Moe Sync

[ronshapiro]

This code has been reviewed and submitted internally. Feel free to discuss on the PR and we can submit follow-up changes as necessary.

Commits:

Stop passing unused VisitorState to SuppressionHelper.

RELNOTES: None

84f3bbc

---

Ignore super(...) invocations in MustBeClosedChecker

RELNOTES: Ignore super(...) invocations in MustBeClosedChecker

75b0055

---

OrphanedFormatString: Only consider StringBuilder.append a likely error if the append(CharSequence, int, int) or append(char[], int, int) overloads are invoked.

RELNOTES: OrphanedFormatString: only consider StringBuilder.append a likely error for some overloads (#1021)

233651e

---

Fix flagging serialization methods in Unused.

Extract the test so it's a bit clearer.

RELNOTES: N/A

ccdc625

---

Make ant example's path handling work on Windows

3f2ca43#commitcomment-30383244

RELNOTES: N/A

2e89884

---

Remove ProtoFieldPreconditionsCheckNotNull, as it is subsumed into ProtoFieldNullComparison.

RELNOTES: N/A

45cc93d

---

Fix treatment of non-trivial receivers of method selects in Unused.

RELNOTES: Fix a false positive in Unused.

c6c010f

---

Provide a marginally better description in BadInstanceof.

It seems slightly friendlier to also include the actual type of the expression to make the issue a bit clearer.

RELNOTES: N/A

65fe799

---

Provide a different description when removing the `default` case in UnnecessaryDefaultInEnumSwitch.

RELNOTES: N/A

5b094ec

---

Remove altName of BadNestedImport from BadImport checker.

There is no need to continue to support both names.

RELNOTES: Remove altName of BadNestedImport from BadImport checker.

135c920

---

Add support for suppression of FieldCanBeFinal

Fixes #1038

RELNOTES: N/A

e5eb00a

---

Add a bug pattern to discourage writing #equals methods by just calling hashCode.

I'm kinda surprised this is a thing. We only match where the hashCode is the last thing compared within a `return, as I think this is a fairly sane pattern, especially with memoized hashCodes:

return hashCode() == o.hashCode() && ;

RELNOTES: [EqualsUsingHashCode] add a bug pattern to discourage writing equals using hashCode.

4fc4181

---

Add a check for obviously inconsistent hashCode implementations.

This will have plenty of false negatives for classes which compare fields via complex getters, but should work fairly well for simple data classes.

Fixes #35

RELNOTES: [InconsistentHashCode] Add a check to flag inconsistent implementations of hashCode.

e8cf0b5

---

Make BooleanParameter match constructors too.

And add a couple of fairly common terms to the list of parameter names to ignore.

RELNOTES: Make BooleanParameter match constructors.

36642fa

---

In BooleanParameter, don't refactor non-boolean parameters, e.g. generic parameters or Object.

RELNOTES: BooleanParameter now skips non-boolean parameters, e.g. generics.

18c54a0

---

Revise docs to try to make them clearer.

4667101

---

Fix ParcelableCreator check for raw creator

RELNOTES: Fix ParcelableCreator check for raw creator

f8d8e93

---

Create namedAnyOf(String... varargs) convenience MethodMatchers method.

RELNOTES: N/A

3926cde

---

Show more Error Prone diagnostics at once

RELNOTES: Report errors for all files with Error Prone diagnostics, not just the first one.

435c6ba

---

Make TypeParameterShadowing rename Javadoc references to type parameters too.

RELNOTES: Make TypeParameterShadowing fix Javadocs too.

93b2a7e

---

Make ParameterComment match non-parameter-comments before arguments too.

Previously it wouldn't match the non-standard form foo(/* bar */ baz).

RELNOTES: Make ParameterComment match before comments too.

09b08e9

---

Make insertion coalescing in Replacements#add explicit

Instead of silently concatenating overlapping insertions together, require
callers of Replacements#add to explicitly requires a policy for concatenating
them (either existing-first, or new-first). The default behaviour now throws
an exception for overlapping insertions.

RELNOTES: N/A

f6d7c53

---

Call out CharSequence in UndefinedEquals docs RELNOTES: N/A

958c0df

---

Fix breakage caused by f6d7c53

f9c9e41

---

Add a qualifyType overload which takes a String of the type name.

A common pattern is qualifyType(..., state.getTypeFromString(typeName)) which stands a fair chance of NPEing if typeName isn't available.

RELNOTES: Add String-taking qualifyType overload.

803cf50

---

Exempt constructors for classes containing the term "boolean" (aimed at AtomicBoolean).

These are an obvious exception to my claim that "constructors accepting a single boolean are rarely self documenting".

(Also include 'defaultValue'.)

RELNOTES: N/A

ebe4d09

[googlebot]

So there's good news and bad news.

👍 The good news is that everyone that needs to sign a CLA (the pull request submitter and all commit authors) have done so. Everything is all good there.

😕 The bad news is that it appears that one or more commits were authored or co-authored by someone other than the pull request submitter. We need to confirm that all authors are ok with their commits being contributed to this project. Please have them confirm that here in the pull request.

Note to project maintainer: This is a terminal state, meaning the cla/google commit status will not change from this state. It's up to you to confirm consent of all the commit author(s), set the cla label to yes (if enabled on your project), and then merge this pull request when appropriate.

[googlebot]

A Googler has manually verified that the CLAs look good.

(Googler, please make sure the reason for overriding the CLA status is clearly documented in these comments.)

[eaftan]

It looks like commit 435c6ba broke the JDK 11 build. @cushon can you PTAL?