Thanks for looking into this Dennis!
I don't have time left before the weekend, but I can review and commit your
patch next Monday.
Concerning the column placement (0-based or 1-based) I have no idea yet as
Steve has refactored it so much I need to dive into it first.
Nevertheless, if you patch fixes at least the API breakage and we can get the
code to compile again I'll be happy to commit it :)
Regards,
Ate
Dennis Dam wrote:
Hi,
I was looking at the broken testcase TestPortletPlacement, which tests
the recently (heavily) refactored PortletPlacementContextImpl. The
constructor of PortletPlacementContextImpl was changed, so I added some
JMock objects to account for this, see the attached patch.
*however*, the tests are still broken, and this is something I don't
know how to fix, because I don't know the specs are. The problem: the
mock test data has 1-based column numbering, but the
PortletPlacementContextImpl seems to need a 0-based column count ( see
the method getColumnFromFragment() ). The result is that the column with
column nr. = column-count is reset to column-count - 1, which results in
all fragments from both column 1 and 2 ending up in the same column.
What should the behaviour be, columns must be 0-based ? In that case all
the column nrs in FragmentUtil and TestPortletPlacemen must decreased by 1.
regards,
Dennis
------------------------------------------------------------------------
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
