• From: Ihe Onwuka <ihe.onwuka@g...>
• To: xml-dev@l...
• Date: Wed, 10 Apr 2013 17:23:38 +0100

On Wed, Apr 10, 2013 at 4:52 PM, Michael Kay <mike@s...> wrote:
>
>
>
>>> Below if as the result of changes to the structure in the XMLe any of
>>> the elements "some" "path" or "to" change, disappear, become reodered,
>>> or now require predicates to be included, changed or removed that test
>>> will not work. Note the multitude of things that the test is
>>> vulnerable to.
>
> The key think with test assertions is to assert the minimum required for success. If it's necessary for the column heading of a table to be bold, include that in the assertion; if this is incidental and irrelevant, then don't. XPath is quite flexible enough to enable you to assert what's necessary and ignore what isn't; that's it's great strength.
>
> Michael Kay
> Saxonica

Right. But XMLUnit doesn't let you do that because it requires an
absolute path (thank you John) to the node you are testing.

• Follow-Ups:
  • Re: Testing XML don't use xUnit
    • From: Fraser Goffin <goffinf@g...>
  • Re: Testing XML don't use xUnit
    • From: Michael Kay <mike@s...>
  • Re: Testing XML don't use xUnit
    • From: Fraser Goffin <goffinf@g...>

• References:
  • Testing XML don't use xUnit
    • From: Ihe Onwuka <ihe.onwuka@g...>
  • Re: Testing XML don't use xUnit
    • From: Andrew Welch <andrew.j.welch@g...>
  • Re: Testing XML don't use xUnit
    • From: Ihe Onwuka <ihe.onwuka@g...>
  • Re: Testing XML don't use xUnit
    • From: Andrew Welch <andrew.j.welch@g...>
  • Re: Testing XML don't use xUnit
    • From: Ihe Onwuka <ihe.onwuka@g...>
  • Re: Testing XML don't use xUnit
    • From: Andrew Welch <andrew.j.welch@g...>
  • Re: Testing XML don't use xUnit
    • From: Ihe Onwuka <ihe.onwuka@g...>
  • Re: Testing XML don't use xUnit
    • From: Andrew Welch <andrew.j.welch@g...>
  • Re: Testing XML don't use xUnit
    • From: Michael Kay <mike@s...>