Implement outer join and friends

[marcelogomez]

This PR implements all of the methods described this comment on #6 :

• outer_join: Takes in a reference to a second HashBag and returns an iterator that yields the union of the key sets for both HashBags (ie. keyset(self) ∪ keyset(other)). Each key is yielded along with its corresponding counts on each HashBag.
  • The rest of the functions added in this PR as well as difference introduced in Implement HashBag::difference(&self, other: &HashBag) #6 can be implemented using this function
• not_in: Takes in a reference to a second HashBag and returns an iterator that yields the set difference between the key set of the first HashBag and the second one (i.e. keyset(self) \ keyset(other)). Each key is yielded along with its corresponding count on the first HashBag.
• signed_difference: Takes in a reference to a second HashBag and returns an iterator that yields the union of the keys sets for both HashBags (ie. keyset(self) ∪ keyset(other). Each key is yielded along with the difference between its count on the first and second HashBags.

[codecov]

Codecov Report

Merging #8 (c551d0c) into main (d1140b8) will increase coverage by 6.6%.
The diff coverage is 100.0%.

Impacted Files	Coverage Δ
src/lib.rs	94.7% <100.0%> (+2.1%)	⬆️
src/serde.rs	95.3% <0.0%> (+95.3%)	⬆️

[jonhoo]

This looks great, thanks for taking it on! The only thing I'm missing is a test that checks that outer_join indeed returns all the elements of self before any elements that are exclusively in other. We promise that in the docs, so we should have a test that the implementation meets that promise.

[marcelogomez]

This looks great, thanks for taking it on! The only thing I'm missing is a test that checks that outer_join indeed returns all the elements of self before any elements that are exclusively in other. We promise that in the docs, so we should have a test that the implementation meets that promise.

Thanks for taking a look! :D

Yeah, will add that test! However, that's not what I meant when I wrote:

The iterator also yields the respective counts in self and other in that order.

The order I was referring was the the order of the values in the tuple that the iterator yields (i.e. it's (key, self_count, other_count). Will change the documentation to make that clearer

[jonhoo]

The order I was referring was the the order of the values in the tuple that the iterator yields (i.e. it's (key, self_count, other_count). Will change the documentation to make that clearer

Ah, I see! Yeah, I think we want to make both promises in the docs (and check in tests) — that the first count in the tuple is from self, and that all elements not in other are yielded first.

[marcelogomez]

Ah, I see! Yeah, I think we want to make both promises in the docs (and check in tests) — that the first count in the tuple is from self, and that all elements not in other are yielded first.

The tests for outer_join assert on the exact values yielded by the iterator so I think the ordering within the tuple is already covered by tests :) I just added new tests for the ordering between yielded elements

[jonhoo]

Looks great, thanks!

[jonhoo]

Published as 0.1.7 🎉