Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Reduce duplication in & between Mock & ObjectMethods #431

Open
wants to merge 34 commits into
base: main
Choose a base branch
from

Conversation

nitishr
Copy link
Contributor

@nitishr nitishr commented Dec 2, 2019

  • Reduce duplication between {Mock,ObjectMethods}#{expects,stubs}
  • Reduce surface area by making on_stubbing{_method_unnecessarily} private (followed by a rename)

Also closes #468

This will allow us to keep the code DRY by extracting a method with the
identical bits
@nitishr nitishr changed the title Reduce duplication between Mock & ObjectMethods Reduce duplication in & between Mock & ObjectMethods Dec 2, 2019
Array will return a single-element array in case the arg is a single value, or a
Hash converted to an array in case the arg is a Hash. We then yield the single
element or the key value pair, and return the result of the last yield
Mock#{expects,stubs} already handle methods_vs_return_values/method_name_or_hash
Note that this change isn't 100% behavior preserving due to the separation of
Mockery#stub_method call from the Mock#{expects,stubs} call. It's possible that
midway through argument iteration, stub_method raises a StubbingError, which
would leave some methods stubbed without having an expectation set for them. However, I think that's still OK, since you wouldn't expect anything useful to happen after a StubbingError anyway. We're changing undocumented/unpublished behavior in a way that shouldn't matter.
Keep the stubba_ prefix to avoid accidental name clash with other valid methods
Comment on lines 359 to 369
def anticipates(method_name_or_hash, backtrace = nil, object = Mock.new(@mockery), &block)
Array(method_name_or_hash).map do |*args|
args = args.flatten
method_name = args.shift
Mockery.instance.stub_method(object, method_name) unless object.is_a?(Mock)
ensure_method_not_already_defined(method_name)
expectation = Expectation.new(self, method_name, backtrace)
expectation.returns(args.shift) unless args.empty?
yield expectation if block
@expectations.add(expectation)
end.last
Copy link
Member

@floehopper floehopper Feb 10, 2020

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Although I appreciate that this reduces the duplication, I'm not convinced that the resultant code is any easier to follow. In fact I think, if anything, it's a bit harder to follow! Let me have a think about how we might better remove the duplication.

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thanks for your review, @floehopper. Would you mind elaborating what aspects of this change you aren't convinced about and/or which ones you think make it harder to follow? If I understand your concerns more specifically, we might be able to come up with a better solution together.

To start with, I can share a bit more of my thinking here...

If it's the anticipates abstraction, I'll admit that I was initially unsure of it as well (and still am, but a lot less than I initially was). The reason I was unsure was I wondered if the need for me to look up (and cite the dictionary definitions in the PR description) was a sign of the concept being too smart/cute/nuanced.

So, I let it mull for a few days in my head and reached the conclusion that anticipates is probably a better abstraction covering both an expectation and an allowance (stub, in mocha's terminology). I've always found expectation to be too strong a term to represent an allowance. (Anticipation might be too strong a term, too - but it does seem less strong than expectation.) So, in the long term, I would even consider/suggest considering renaming, for instance, the class Expectation to Anticipation (and maybe introduce subclasses - Expectation and Allowance). I have of course not fully analyzed that chain of reasoning, and certainly not considered it from an API compatibility point of view. This was only trying to clarify the concepts and their nuances for myself.

Having said that, I'm not a native English speaker and may well have misinterpreted the shades of meanings of the various terms. So, I'd appreciate any corrections to my understanding or thinking.

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Sorry for the slow reply. I'll try to explain my reservations:

  • Although the Mock#anticipates method has removed all the duplication between ObjectMethods#expects, ObjectMethods#stubs, Mock#expects & Mock#stubs (which is great BTW!), it's now really quite a long and "dense" method and, what's more, three of the lines are only executed conditionally. Coming at the final version of this method definition cold, I found it quite hard to follow.

  • I'm not completely convinced that the inlining of the ArgumentIterator is a net gain. Encapsulating the common processing of the arguments for the four API methods in a single place still makes sense to me, even though its exact abstraction may not be quite right. And having to work out what the combination of calls to Kernel#Array, Array#map, Array#flatten, Array#shift & Array#last is doing adds quite a lot of cognitive overhead to understanding the method definition.

  • The naming of Mock#anticipates doesn't bother me too much, although I don't think it is quite right. I wonder whether it would make more sense if we made an Expectation have a cardinality of Cardinality.at_least(0) (i.e. stubbing behaviour) by default, inline Mock#anticipates into Mock#stubs, and call Mock#stubs from Mock#expects passing a block setting the expectation cardinality to Cardinality.exactly(1) (i.e. default expecting behaviour). It feels as if the expecting behaviour is a superset of the stubbing behaviour. What do you think...?

Does make sense? I'm sorry I've struggled a bit to explain my thinking!

I've had time to go through the commits one-by-one in more detail this afternoon and I'd like to mention that I found them very easy to follow - so thank you for taking so much care to curate the commits and write useful commit messages.

Having looked through the commits in more detail, I'm inclined to have a go at merging a version of this PR, albeit with some changes as per my comments above. Is that OK with you?

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

That's very helpful, @floehopper. I now understand your concerns much better.

I like the suggestion of expects constraining the Expectation, rather than stubs loosening it. Part of the reason I extracted anticipates rather than reuse expects, BTW, was to avoid changing the public API of the existing methods, even though the changes would be fully backward compatible (i.e. accepting but not expecting a block or optional arguments). So, that's a consideration.

I'd be happy with you merging a modified version of the PR with the changes you feel comfortable with. We could iterate on top of that for the more controversial/difficult aspects.

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

👍

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@floehopper, I'm planning to push another commit which I think will alleviate some of your concerns, while at the same time, improving mocha's behavior with expectation chaining by making it more consistent across single-method and multiple-method stubs/expects.

The code in that commit or even the approach could possibly be improved, but I'd like to convey the general idea/direction before I go too far.

Is that OK? If you'd rather that I wait for you to finish merging (a version of) this PR, that's fine, too. Just let me know.

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

BTW, was to avoid changing the public API of the existing methods, even though the changes would be fully backward compatible (i.e. accepting but not expecting a block or optional arguments). So, that's a consideration.

Good point! 😄

Is that OK? If you'd rather that I wait for you to finish merging (a version of) this PR, that's fine, too. Just let me know.

Yes, that's fine - please go ahead!

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Done.

I'm completely fine taking any or all of the newly added commits to a different PR so they can be discussed and reviewed on their own (or altogether dropping any/all of them). The newly introduced ExpectationSetting may need some redesign, but I hope the changes convey the general concept and approach.

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I like the suggestion of expects constraining the Expectation, rather than stubs loosening it.

Looking at the code after my last few updates, I think the current form expresses the intent better than the change we were thinking of.

def stubs(stubbed_methods_vs_return_values)
  expects(stubbed_methods_vs_return_values).at_least(0)
end

seem more intention-revealing and logical than

def expects(stubbed_methods_vs_return_values)
  stubs(stubbed_methods_vs_return_values).exactly(1)
end

Copy link
Member

@floehopper floehopper Mar 21, 2020

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@floehopper, I'm planning to push another commit which I think will alleviate some of your concerns, while at the same time, improving mocha's behavior with expectation chaining by making it more consistent across single-method and multiple-method stubs/expects.

The code in that commit or even the approach could possibly be improved, but I'd like to convey the general idea/direction before I go too far.

Is that OK? If you'd rather that I wait for you to finish merging (a version of) this PR, that's fine, too. Just let me know.

Done.

I'm completely fine taking any or all of the newly added commits to a different PR so they can be discussed and reviewed on their own (or altogether dropping any/all of them). The newly introduced ExpectationSetting may need some redesign, but I hope the changes convey the general concept and approach.

Sorry for the slow reply. Thanks for adding the extra commit - I'm going to take a look at it shortly.

@floehopper floehopper self-assigned this Feb 10, 2020
@floehopper
Copy link
Member

I've added a comment inline relating to the most significant change in this PR. While I agree that removing the duplication is good, I wonder whether there are some better abstractions we could use so the resultant code is easier to understand.

@nitishr nitishr force-pushed the mock-argument-iterator branch 3 times, most recently from 4214427 to 9bf000a Compare February 19, 2020 18:44
When calling expects/stubs with a hash (method_names_vs_return_values),
only the last expectation would get returned. Any further 'expectation
setting' methods chained to that call would, therefore, get called only
on the last expectation. This seems arbitrary, and neither evident nor
intuitive. We now 'extract' the expectation setting methods into a
_virtual_ interface, 'implemented' by both Expectation and
ExpectationSetting, and return ExpectationSetting instead of Expectation
from the expects/stubs methods. This allows us to pass any further
expectation setting method calls on to _each_ of the multiple
expectations, rather than just the _last_ (or some other arbitrary)
single expectation.
@nitishr nitishr force-pushed the mock-argument-iterator branch 2 times, most recently from 74d8631 to 5222998 Compare February 20, 2020 09:25
@nitishr
Copy link
Contributor Author

nitishr commented Feb 21, 2020

Please merge this to stop me pushing more on this branch :-)

Mock wasn't the right place for calling it. The new yield mechanism is
more general and less ugly, and allocates responsibilities more
appropriately.
Closes freerange#468 and sets us up to reduce the surface area of the internal
API exposed to users, by inlining anticipates into expects and using
that from stubs
This reduces the surface area of the internal API exposed to users as
suggested in freerange#467
Given that the public API of Expectation consists solely of expectation
setting methods, and that the matching, verifying and other misc methods
aren't part of the public API, it makes sense to equate the Expectation
protocol implicitly to the ExpectationSetting protocol. If we were to
segregate interfaces, we might wanna have other internal-facing
interfaces like ExpectationMatching, etc.

By naming this CompositeExpectation, we don't need (as much) to expose
it or to include it in/update the documentation because as far as
clients are concerned, they still get back an Expectation (technically,
something that quacks like an Expectation).

%w[
times twice once never at_least at_least_once at_most at_most_once
with yields multiple_yields returns raises throws then when in_sequence
Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Remember to add with_block_given and with_no_block_given when merging

@@ -109,14 +109,15 @@ class Mock
# object.expects(:expected_method_one).returns(:result_one)
# object.expects(:expected_method_two).returns(:result_two)
def expects(method_name_or_hash, backtrace = nil)
iterator = ArgumentIterator.new(method_name_or_hash)
iterator.each do |*args|
CompositeExpectation.new(Array(method_name_or_hash).map do |*args|
Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Looking at the documentation:

@return [Expectation] last-built expectation which can be further modified by methods on {Expectation}

This introduces a breaking change. So, even if we agree that the new behavior is preferable, we'd have to think about deprecating the old behavior.

There are no existing tests to check for the documented behavior, though.

@floehopper floehopper changed the base branch from master to main July 24, 2020 16:40
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

Successfully merging this pull request may close these issues.

Consider removing the Spanish Inquisition special case
2 participants