| Mike,
The normative statements this would test is ASM60030 and ASM60042. Using a name that cannot be mapped to a file name is a further check that qname resolution is being performed as opposed to some other mechanism, which is mandated by the Assembly spec.
Another way to look at this is: would a runtime be SCA "compliant" if it required composite file names to be mapped to the local part of a composite qname? I personally don't think so.
Jim
On Mar 11, 2010, at 4:07 PM, Mike Edwards wrote:
Jim,
There is no explicit test that checks
that an SCA runtime can resolve the QName of a composite where the
filename which contains that composite
is different from the local part of the composite QName.
The reason that there is no such explicit
test is that there is no normative statement in the SCA Assembly spec
that says anything about the relationship
between the filename of a composite file and the QName of the
composite contained within that file.
I think that if this is a problem, then
the first step would be to raise an issue against the Assembly spec to
add
the required normative statement - that
would then cause the necessary TA and Testcase to be added.
Yours, Mike.
Strategist - Emerging Technologies, SCA & SDO.
Co Chair OASIS SCA Assembly TC.
IBM Hursley Park, Mail Point 146, Winchester, SO21 2JN, Great Britain.
Phone & FAX: +44-1962-818014 Mobile: +44-7802-467431
Email: mike_edwards@uk.ibm.com
Mike,
What I think should be tested is the condition where a
QName does not naturally map to the filename of a composite. For example:
qname: {urn:foo:bar}MyComposite
file name: thecomposite.composite
The one testcase where this comes up is ASM_12008, which
you identified previously. However, I think there should be an explicit
test for this. I don't believe that is the case but I could be mistaken.
So, let me phrase it differently: is there a reason why
we would not include an explicit test for this condition?
Jim
On Mar 11, 2010, at 12:56 PM, Mike Edwards wrote:
Jim,
If your concern is to ensure that there are appropriate testcases for the
<implementation.composite/> and
<include/> elements, then I believe that the work is already done
and that there is no need to write any further tests.
For <implementation.composite/>, the requirement for the @name attribute
to resolve to the QName of a
<composite/> element in the Domain is described by ASM60030.
Looking at the TestAssertions and TestCases documents:
ASM60030 --> ASM-TA-6031 --> ASM_6024_TestCase
For <include/>, the requirement for the @name attribute to resolve
to the QName of a <composite/>
element in the Domain is described by ASM60042. For this:
ASM60042 --> ASM-TA-6041 --> ASM_6030_TestCase
Note that these 2 testcases cover the negative version of the test ie where
the name does NOT resolve correctly.
The positive test (that the name resolves correctly) is actually covered
by a whole pile of other testcases
that use <implementation.composite/> and <include/> and expect
them to work correctly (ie if the resolution
did not work, then the testcase would fail and that would pick up the problem).
Best wishes,
Yours, Mike.
Strategist - Emerging Technologies, SCA & SDO.
Co Chair OASIS SCA Assembly TC.
IBM Hursley Park, Mail Point 146, Winchester, SO21 2JN, Great Britain.
Phone & FAX: +44-1962-818014 Mobile: +44-7802-467431
Email: mike_edwards@uk.ibm.com
Jim Marino <jim.marino@gmail.com>
wrote on 11/03/2010 10:08:45:
> [image removed]
>
> [sca-assembly] Re: [ASSEMBLY 204] Assembly Tests do not verify
> composite name resolution
>
> Jim Marino
>
> to:
>
> OASIS Assembly
>
> 11/03/2010 10:09
>
> Sorry for the delayed response but I was out on vacation.
>
> I will be happy to submit a testcase for this.
>
> I'm confused by your last statement about the name of the composite
> file. The issue is not about mandating the contents of a composite
> file name but about verifying qname resolution is performed
> correctly when a composite is referenced from the
> <implementation.composite/> or <include/> elements. The
behavior is
> spelled out in the Assembly spec in a number of places, for example
> when detailing the <implementation.composite/> elements:
>
> "name (1..1) – the name of the composite used as an implementation.
> The @name attribute of an <implementation.composite/> element
MUST
> contain the QName of a composite in the SCA Domain"
>
> Jim
>
>
> On Mar 3, 2010, at 2:32 PM, Mike Edwards wrote:
>
>
> Jim,
>
> This note follows from a discussion of ASSEMBLY 204 on the Assembly
> TC call yesterday.
>
> The proposal is that we shall close this issue with no action,
> unless you are able to submit
> a complete testcase for this situation.
>
> One thing to note is that there is ONE existing testcase where the
> testcase uses a composite
> by reference from an <implementation.composite /> element, where
the
> name of the composite
> element does not match the filename.
>
> The testcase concerned is ASM_12008 - and here there is a reference
> to a composite with the
> name test2:TestComposite5 that is expected to resolve to a composite
> contained within a file
> with the name "TestComposite5a.composite" - ie. the filename
and the
> composite name are
> not the same.
>
> However, testing the name resolution is not the main goal of this
> testcase - but it does depend
> on getting the resolution done correctly.
>
>
> The second thing to note is that there is no specific normative
> statement which relates to the name
> of the file containing a composite element other than that it should
> have the suffix ".composite" -
> so it is an interesting question as to which normative statement
> would be the subject of a test
> specifically checking that a composite can be contained in a file
> with a name that is different from
> the name of the composite itself.
>
> Yours, Mike.
>
> Strategist - Emerging Technologies, SCA & SDO.
> Co Chair OASIS SCA Assembly TC.
> IBM Hursley Park, Mail Point 146, Winchester, SO21 2JN, Great Britain.
> Phone & FAX: +44-1962-818014 Mobile: +44-7802-467431
> Email: mike_edwards@uk.ibm.com
>
>
>
> Unless stated otherwise above:
> IBM United Kingdom Limited - Registered in England and Wales with
> number 741598.
> Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire
PO6 3AU
>
>
>
>
Unless stated otherwise above:
IBM United Kingdom Limited - Registered in England and Wales with number
741598.
Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire PO6
3AU
Unless stated otherwise above:
IBM United Kingdom Limited - Registered in England and Wales with number
741598.
Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire PO6
3AU
|