Cloning layers creates clones with the initial state of the original, not the current one

[ahocevar]

Because all our layers' clone() methods create a new instance with the original's options, the clones don't have the current state of the original. I think this is not what people expect.

[ahocevar]

(In [10040]) Acceptance test checking for proper clone() behavior (see #2477)

[ahocevar]

Patch with tests attached. Here is what it does: the new Layer::getOptions function walks through the options object, creating a new options object with the values from the instance. At the end of the clone process, the original's options are copied to the clone's options property.

All layer tests still pass, please review.

[tschaub]

As discussed elsewhere, it feels a bit inconsistent to me to redefine clone as something like "the clone method creates a new layer based on the current state of an existing layer" and then to find that the clone.options object isn't actually what it was constructed with but instead what the other layer was constructed with. But if there is some reason to set the clone's options to those of the original, it's ok by me.

If tests pass, please commit.

[ahocevar]

as suggested by tschaub, don't copy the original's options to the clone

[ahocevar]

(In [10045]) Changing our layers' clone methods so they create a clone of the layer state at the time of clone creation, not at the time the original was created. r=tschaub (closes #2477)

[pgiraud]

With this patch, tests fail on Layer.html.

[bartvde]

I'll look into the test breakage, thanks for the report fredj.

[bartvde]

(In [10120]) fix up tests in Layer.html after clone method was changed, since the clone now represents the original in its current state, and not the state at creation, two tests had to be adapted, non-functional change (closes #2477)

[pgiraud]

Bart got quicker than me.

OK, as far as I understand, the problem is related to the max and min resolutions. Given a maxResolution option set to "auto", a layer will compute its maxResolution (and then its minResolution) given the size of the map div. Now that the clone method creates a clone with the layer state at the time of the clone creation, the maxResolution of the cloned layer value is the computed one. For now, I don't see any issue with this. If others are also ok with this, I think we can remove the tests on min/maxResolution (See attached patch). The previous test which checks that optionschicken? is correct is sufficient, in my opinion.

[bartvde]

I am fine with removing these two test cases.

[tschaub]

I'm ok with this as well.

[pgiraud]

(In [10140]) let's remove unrequired tests, r=bartvde, (Closes #2477)