tree 980d8a4fd352edca4bfdca34d18d6570b3122007
parent fec8c4f5123d557eb3b0cfd238e0ecc673bb6840
author peteg <peteg@google.com> 1553162693 -0700
committer Ron Shapiro <shapiro.rd@gmail.com> 1554214040 -0400

Make `IterableSubject.<A, E>comparingElementsUsing` accept `Correspondence<? super A, ? super E>` instead of `Correspondence<A, E>`. And similarly for MapSubject and MultimapSubject.

This accurately reflects the correct constraint. It also allows things which aren't possible at the moment, e.g. making assertions about an iterable of a subtype using a correspondence which accepts the supertype and then doing displayingDiffsPairedBy using a key function which accepts the subtype. (The caller will need to specify the generics rather than letting them be inferred, but at least it's possible --- right now, doing this would require an unchecked cast.)

While we're at it, replace `? extends Object` with the simpler `?` in a few places.

RELNOTES=n/a

-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=239564279