[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]
Subject: Re: [ebxml-iic] Errata list for versoin 2.0 of IIC Test Framework
----- Original Message -----
From: Tim Sakach
Cc: Brandon Bethke ; Brian Fliege
Sent: Thursday, September 25, 2003 3:20
PM
Subject: RE: [ebxml-iic] Errata list for
versoin 2.0 of IIC Test Framework
Here are comments from my team on the erratta
list:
1. Would like
to specify a schematron file to use for
validation. It could be as simple as adding an additional type for the existing
attribute contentType
on <ValidateContent>
For
instance:
contentType="SchematronFile" or "Schematron"
<ValidateContent contentType="SchematronFile"
schemaLocation="oag/processpurchaseorder/ProcessPurchaseOrder.sch"/>
[MIKE]
- This looks doable. I will add this to the proposed new schema and
Errata.
2.
Please clarify or elaborate the difference between <SetParameter> in
GetMessage and the nested SetParameter in <GetMessage><GetPayload>.
The description has them both doing the same thing. "Container for user-defined parameter to be made available to other Test
Steps ( source can be this message's XML content retrieved via XPath statement
or simply a user-supplied string value)".
[MIKE] - You are correct... they appear to
do the same thing. I need to specify more clearly that
<SetParameter> ( in the context of being a child of a <GetPayload>
parent ) means to possible set a parameter value based upon the XPath evaluation
of XML payload content. <SetParameter) ( in the context of being a
child of a <GetMessage> parent ) means means to possible set a parameter
value based upon the XPath evaluation of the XML message envelope
content. In both cases, I propose moving <SetParameter>
to a location "preceeding" <TestPreCondition> and
<TestAssertion>, as I think that would make more sense. Also, I
suggest moving <SetParamter> to a location preceeding
<PutMessage> ( Test Framework Errata #5 and #6
)
3. Consider the possiblity of standard
variables that can be used to cause the engine to generate a date-time value.
For example:
TestDriverDateTime
<SetParameter>
<Name>MyDateTimeVar</Name> <Value>TestDriverDateTime</Value> <SetParameter> This would cause the test driver to create
a variable named " MyDateTimeVar". The value of the variable would be a date
time value generated by the test
driver.
If the name TestDriverDateTime doesn't
isn't appealing, then maybe we could make it be more function like:
test-driver-date-time()
Another useful
example,
<SetParameter> <Name>MyGUID</Name> <Value>TestDriverGUID</Value> <SetParameter> [MIKE] - I support both of these. I
will add this to the Errata. Any other possible "intrinsic" run-time
variables you can think of?
4. For optimization, is is possible for
the TestRequirement documents to refer to the TestSuite document that contains
the necessary TestCases that fulfil the requirement? Right now it requires a
disk search and file scanning to determine the correct test cases for a
particular requirement. For optimization, and to avoid a disk lookup and file
scanning, we have to specify in a database what requirement goes with what test
case.
Profile ->
requirement1
-> requirement2
requirement1 -> testsuite1/test
case1
->
testsuite2/test case5
requirement2 ->
testsuite1/testcase3
[MIKE] - We could do that.. I'll propose
that in the Errata.
Regards,
Mike
|
[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]