Fix character ranges in case insensitive regexp

R=jgruber@chromium.org Bug: chromium:971383 Change-Id: I39d26a63c0735f595a809959c06cb2ac1c141451 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1648098 Commit-Queue: Frank Tang <ftang@chromium.org> Auto-Submit: Yang Guo <yangguo@chromium.org> Reviewed-by: Frank Tang <ftang@chromium.org> Cr-Commit-Position: refs/heads/master@{#62044}
v8 · Jun 7, 2019 · 9bcacf6 · 9bcacf6
1 parent ac30897
commit 9bcacf6
Show file tree

Hide file tree

Showing 4 changed files with 28 additions and 2 deletions.
diff --git a/src/d8/d8.cc b/src/d8/d8.cc
@@ -48,6 +48,10 @@
 #include "src/utils/utils.h"
 #include "src/wasm/wasm-engine.h"
 
+#ifdef V8_INTL_SUPPORT
+#include "unicode/locid.h"
+#endif  // V8_INTL_SUPPORT
+
 #if !defined(_WIN32) && !defined(_WIN64)
 #include <unistd.h>  // NOLINT
 #else
@@ -2826,6 +2830,9 @@ bool Shell::SetOptions(int argc, char* argv[]) {
     } else if (strncmp(argv[i], "--icu-data-file=", 16) == 0) {
       options.icu_data_file = argv[i] + 16;
       argv[i] = nullptr;
+    } else if (strncmp(argv[i], "--icu-locale=", 13) == 0) {
+      options.icu_locale = argv[i] + 13;
+      argv[i] = nullptr;
 #ifdef V8_USE_EXTERNAL_STARTUP_DATA
     } else if (strncmp(argv[i], "--natives_blob=", 15) == 0) {
       options.natives_blob = argv[i] + 15;
@@ -3325,8 +3332,17 @@ void Shell::CleanupWorkers() {
 int Shell::Main(int argc, char* argv[]) {
   v8::base::EnsureConsoleOutput();
   if (!SetOptions(argc, argv)) return 1;
+
   v8::V8::InitializeICUDefaultLocation(argv[0], options.icu_data_file);
 
+#ifdef V8_INTL_SUPPORT
+  if (options.icu_locale != nullptr) {
+    icu::Locale locale(options.icu_locale);
+    UErrorCode error_code = U_ZERO_ERROR;
+    icu::Locale::setDefault(locale, error_code);
+  }
+#endif  // V8_INTL_SUPPORT
+
   v8::platform::InProcessStackDumping in_process_stack_dumping =
       options.disable_in_process_stack_traces
           ? v8::platform::InProcessStackDumping::kDisabled

diff --git a/src/d8/d8.h b/src/d8/d8.h
@@ -329,6 +329,7 @@ class ShellOptions {
   CodeCacheOptions code_cache_options = CodeCacheOptions::kNoProduceCache;
   SourceGroup* isolate_sources = nullptr;
   const char* icu_data_file = nullptr;
+  const char* icu_locale = nullptr;
   const char* natives_blob = nullptr;
   const char* snapshot_blob = nullptr;
   bool trace_enabled = false;

diff --git a/src/regexp/jsregexp.cc b/src/regexp/jsregexp.cc
@@ -32,6 +32,7 @@
 #include "src/zone/zone-list-inl.h"
 
 #ifdef V8_INTL_SUPPORT
+#include "unicode/locid.h"
 #include "unicode/uniset.h"
 #include "unicode/utypes.h"
 #endif  // V8_INTL_SUPPORT
@@ -5948,17 +5949,18 @@ void CharacterRange::AddCaseEquivalents(Isolate* isolate, Zone* zone,
       if (top > String::kMaxOneByteCharCode) top = String::kMaxOneByteCharCode;
     }
     already_added.add(bottom, top);
+    icu::Locale locale = icu::Locale::getRoot();
     while (bottom <= top) {
       icu::UnicodeString upper(bottom);
-      upper.toUpper();
+      upper.toUpper(locale);
       icu::UnicodeSet expanded(bottom, bottom);
       expanded.closeOver(USET_CASE_INSENSITIVE);
       for (int32_t i = 0; i < expanded.getRangeCount(); i++) {
         UChar32 start = expanded.getRangeStart(i);
         UChar32 end = expanded.getRangeEnd(i);
         while (start <= end) {
           icu::UnicodeString upper2(start);
-          upper2.toUpper();
+          upper2.toUpper(locale);
           // Only add if the upper case are the same.
           if (upper[0] == upper2[0]) {
             others.add(start);

diff --git a/test/mjsunit/regress/regress-crbug-971383.js b/test/mjsunit/regress/regress-crbug-971383.js
@@ -0,0 +1,7 @@
+// Copyright 2019 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.
+
+// Flags: --icu-locale=tr
+
+assertEquals(["HIJK"], "HIJK".match(/[a-z]+/gi));