Detect calls to scalb that should be using the double overload in…

…stead

PiperOrigin-RevId: 711799309

core/src/main/java/com/google/errorprone/bugpatterns/IntFloatConversion.java

@@ -0,0 +1,71 @@
+/*
+ * Copyright 2024 The Error Prone Authors.
+ *
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ *     http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+package com.google.errorprone.bugpatterns;
+
+import static com.google.errorprone.BugPattern.SeverityLevel.WARNING;
+import static com.google.errorprone.fixes.SuggestedFix.prefixWith;
+import static com.google.errorprone.matchers.Description.NO_MATCH;
+import static com.google.errorprone.util.ASTHelpers.getType;
+import static com.google.errorprone.util.ASTHelpers.targetType;
+
+import com.google.errorprone.BugPattern;
+import com.google.errorprone.VisitorState;
+import com.google.errorprone.bugpatterns.BugChecker.MethodInvocationTreeMatcher;
+import com.google.errorprone.matchers.Description;
+import com.google.errorprone.matchers.Matcher;
+import com.google.errorprone.matchers.method.MethodMatchers;
+import com.google.errorprone.util.ASTHelpers.TargetType;
+import com.sun.source.tree.ExpressionTree;
+import com.sun.source.tree.MethodInvocationTree;
+import com.sun.source.tree.Tree;
+import com.sun.tools.javac.code.TypeTag;
+
+/** A {@link BugChecker}; see the associated {@link BugPattern} annotation for details. */
+@BugPattern(
+    summary =
+        "Conversion from int to float may lose precision; use an explicit cast to float if this"
+            + " was intentional",
+    severity = WARNING)
+public class IntFloatConversion extends BugChecker implements MethodInvocationTreeMatcher {
+
+  /**
+   * int to float conversions aren't always problematic, this specific issue is that there are float
+   * and double overloads, and when passing an int the float will be resolved in situations where
+   * double precision may be desired.
+   */
+  private static final Matcher<ExpressionTree> MATCHER =
+      MethodMatchers.staticMethod()
+          .onClass("java.lang.Math")
+          .named("scalb")
+          .withParameters("float", "int");
+
+  @Override
+  public Description matchMethodInvocation(MethodInvocationTree tree, VisitorState state) {
+    if (!MATCHER.matches(tree, state)) {
+      return NO_MATCH;
+    }
+    Tree arg = tree.getArguments().get(0);
+    if (!getType(arg).hasTag(TypeTag.INT)) {
+      return NO_MATCH;
+    }
+    TargetType targetType = targetType(state);
+    if (targetType == null || !targetType.type().hasTag(TypeTag.DOUBLE)) {
+      return NO_MATCH;
+    }
+    return describeMatch(arg, prefixWith(arg, "(double) "));
+  }
+}

core/src/main/java/com/google/errorprone/scanner/BuiltInCheckerSuppliers.java

@@ -190,6 +190,7 @@
 import com.google.errorprone.bugpatterns.InputStreamSlowMultibyteRead;
 import com.google.errorprone.bugpatterns.InsecureCipherMode;
 import com.google.errorprone.bugpatterns.InstanceOfAndCastMatchWrongType;
+import com.google.errorprone.bugpatterns.IntFloatConversion;
 import com.google.errorprone.bugpatterns.IntLongMath;
 import com.google.errorprone.bugpatterns.InterfaceWithOnlyStatics;
 import com.google.errorprone.bugpatterns.InterruptedExceptionSwallowed;
@@ -963,6 +964,7 @@ public static ScannerSupplier warningChecks() {
           Inliner.class,
           InputStreamSlowMultibyteRead.class,
           InstanceOfAndCastMatchWrongType.class,
+          IntFloatConversion.class,
           IntLongMath.class,
           InvalidBlockTag.class,
           InvalidInlineTag.class,

core/src/test/java/com/google/errorprone/bugpatterns/IntFloatConversionTest.java

@@ -0,0 +1,60 @@
+/*
+ * Copyright 2024 The Error Prone Authors.
+ *
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ *     http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+package com.google.errorprone.bugpatterns;
+
+import com.google.errorprone.CompilationTestHelper;
+import org.junit.Test;
+import org.junit.runner.RunWith;
+import org.junit.runners.JUnit4;
+
+@RunWith(JUnit4.class)
+public class IntFloatConversionTest {
+
+  private final CompilationTestHelper testHelper =
+      CompilationTestHelper.newInstance(IntFloatConversion.class, getClass());
+
+  @Test
+  public void positive() {
+    testHelper
+        .addSourceLines(
+            "Test.java",
+            """
+            class Test {
+              double f(int x) {
+                // BUG: Diagnostic contains:
+                return Math.scalb(x, 2);
+              }
+            }
+            """)
+        .doTest();
+  }
+
+  @Test
+  public void negative() {
+    testHelper
+        .addSourceLines(
+            "Test.java",
+            """
+            class Test {
+              float f(int x) {
+                return Math.scalb(x, 2);
+              }
+            }
+            """)
+        .doTest();
+  }
+}

docs/bugpattern/IntFloatConversion.md

@@ -0,0 +1,20 @@
+Implicit conversions from `int` to `float` may lose precision when when calling
+methods with overloads that accept both`float` and `double`.
+
+For example, `Math.scalb` has overloads
+[`Math.scalb(float, int)`](https://docs.oracle.com/en/java/javase/21/docs/api/java.base/java/lang/Math.html#scalb\(float,int\))
+and
+[`Math.scalb(double, int)`](https://docs.oracle.com/en/java/javase/21/docs/api/java.base/java/lang/Math.html#scalb\(double,int\)).
+
+When passing an `int` as the first argument, `Math.scalb(float, int)` will be
+selected. If the result of `Math.scalb(float, int)` is then used as a `double`,
+this may result in a loss of precision.
+
+To avoid this, an explicit cast to `double` can be used to call the
+`Match.scalb(double, int)` overload:
+
+```java
+int x = ...
+int y = ...
+double f = Math.scalb((double) x, 2);
+```