GitLab

When backporting patches the different branch info comes thru and the
first item in array is not necessarily the one we want to check so we
need to check the object with right branch information.
Signed-off-by: Robert Lyon <robertl@catalyst.net.nz>

Signed-off-by: Robert Lyon <robertl@catalyst.net.nz>

And alerting user to the problem
Signed-off-by: Robert Lyon <robertl@catalyst.net.nz>

This reverts commit af3ec01f.

Signed-off-by: Robert Lyon <robertl@catalyst.net.nz>

When the first patch being checked is already merged.

We need to reject out before checking gerrit as we don't have a commit
id to check against
Signed-off-by: Robert Lyon <robertl@catalyst.net.nz>

Signed-off-by: Robert Lyon <robertl@catalyst.net.nz>

Now automated testing will fail for all patches that don't make a change
to at least one Behat test file. Unless the patch commit message includes
the phrase "behatnotneeded".

This happens if the job is added to jenkins queue and then patchset is
updated/rebased before the job is run.

Patch will stop running tests on obsolete patchset
Signed-off-by: Robert Lyon <robertl@catalyst.net.nz>

But not for any parents it has.
This is useful if we retrigger the test from jenkins
Signed-off-by: Robert Lyon <robertl@catalyst.net.nz>

unknown revision or path not in the working tree.
Signed-off-by: Robert Lyon <robertl@catalyst.net.nz>

Signed-off-by: Robert Lyon <robertl@catalyst.net.nz>

Signed-off-by: Robert Lyon <robertl@catalyst.net.nz>

Signed-off-by: Robert Lyon <robertl@catalyst.net.nz>

To Do - work out why the call to the php file is failing
Signed-off-by: root <root@mahara-jenkins>

Eg. if the patch or one of it's parents has a reject status then there
is no point running the behat - we should exit out and let user know
to fix up patches first
Signed-off-by: Robert Lyon <robertl@catalyst.net.nz>

This reverts commit d3450897.