                      Using doc/literal mode, does it make sense to have a soap message
                      where
                      the Body has more than one child node?

                      <soap-env:envelope>

                        <soap-env:body>

                          <tns:foo>...</tns:foo>

                          <tns:bar>...</tns:bar>

                        </soap-env:body>

                      </soap-env:envelope>



                      My gut feeling is that (a) I'm not sure this makes sense; and (b) it

                      will cause interop problems, esp if tns:bar becomes

                      some-other-namespace:bar.



                      Comments?

                      /r$





                      ------------------------ Yahoo! Groups Sponsor



                      -----------------------------------------------------------------

                      This group is a forum for builders of SOAP implementations to discuss

                      implementation and interoperability issues.  Please stay on-topic.



                      To unsubscribe from this group, send an email to:

                      soapbuilders-unsubscribe@yahoogroups.com







                      Your use of Yahoo! Groups is subject to
                      http://docs.yahoo.com/info/terms/






                      ------------------------ Yahoo! Groups Sponsor ---------------------~-->

                      Free $5 Love Reading

                      Risk Free!

                      http://us.click.yahoo.com/NsdPZD/PfREAA/Ey.GAA/W6uqlB/TM

                      ---------------------------------------------------------------------~->



                      -----------------------------------------------------------------

                      This group is a forum for builders of SOAP implementations to discuss implementation and interoperability issues.  Please stay on-topic.



                      To unsubscribe from this group, send an email to:

                      soapbuilders-unsubscribe@yahoogroups.com



                       



                      Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/






                      -----------------------------------------------------------------
                      This group is a forum for builders of SOAP implementations to discuss implementation and interoperability issues.  Please stay on-topic.

                      To unsubscribe from this group, send an email to:
                      soapbuilders-unsubscribe@yahoogroups.com



                      Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service .



                      -----------------------------------------------------------------
                      This group is a forum for builders of SOAP implementations to discuss implementation and interoperability issues.  Please stay on-topic.

                      To unsubscribe from this group, send an email to:
                      soapbuilders-unsubscribe@yahoogroups.com



                      Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service .

                    • Rich Salz
                      ... By validating I assume you mean XSD schema-valid, not basic XML valid. ... Can t you determine everything from the WSDL? /r$
                      Message 10 of 22 , Jul 31, 2002
                        > On the subject of using a validating parser, how can the parser know
                        > that the body should contain 2 elements, or what elements they should
                        > be? What I mean is you can use a validating parser to validate that
                        > *each* element is valid but the parser can't tell what elements were
                        > supposed to be in the message in the first place.

                        By validating I assume you mean XSD schema-valid, not basic XML valid.

                        > I find this a flaw in
                        > SOAP where a message cannot be validated as the correct message using a
                        > validating parser.

                        Can't you determine everything from the WSDL?
                        /r$
                      • Matt Long
                        Pete, I must be missing something. If the wsdl:message contains wsdl:part (s) then the message parts are known. If some of those wsdl:part are omitted, then
                        Message 11 of 22 , Jul 31, 2002

                          Pete,

                           

                          I must be missing something.  If the wsdl:message contains wsdl:part (s) then the message parts are known.  If some of those wsdl:part are omitted, then the message is invalid.  What am I missing here?

                           

                          Thx,

                           

                          -Matt Long

                          Phalanx Systems, LLC

                           

                           

                          That's my point. Each body element can be validated as the parser can lookup the element name in the schema. But, the parser can't know what elements are supposed to be in the body and whether the elements that are there are valid against any valid message format.

                          Pete

                          Matt Long wrote:

                          Pete,

                           

                          If the message is literal (rpc | document), each message part can be validated against the schema in WSDL.   Likewise, literal headers can be validated also.

                           

                          “.. how can the parser know that the body should contain 2 elements”

                           

                          The wsdl:message will contain 2 wsdl:parts, _ each_ part (not element) is then validated independent of the other (just like a literal header).  The part can be validated against the schema (after a check of the required namespaces).

                           

                           

                          Thx,

                           

                          -Matt Long

                          Phalanx Systems, LLC

                           

                           

                          -----Original Message-----
                          From: Pete Hendry [mailto:peter.hendry@...]
                          Sent: Wednesday, July 31, 2002 1:41 AM
                          To: soapbuilders@yahoogroups.com
                          Subject: Re: [soapbuilders] multiple immediate children in doc/literal?

                           

                          On the subject of using a validating parser, how can the parser know that the body should contain 2 elements, or what elements they should be? What I mean is you can use a validating parser to validate that *each* element is valid but the parser can't tell what elements were supposed to be in the message in the first place. I find this a flaw in SOAP where a message cannot be validated as the correct message using a validating parser. However, I can't see how you could define a generic content holder like Body and solve this problem. My only suggestion would be that the Body could be defined as a restriction of env:Body defined in your target schema and defining the exact content of the body. However, this would not be allowed by the spec and would complicate things somewhat (although I would love to be able to fire the message at a validating parser and know it was valid).

                          Pete


                          Matt Long wrote:

                          Rich,



                          I think it does make sense in the fact that two documents can be

                          contained in the soap message and it is allowable under spec.  I don't

                          believe that the namespace issue is an issue.  A validating parser (for

                          literals) must understand all namespaces and their respective scopes.

                          The prefixes are merely place holders for the resolving uri and must be

                          resolved (validating parser or not).



                          Thx,



                          -Matt Long

                          Phalanx Systems, LLC
                          -----Original Message-----

                          From: Rich Salz [mailto:r.salz@...]

                          Sent: Tuesday, July 30, 2002 10:11 AM

                          To: soapbuilders@yahoogroups.com

                          Subject: [soapbuilders] multiple immediate children in doc/literal?



                          Using doc/literal mode, does it make sense to have a soap message
                          where
                          the Body has more than one child node?

                                  <soap-env:envelope>

                                    <soap-env:body>

                                      <tns:foo>...</tns:foo>

                                      <tns:bar>...</tns:bar>

                                    </soap-env:body>

                                  </soap-env:envelope>



                          My gut feeling is that (a) I'm not sure this makes sense; and (b) it

                          will cause interop problems, esp if tns:bar becomes

                          some-other-namespace:bar.



                          Comments?

                           /r$





                          ------------------------ Yahoo! Groups Sponsor



                          -----------------------------------------------------------------

                          This group is a forum for builders of SOAP implementations to discuss

                          implementation and interoperability issues.  Please stay on-topic.



                          To unsubscribe from this group, send an email to:

                          soapbuilders-unsubscribe@yahoogroups.com







                          Your use of Yahoo! Groups is subject to
                          http://docs.yahoo.com/info/terms/






                          ------------------------ Yahoo! Groups Sponsor ---------------------~-->

                          Free $5 Love Reading

                          Risk Free!

                          http://us.click.yahoo.com/NsdPZD/PfREAA/Ey.GAA/W6uqlB/TM

                          ---------------------------------------------------------------------~->



                          -----------------------------------------------------------------

                          This group is a forum for builders of SOAP implementations to discuss implementation and interoperability issues.  Please stay on-topic.



                          To unsubscribe from this group, send an email to:

                          soapbuilders-unsubscribe@yahoogroups.com



                           




                          Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/





                          -----------------------------------------------------------------
                          This group is a forum for builders of SOAP implementations to discuss implementation and interoperability issues.  Please stay on-topic.

                          To unsubscribe from this group, send an email to:
                          soapbuilders-unsubscribe@yahoogroups.com



                          Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service .

                           



                          -----------------------------------------------------------------
                          This group is a forum for builders of SOAP implementations to discuss implementation and interoperability issues.  Please stay on-topic.

                          To unsubscribe from this group, send an email to:
                          soapbuilders-unsubscribe@yahoogroups.com



                          Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service .

                           

                        • Matt Long
                          Here is my spin. 1) The validating parser only validates a literal message part against the schema in WSDL. 2) The message is valid only is the wsdl:part (s)
                          Message 12 of 22 , Jul 31, 2002
                            Here is my spin.

                            1) The validating parser only validates a literal message part against
                            the schema in WSDL.
                            2) The message is valid only is the wsdl:part (s) in Body are present.

                            If a message contains 'n' parts in Body, then each part must correspond
                            to a wsdl:part in then message, else invalid message. Each part in the
                            message can be validated against the schema in the WSDL.

                            Make sense?

                            Thx,

                            -Matt Long
                            Phalanx Systems, LLC



                            > -----Original Message-----
                            > From: Rich Salz [mailto:r.salz@...]
                            > Sent: Wednesday, July 31, 2002 9:10 AM
                            > To: Pete Hendry
                            > Cc: soapbuilders@yahoogroups.com
                            > Subject: Re: [soapbuilders] multiple immediate children in
                            doc/literal?
                            >
                            > > On the subject of using a validating parser, how can the parser know
                            > > that the body should contain 2 elements, or what elements they
                            should
                            > > be? What I mean is you can use a validating parser to validate that
                            > > *each* element is valid but the parser can't tell what elements were
                            > > supposed to be in the message in the first place.
                            >
                            > By validating I assume you mean XSD schema-valid, not basic XML valid.
                            >
                            > > I find this a flaw in
                            > > SOAP where a message cannot be validated as the correct message
                            using a
                            > > validating parser.
                            >
                            > Can't you determine everything from the WSDL?
                            > /r$
                            >
                            >
                            > ------------------------ Yahoo! Groups Sponsor
                            >
                            > -----------------------------------------------------------------
                            > This group is a forum for builders of SOAP implementations to discuss
                            > implementation and interoperability issues. Please stay on-topic.
                            >
                            > To unsubscribe from this group, send an email to:
                            > soapbuilders-unsubscribe@yahoogroups.com
                            >
                            >
                            >
                            > Your use of Yahoo! Groups is subject to
                            http://docs.yahoo.com/info/terms/
                            >
                          • Pete Hendry
                            ... Yes. ... You can by reading the WSDL and writing your own validation. However, I think if you are using a standard based on XML then a basic feature should
                            Message 13 of 22 , Jul 31, 2002
                              Rich Salz wrote:

                              >By validating I assume you mean XSD schema-valid, not basic XML valid.
                              >

                              Yes.

                              >Can't you determine everything from the WSDL?
                              >

                              You can by reading the WSDL and writing your own validation. However, I
                              think if you are using a standard based on XML then a basic feature
                              should be that you can validate it using an out-of-the-box XML parser.
                              But an XML parser can't read WSDL. It can find the schema using XPath or
                              whatever (another thing about the way things are at the moment is that
                              the schemas are not tagged with for example an ID so they are easy to
                              reference for a parser) but there is other information in the WSDL they
                              would need to validate the message. I think this breaks with part of the
                              purpose of using XML in the first place - schema + validating parser.

                              Pete
                            • Rich Salz
                              You already need out of band information since PI s aren t legal in SOAP, so you need a way to link the schema to the message. Basically you re saying that XML
                              Message 14 of 22 , Jul 31, 2002
                                You already need out of band information since PI's aren't legal in
                                SOAP, so you need a way to link the schema to the message.

                                Basically you're saying that XML Schema isn't rich enough to be able to
                                validate all SOAP messages.

                                I have no problem with that. :)
                                /r$
                              • noah_mendelsohn@us.ibm.com
                                FWIW, SOAP 1.2 clarifies the implications of multiple children of body [1]: An ultimate SOAP receiver MUST correctly process the immediate children of the
                                Message 15 of 22 , Jul 31, 2002
                                  FWIW, SOAP 1.2 clarifies the implications of multiple children of body
                                  [1]:

                                  "An ultimate SOAP receiver MUST correctly process the immediate children of
                                  the SOAP body (see 5.3 SOAP Body). However, with the exception of SOAP
                                  faults (see 5.4 SOAP Fault), Part 1 of this specification (this document)
                                  mandates no particular structure or interpretation of these elements, and
                                  provides no standard means for specifying the processing to be done."

                                  Note it says children, not child. Note that it (intentionally) does not
                                  say anything about the order in which processing occurs. Note that it
                                  specifically declines to impose a "structure or interpretation" that's
                                  standard across SOAP. So, in particular:

                                  * Multiple elements may be need not represent multiple units of work
                                  (boxcarring). Alternatively, one or another element could be data to be
                                  used by the other(s)
                                  * If they do represent multiple units of work, the order of processing is
                                  not determined (unless some header targeted to the endpoint determines an
                                  order)
                                  * If there is a single unit of work + data, nothing says whether the
                                  first, middle, or last is actually the operation to be performed.

                                  In short: the ultimate receiver must understand the implications of all
                                  the elements in the <body> taken together. How it understands that and
                                  whether the model for one endpoint is the same or different as for others
                                  is beyond the scope of the SOAP 1.2 specification. The pros and cons of
                                  this choice were debated at great length. Note that it would be trivial
                                  to build a header like:

                                  <env:Soap>
                                  <env:Header>
                                  <someNs:boxcarBodies mustUnderstand="true">
                                  <someNs:bodyProcessingOrder>
                                  firstToLast
                                  </someNs:bodyProcessingOrder>
                                  </someNs:boxcarBodies mustUnderstand>
                                  </env:Header>
                                  <env:Body>
                                  <n:b1>....</n:b1>
                                  <n2:b2>....</n2:b2>
                                  </env:Body>
                                  </env:Soap>

                                  Or some such. The point is: you make SOAP as generic as possible, and
                                  use its own extension mechanisms to tailor it. If some of these headers
                                  become so widely agreed-upon that they are nearly universally understood,
                                  that's terrific. Of course, the specification for boxcarBodies must
                                  describe the exact semantics impossed on the bodies, when to fault, etc.
                                  Other conventions can be standardized using other headers.

                                  Lacking such a header, if the endpoint sees a body that it is not
                                  completely prepared to process (I.e. is confused about why there might be
                                  multiple children), it MUST fault. It is NOT OK to just randomly process
                                  them in order, or process the first one only, in some vague hope that it
                                  might be the right thing to do. We're trying to make this all industrial
                                  strength. The spec mandates that if you aren't truly sure what to do with
                                  the body, you fault.

                                  [1]
                                  http://www.w3.org/2000/xp/Group/2/06/LC/soap12-part1.html#structinterpbodies

                                  ------------------------------------------------------------------
                                  Noah Mendelsohn Voice: 1-617-693-4036
                                  IBM Corporation Fax: 1-617-693-8676
                                  One Rogers Street
                                  Cambridge, MA 02142
                                  ------------------------------------------------------------------
                                • Rosimildo daSIlva
                                  ... I read this way: SOAP 1.2 is a beast with 100+ page spec, and the only thing it defines is two elements, Header and Body. The content of these
                                  Message 16 of 22 , Jul 31, 2002
                                    --- noah_mendelsohn@... wrote:
                                    > FWIW, SOAP 1.2 clarifies the implications of
                                    > multiple children of body
                                    > [1]:
                                    >
                                    > "An ultimate SOAP receiver MUST correctly process
                                    > the immediate children of
                                    > the SOAP body (see 5.3 SOAP Body). However, with the
                                    > exception of SOAP
                                    > faults (see 5.4 SOAP Fault), Part 1 of this
                                    > specification (this document)
                                    > mandates no particular structure or interpretation
                                    > of these elements, and
                                    > provides no standard means for specifying the
                                    > processing to be done."
                                    >
                                    > Note it says children, not child. Note that it
                                    > (intentionally) does not
                                    > say anything about the order in which processing
                                    > occurs. Note that it
                                    > specifically declines to impose a "structure or
                                    > interpretation" that's
                                    > standard across SOAP.

                                    <rant on>

                                    I read this way:

                                    SOAP 1.2 is a beast with 100+ page spec, and
                                    the only thing it defines is two elements,
                                    Header and Body. The content of these elements
                                    are "vaguely" defined, and they can contain
                                    just about anything that you can do with XML.

                                    One could question: What is the value of SOAP
                                    at all ? Why not just transport XML documents
                                    over any prococol as SOAP does ?

                                    After doing SOAP, along with your guys here,
                                    I have this feeling that I understand it less
                                    each day.

                                    </rant on>











                                    __________________________________________________
                                    Do You Yahoo!?
                                    Yahoo! Health - Feel better, live better
                                    http://health.yahoo.com
                                  • Rich Salz
                                    SOAP 1.1 hit the 80/20 rule. SOAP 1.2 was taken over by comp sci theorists and unfortunately it s now 20/80. SOAP 1.1 could be tricky because big things
                                    Message 17 of 22 , Jul 31, 2002
                                      SOAP 1.1 hit the 80/20 rule.
                                      SOAP 1.2 was taken over by comp sci theorists and unfortunately it's now
                                      20/80. SOAP 1.1 could be tricky because big things lurked in terse
                                      corners of its language. SOAP 1.2 is tricky because you can't stay away
                                      while reading it. Infoset allowing alternate serializations -- for a wire
                                      protocol!? Bah.
                                      /r$
                                    • graham glass
                                      hi rosimildo, if you think the SOAP 1.2 specification is large, you should check out the 500 page UDDI 3.0 specification! cheers, graham ... From: Rosimildo
                                      Message 18 of 22 , Jul 31, 2002
                                        hi rosimildo,
                                         
                                        if you think the SOAP 1.2 specification is large, you should check out
                                        the 500 page UDDI 3.0 specification!
                                         
                                        cheers,
                                        graham
                                        -----Original Message-----
                                        From: Rosimildo daSIlva [mailto:rosimildo@...]
                                        Sent: Wednesday, July 31, 2002 3:33 PM
                                        To: soapbuilders@yahoogroups.com
                                        Subject: Re: [soapbuilders] multiple immediate children in doc/literal?

                                        --- noah_mendelsohn@... wrote:
                                        > FWIW, SOAP 1.2 clarifies the implications of
                                        > multiple children of body
                                        > [1]:
                                        >
                                        > "An ultimate SOAP receiver MUST correctly process
                                        > the immediate children of
                                        > the SOAP body (see 5.3 SOAP Body). However, with the
                                        > exception of SOAP
                                        > faults (see 5.4 SOAP Fault), Part 1 of this
                                        > specification (this document)
                                        > mandates no particular structure or interpretation
                                        > of these elements, and
                                        > provides no standard means for specifying the
                                        > processing to be done."
                                        >
                                        > Note it says children, not child.  Note that it
                                        > (intentionally) does not
                                        > say anything about the order in which processing
                                        > occurs.  Note that it
                                        > specifically declines to impose a "structure or
                                        > interpretation" that's
                                        > standard across SOAP.

                                        <rant on>

                                        I read this way:

                                        SOAP 1.2 is a beast with 100+ page spec, and
                                        the only thing it defines is two elements,
                                        Header and Body. The content of these elements
                                        are "vaguely" defined, and they can contain
                                        just about anything that you can do with XML.

                                        One could question: What is the value of SOAP
                                        at all ? Why not just transport XML documents
                                        over any prococol as SOAP does ?

                                        After doing SOAP, along with your guys here,
                                        I have this feeling that I understand it less
                                        each day.

                                        </rant on>











                                        __________________________________________________
                                        Do You Yahoo!?
                                        Yahoo! Health - Feel better, live better
                                        http://health.yahoo.com


                                        -----------------------------------------------------------------
                                        This group is a forum for builders of SOAP implementations to discuss implementation and interoperability issues.  Please stay on-topic.

                                        To unsubscribe from this group, send an email to:
                                        soapbuilders-unsubscribe@yahoogroups.com



                                        Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service.
                                      • Rosimildo daSIlva
                                        ... Outch ! That hurts. Sorry everybody, for my fustrastion. It has been a long day ! Rosimildo. __________________________________________________ Do You
                                        Message 19 of 22 , Jul 31, 2002
                                          --- graham glass <graham@...> wrote:
                                          > hi rosimildo,
                                          >
                                          > if you think the SOAP 1.2 specification is large,
                                          > you should check out
                                          > the 500 page UDDI 3.0 specification!
                                          >

                                          Outch ! That hurts.

                                          Sorry everybody, for my fustrastion. It has been
                                          a long day !

                                          Rosimildo.

                                          __________________________________________________
                                          Do You Yahoo!?
                                          Yahoo! Health - Feel better, live better
                                          http://health.yahoo.com
                                        • Matt Long
                                          ... I have to disagree with you both; you re both right! Thx, -Matt Long Phalanx Systems, LLC
                                          Message 20 of 22 , Aug 3, 2002
                                            Robert van Engelen wrote:

                                            > I disagree with Rich' opinion that comp. sci. theorists are to blame
                                            > (no offense, Rich ;-) ). I suspect that too many parties attempt to be
                                            > stakeholders in SOAP 1.2 to tweak it to be as widely applicable as
                                            > possible in order to claim IP rights for anything that uses XML over
                                            > HTTP/SMTP/TCP (I am not even sure anymore whether SOAP 1.2. requires
                                            > XML at all which used to be main advantage that SOAP had over other
                                            > protocols by utilizing XML as the lingua franca for RPC).

                                            I have to disagree with you both; you're both right!


                                            Thx,

                                            -Matt Long
                                            Phalanx Systems, LLC
                                          • Robert van Engelen
                                            ... Here is my two cents worth of unsollicited opinion (read: rant). It is certainly ironic that the bloat has crept into protocols such as UDDI 3.0 and SOAP
                                            Message 21 of 22 , Aug 3, 2002
                                              graham glass wrote:
                                              >
                                              > hi rosimildo,
                                              >
                                              > if you think the SOAP 1.2 specification is large, you should check out
                                              > the 500 page UDDI 3.0 specification!

                                              Here is my two cents worth of unsollicited opinion (read: rant).

                                              It is certainly ironic that the bloat has crept into protocols such
                                              as UDDI 3.0 and SOAP 1.2, while the original goal of these protocols
                                              was "lightweight" RPC and/or message exchange with lookup/discovery.

                                              On the one hand the usage context of SOAP 1.2 is greatly enlarged,
                                              while on the other hand SOAP 1.2 RPC's Section 5 encoding has been
                                              simplified in such a way that it breaks compatibility with SOAP 1.1 RPC
                                              encoding. I get the feeling that RPC encoding will be on its way out.
                                              As a result, SOAP's usefulness for language/platform interoperability
                                              will be diminished.

                                              Certainly, SOAP 1.1 promised to be a viable alternative to ORB systems
                                              because of its lightweight properties that allowed small-scale systems
                                              such as cellphones and embedded systems to use it rather than existing
                                              heavyweight ORB architectures. I am not sure if SOAP 1.2 will be able
                                              to fulfill this promise.

                                              I disagree with Rich' opinion that comp. sci. theorists are to blame
                                              (no offense, Rich ;-) ). I suspect that too many parties attempt to be
                                              stakeholders in SOAP 1.2 to tweak it to be as widely applicable as
                                              possible in order to claim IP rights for anything that uses XML over
                                              HTTP/SMTP/TCP (I am not even sure anymore whether SOAP 1.2. requires
                                              XML at all which used to be main advantage that SOAP had over other
                                              protocols by utilizing XML as the lingua franca for RPC).

                                              - Robert van Engelen, Computer Science Dept., FSU.
                                            • Rich Salz
                                              ... Hey, did you see that UDDI just turned spec development over to OASIS? That s interesting.
                                              Message 22 of 22 , Aug 6, 2002
                                                > It is certainly ironic that the bloat has crept into protocols such
                                                > as UDDI 3.0 and SOAP 1.2, while the original goal of these protocols
                                                > was "lightweight" RPC and/or message exchange with lookup/discovery.

                                                Hey, did you see that UDDI just turned spec development over to OASIS?

                                                That's interesting.
                                              Your message has been successfully submitted and would be delivered to recipients shortly.