Moves KeyError into Shared spec #576
Conversation
This refactors my work on KeyError features into a shared spec to reduce the duplication.
shared/hash/key_error.rb
Outdated
| -> { | ||
| @method.call(@object, 'foo') | ||
| }.should raise_error(KeyError) { |err| | ||
| err.receiver.should == @object |
There was a problem hiding this comment.
This could be should equal(@object) to make sure it's the same object.
core/hash/fetch_values_spec.rb
Outdated
| it "returns the default value from block" do | ||
| @hash.fetch_values(:z) { |key| "`#{key}' is not found" }.should == ["`z' is not found"] | ||
| @hash.fetch_values(:a, :z) { |key| "`#{key}' is not found" }.should == [1, "`z' is not found"] | ||
| end |
There was a problem hiding this comment.
It's the group describe "with unmatched keys" do below which should be replace by it_behaves_like.
| lambda { {}.fetch(:a) }.should raise_error(KeyError) | ||
| lambda { Hash.new(5).fetch(:a) }.should raise_error(KeyError) | ||
| lambda { Hash.new { 5 }.fetch(:a) }.should raise_error(KeyError) | ||
| end |
There was a problem hiding this comment.
It would be good to keep these examples which use different kinds of Hash.
There was a problem hiding this comment.
I've added 3 more it_behaves_like calls so it can test all the shared specs in each of the Hash creation methods.
That seems unintended bugs. Could you file an issue on https://github.com/ruby/mspec ? |
|
I've updated the code per review comments, and I'll file an issue with ruby/mspec detailing the scoping problems of |
|
Thank you for the refactoring! |
This refactors my work on KeyError (#574) features into a shared spec to reduce the duplication.
There are two oddities that I should point out, both are limitations I've found with the implementation of
it_behaves_like:it_behaves_likecalls must have an identical@methodand@objectcalls within a given scope, or they will affect each others' state. To overcome this, you have to wrap the differing parameter in acontextordescribeblock.@methodhas shared scope. To get around this here, I have saved the outer@methodinto a new@base_methodinside abeforeand referred to it via@base_methodto solve the collision.I'm not sure if the two above issues are intentional side-effects of mspec's implementation or unintended bugs so I'm not sure if I should file an issue, but I thought I'd bring it up since it was relevant to my work on this.