[clang-tools-extra] [clang-tidy] Add new check bugprone-tagged-union-member-count (PR #89925)

[via cfe-commits]

=?utf-8?b?R8OhYm9yIFTDs3RodsOhcmk=?= <tigbrcode at protonmail.com>,
=?utf-8?b?R8OhYm9yIFTDs3RodsOhcmk=?= <tigbrcode at protonmail.com>,
=?utf-8?b?R8OhYm9yIFTDs3RodsOhcmk=?= <tigbrcode at protonmail.com>,
=?utf-8?b?R8OhYm9yIFTDs3RodsOhcmk=?= <tigbrcode at protonmail.com>,
=?utf-8?b?R8OhYm9yIFTDs3RodsOhcmk=?= <tigbrcode at protonmail.com>,
=?utf-8?b?R8OhYm9yIFTDs3RodsOhcmk=?= <tigbrcode at protonmail.com>
Message-ID:
In-Reply-To: <llvm.org/llvm/llvm-project/pull/89925 at github.com>


================
@@ -0,0 +1,214 @@
+//===--- TaggedUnionMemberCountCheck.cpp - clang-tidy ---------------------===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+
+#include "TaggedUnionMemberCountCheck.h"
+#include "../utils/OptionsUtils.h"
+#include "clang/ASTMatchers/ASTMatchFinder.h"
+#include "llvm/ADT/STLExtras.h"
+#include "llvm/ADT/SmallSet.h"
+
+using namespace clang::ast_matchers;
+
+namespace clang::tidy::bugprone {
+
+const char StrictModeOptionName[] = "StrictMode";
+const char EnableCountingEnumHeuristicOptionName[] =
+    "EnableCountingEnumHeuristic";
+const char CountingEnumPrefixesOptionName[] = "CountingEnumPrefixes";
+const char CountingEnumSuffixesOptionName[] = "CountingEnumSuffixes";
+
+TaggedUnionMemberCountCheck::TaggedUnionMemberCountCheck(
+    StringRef Name, ClangTidyContext *Context)
+    : ClangTidyCheck(Name, Context),
+      StrictMode(Options.get(StrictModeOptionName, false)),
+      EnableCountingEnumHeuristic(
+          Options.get(EnableCountingEnumHeuristicOptionName, true)),
+      RawCountingEnumPrefixes(Options.get(CountingEnumPrefixesOptionName, "")),
+      RawCountingEnumSuffixes(
+          Options.get(CountingEnumSuffixesOptionName, "count")),
+      ParsedCountingEnumPrefixes(
+          utils::options::parseStringList(RawCountingEnumPrefixes)),
+      ParsedCountingEnumSuffixes(
+          utils::options::parseStringList(RawCountingEnumSuffixes)),
+      CountingEnumPrefixesSet(
+          Options.get(CountingEnumPrefixesOptionName).has_value()),
+      CountingEnumSuffixesSet(
+          Options.get(CountingEnumSuffixesOptionName).has_value()),
+      CountingEnumConstantDecl(nullptr) {
+  if (!EnableCountingEnumHeuristic) {
+    if (CountingEnumPrefixesSet)
+      configurationDiag("%0: Counting enum heuristic is disabled but "
+                        "CountingEnumPrefixes is set")
+          << Name;
+    if (CountingEnumSuffixesSet)
+      configurationDiag("%0: Counting enum heuristic is disabled but "
+                        "CountingEnumSuffixes is set")
+          << Name;
+  }
+}
+
+void TaggedUnionMemberCountCheck::storeOptions(
+    ClangTidyOptions::OptionMap &Opts) {
+  Options.store(Opts, StrictModeOptionName, StrictMode);
+  Options.store(Opts, EnableCountingEnumHeuristicOptionName,
+                EnableCountingEnumHeuristic);
+  Options.store(Opts, CountingEnumPrefixesOptionName, RawCountingEnumPrefixes);
+  Options.store(Opts, CountingEnumSuffixesOptionName, RawCountingEnumSuffixes);
+}
+
+void TaggedUnionMemberCountCheck::registerMatchers(MatchFinder *Finder) {
+  Finder->addMatcher(
+      recordDecl(
+          anyOf(isStruct(), isClass()),
+          has(fieldDecl(hasType(qualType(hasCanonicalType(recordType()))))
+                  .bind("union")),
+          has(fieldDecl(hasType(qualType(hasCanonicalType(enumType()))))
+                  .bind("tags")))
+          .bind("root"),
+      this);
+}
+
+static bool isUnion(const FieldDecl *R) {
+  return R->getType().getCanonicalType().getTypePtr()->isUnionType();
+}
+
+static bool isEnum(const FieldDecl *R) {
+  return R->getType().getCanonicalType().getTypePtr()->isEnumeralType();
+}
+
+static bool hasMultipleUnionsOrEnums(const RecordDecl *Rec) {
+  return llvm::count_if(Rec->fields(), isUnion) > 1 ||
+         llvm::count_if(Rec->fields(), isEnum) > 1;
+}
+
+static bool signEquals(const llvm::APSInt &A, const llvm::APSInt &B) {
+  return (A.isNegative() && B.isNegative()) ||
+         (A.isStrictlyPositive() && B.isStrictlyPositive()) ||
+         (A.isZero() && B.isZero());
+}
----------------
whisperity wrote:

**If** the following is equivalent (I would guess `isNegative`/`isStrictlyPositive`/`isZero` are all mutually exclusive properties) then perhaps expressing the query like this would better reflect that we are checking for some *equality* here:

```suggestion
static bool signEquals(const llvm::APSInt &A, const llvm::APSInt &B) {
  return (A.isNegative() == B.isNegative()) &&
         (A.isStrictlyPositive() == B.isStrictlyPositive()) &&
         (A.isZero() == B.isZero());
}
```

-----

Also, I am **very** surprised that `APInt`/`APSInt` does not have the facility for querying the [sign function](https://en.wikipedia.org/wiki/Sign_function) as a single value. Tangentially, perhaps you could add that in a separate patch with a few tests, a function that returns a 3-way enum `NEGATIVE, ZERO, POSITIVE`, and then this function simplifies to `A.getSign() == B.getSign()`, which is the most trivial expansion of the function's name.

https://github.com/llvm/llvm-project/pull/89925