Fixes a couple segment handling regressions in the UI #15110
Conversation
…segment since it appears to be more common.
…d segment is one being compared.
|
@diosmosis there are quite a few system tests failing: https://travis-ci.org/matomo-org/matomo/jobs/607490406#L849 Not sure this is expected? |
…t way to avoid side effects.
|
Tried looking for the cause of the regressions, but couldn't figure out the other encoding issue so I ended up solving this a different way. |
|
@diosmosis looks good. There's a failing system tests still though https://travis-ci.org/matomo-org/matomo/jobs/607961524#L810 probably from the fixture change in the PR. Not sure that fixture change is still needed or expected? Do you reckon it's possible to add a test for the case it fixes? |
|
Shouldn't be a fixture change here, that's probably something that happens locally on os x. Maybe. There's a change to the test file in this PR I'll just remove. |
|
@diosmosis if a test can be added for this that be great, otherwise feel free to merge. |
FIxes #15080
Changes:
&segment=&segment=...).