[icu] Fix LookupAndValidateUnicodeExtensions() for ICU 77

kojiishi · V8 LUCI CQ · commit 97e76199e83e · 2025-11-12T01:45:56.000-08:00
ICU 77 `getKeywordValue()` changed to return `U_STRING_NOT_TERMINATED_WARNING` when the value is exactly the same size as the given buffer size. In this case, the buffer is not null-terminated[1]. This patch fixes `LookupAndValidateUnicodeExtensions()` for this change. [1] https://unicode-org.github.io/icu/userguide/strings/#using-c-strings-nul-terminated-vs-length-parameters Bug: 458186954 Change-Id: Iaff890f93cc392d4deb71ab881ab4bcce97f5b34 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/7137918 Auto-Submit: Koji Ishii <kojii@chromium.org> Commit-Queue: Jakob Linke <jgruber@chromium.org> Reviewed-by: Jakob Linke <jgruber@chromium.org> Reviewed-by: Michael Lippautz <mlippautz@chromium.org> Cr-Commit-Position: refs/heads/main@{#103662}
diff --git a/src/objects/intl-objects.cc b/src/objects/intl-objects.cc
@@ -2445,7 +2445,7 @@ std::map<std::string, std::string> LookupAndValidateUnicodeExtensions(
   if (U_FAILURE(status)) return extensions;
 
   if (!keywords) return extensions;
-  char value[ULOC_FULLNAME_CAPACITY];
+  char value[ULOC_FULLNAME_CAPACITY + 1];
 
   int32_t length;
   status = U_ZERO_ERROR;
@@ -2459,7 +2459,8 @@ std::map<std::string, std::string> LookupAndValidateUnicodeExtensions(
       continue;
     }
 
-    icu_locale->getKeywordValue(keyword, value, ULOC_FULLNAME_CAPACITY, status);
+    const int32_t value_len = icu_locale->getKeywordValue(
+        keyword, value, ULOC_FULLNAME_CAPACITY, status);
 
     // Ignore failures in ICU and skip to the next keyword.
     //
@@ -2468,6 +2469,9 @@ std::map<std::string, std::string> LookupAndValidateUnicodeExtensions(
       status = U_ZERO_ERROR;
       continue;
     }
+    // Ensure the `value` is null-terminated even when the `status` is
+    // `U_STRING_NOT_TERMINATED_WARNING`.
+    value[value_len] = '\0';
 
     const char* bcp47_key = uloc_toUnicodeLocaleKey(keyword);
 
diff --git a/test/intl/date-format/long-locale.js b/test/intl/date-format/long-locale.js
@@ -0,0 +1,15 @@
+// Copyright 2025 the V8 project authors. All rights reserved.
+// Use of this source code is governed by a BSD-style license that can be
+// found in the LICENSE file.
+
+let locale = "zh-TW-u-ca-chinese";
+for (let i = 0; i < 300; ++i) {
+  try {
+    Intl.DateTimeFormat(locale);
+  } catch (e) {
+    // "RangeError: Invalid language tag", for locales ending with "-", or
+    // sub-tags of one character, are not relevant to this test.
+  }
+  locale += (i % 5) ? "a" : "-";
+}
+// Pass if this test doesn't crash.
