Add --incompatible_utf8_locale_in_action_env

bazelbuild · Apr 8, 2022 · cdedef8 · cdedef8
1 parent 2b3bd3e
commit cdedef8
Show file tree

Hide file tree

Showing 2 changed files with 46 additions and 0 deletions.
diff --git a/src/main/java/com/google/devtools/build/lib/bazel/rules/BazelRuleClassProvider.java b/src/main/java/com/google/devtools/build/lib/bazel/rules/BazelRuleClassProvider.java
@@ -156,6 +156,15 @@ public static class StrictActionEnvOptions extends FragmentOptions {
                 + "can prevent cross-user caching if a shared cache is used.")
     public boolean useStrictActionEnv;
 
+    @Option(
+        name = "incompatible_utf8_locale_in_action_env",
+        defaultValue = "true",
+        documentationCategory = OptionDocumentationCategory.UNCATEGORIZED,
+        effectTags = {OptionEffectTag.LOADING_AND_ANALYSIS},
+        metadataTags = {OptionMetadataTag.INCOMPATIBLE_CHANGE}
+    )
+    public boolean setUtf8LocaleInActionEnv;
+
     @Override
     public StrictActionEnvOptions getHost() {
       StrictActionEnvOptions host = (StrictActionEnvOptions) getDefault();
@@ -184,6 +193,8 @@ public StrictActionEnvConfiguration(BuildOptions buildOptions) {}
   public static final Function<BuildOptions, ActionEnvironment> SHELL_ACTION_ENV =
       (BuildOptions options) -> {
         boolean strictActionEnv = options.get(StrictActionEnvOptions.class).useStrictActionEnv;
+        boolean utf8LocaleInActionEnv = options.get(
+            StrictActionEnvOptions.class).setUtf8LocaleInActionEnv;
         OS os = OS.getCurrent();
         PathFragment shellExecutable = SHELL_EXECUTABLE.apply(options);
         TreeMap<String, String> env = new TreeMap<>();
@@ -211,6 +222,13 @@ public StrictActionEnvConfiguration(BuildOptions buildOptions) {}
           env.put("PATH", null);
         }
 
+        if (utf8LocaleInActionEnv) {
+          // Without setting this variable, tools (e.g. the JVM) may fall back to supporting only
+          // ASCII file paths and/or contents. The semantically ideal value would be "C.UTF-8", but
+          // this locale is not universally supported.
+          env.put("LC_CTYPE", "en_US.UTF-8");
+        }
+
         // Shell environment variables specified via options take precedence over the
         // ones inherited from the fragments. In the long run, these fragments will
         // be replaced by appropriate default rc files anyway.

diff --git a/src/test/shell/bazel/unicode_filenames_test.sh b/src/test/shell/bazel/unicode_filenames_test.sh
@@ -206,4 +206,32 @@ function test_utf8_source_artifact_in_bep() {
   expect_log '"name":"pkg/srcs/ünïcödë fïlë.txt"'
 }
 
+function test_utf8_filename_in_java_test() {
+  touch WORKSPACE
+  mkdir pkg
+
+  cat >pkg/BUILD <<'EOF'
+java_test(
+    name = "Test",
+    srcs = ["Test.java"],
+    main_class = "Test",
+    use_testrunner = False,
+)
+EOF
+
+  cat >pkg/Test.java <<'EOF'
+import java.nio.file.Files;
+import java.io.IOException;
+
+class Test {
+    public static void main(String[] args) throws IOException {
+        Files.createTempFile("æøå", null);
+    }
+}
+EOF
+
+  bazel test //pkg:Test --incompatible_utf8_locale_in_action_env \
+      --test_output=errors 2>$TEST_log || fail "Test should pass"
+}
+
 run_suite "Tests for handling of Unicode filenames"