tests: unwind BaseModel test side effect on AclConfig
Give the test its own empty config to migrate into, which makes
resetting it irrelevant. A tiny bit of magic is needed to bootstrap
a config.xml so it's not the one that is checked in and we also
remove the temporary files from the tree so it does not confuse git.
save() is a far-reaching side effect in the test, but more or less
necessary for better test coverage. But I'd still want a simpler
test around Core\Config and this offers merely offers the groundwork
also removing the need to reset this from make commands.
firewall: slight simplification in filter sync script
Usually we can let static command line switches live in the format
string. While here omit the use of the first $output assignment.
isc-dhcp: interalize interfaces_staticarp_configure(); closes #9476
Instead of making the interface code pluggable, push the code that
causes the persistent side effect to the ISC DHCP plugin which then
gets to fix the stuck static ARP flag after disable/deinstall and a
reboot. The situation isn't ideal, but much better than before.
mvc: FilterRuleField: remove the other strpos() and reformat
We don't actually know what the separator char is although it's very
likely the default getValues() will always do the right thing and
make the following test a tiny bit easier.
While unwiwnding the if-else we can do a few simplifications along
the way.
system: use is_int()/array_key_first() in toArray() and fromArray() #9485
The approximation of the magic here is that we are looking for array
elements created by a natural append [] = or equivalent which has
an integer key of a rough range of 0 to count() - 1, but not always as
we can see from the ticket.
unset() breaks the pledge of sequential lists and makes array_is_list()
fail. Sorting would also break the sequential pledge without resetting
the keyes using array_values() but that approach is too broad.
Instead, get a single key we can do a strict int type check on so that
we are as likely to succeed as was the case before the change in 7ee3b2c.
It's also fast. ;)
system: use is_int()/array_key_first() in toArray() and fromArray() #9485
The approximation of the magic here is that we are looking for array
elements created by a natural append [] = or equivalent which has
an integer key of a rough range of 0 to count() - 1, but not always as
we can see from the ticket.
unset() breaks the pledge of sequential lists and makes array_is_list()
fail. Sorting would also break the sequential pledge without resetting
the keyes using array_values() but that approach is too broad.
Instead, get a single key we can do a strict int type check on so that
we are as likely to succeed as was the case before the change in 7ee3b2c.
It's also fast. ;)
