From e61e2b17851818c993d4fb509b4076616ea8292b Mon Sep 17 00:00:00 2001
From: eroman <eroman@chromium.org>
Date: Tue, 13 Oct 2015 16:39:22 -0700
Subject: [PATCH] Make base::IsUnicodeWhitespace('\0') return false.

BUG=542759

Review URL: https://codereview.chromium.org/1399253004

Cr-Commit-Position: refs/heads/master@{#353905}
---
 base/strings/string_util.cc          |  9 +++++++++
 base/strings/string_util.h           |  4 +---
 base/strings/string_util_unittest.cc | 20 ++++++++++++++++++++
 3 files changed, 30 insertions(+), 3 deletions(-)

diff --git a/base/strings/string_util.cc b/base/strings/string_util.cc
index d7a1d54c72210a..a978891d57e291 100644
--- a/base/strings/string_util.cc
+++ b/base/strings/string_util.cc
@@ -674,6 +674,15 @@ char HexDigitToInt(wchar_t c) {
   return 0;
 }
 
+bool IsUnicodeWhitespace(wchar_t c) {
+  // kWhitespaceWide is a NULL-terminated string
+  for (const wchar_t* cur = kWhitespaceWide; *cur; ++cur) {
+    if (*cur == c)
+      return true;
+  }
+  return false;
+}
+
 static const char* const kByteStringsUnlocalized[] = {
   " B",
   " kB",
diff --git a/base/strings/string_util.h b/base/strings/string_util.h
index 169726bca5a128..2578569ce014a4 100644
--- a/base/strings/string_util.h
+++ b/base/strings/string_util.h
@@ -358,9 +358,7 @@ inline bool IsHexDigit(Char c) {
 BASE_EXPORT char HexDigitToInt(wchar_t c);
 
 // Returns true if it's a Unicode whitespace character.
-inline bool IsUnicodeWhitespace(wchar_t c) {
-  return wcschr(base::kWhitespaceWide, c) != NULL;
-}
+BASE_EXPORT bool IsUnicodeWhitespace(wchar_t c);
 
 // Return a byte string in human-readable format with a unit suffix. Not
 // appropriate for use in any UI; use of FormatBytes and friends in ui/base is
diff --git a/base/strings/string_util_unittest.cc b/base/strings/string_util_unittest.cc
index 187e49e6698160..765ba83508c6a1 100644
--- a/base/strings/string_util_unittest.cc
+++ b/base/strings/string_util_unittest.cc
@@ -1080,6 +1080,26 @@ TEST(StringUtilTest, EqualsCaseInsensitiveASCII) {
   EXPECT_FALSE(EqualsCaseInsensitiveASCII("Asdf", "aSDFz"));
 }
 
+TEST(StringUtilTest, IsUnicodeWhitespace) {
+  // NOT unicode white space.
+  EXPECT_FALSE(IsUnicodeWhitespace(L'\0'));
+  EXPECT_FALSE(IsUnicodeWhitespace(L'A'));
+  EXPECT_FALSE(IsUnicodeWhitespace(L'0'));
+  EXPECT_FALSE(IsUnicodeWhitespace(L'.'));
+  EXPECT_FALSE(IsUnicodeWhitespace(L';'));
+  EXPECT_FALSE(IsUnicodeWhitespace(L'\x4100'));
+
+  // Actual unicode whitespace.
+  EXPECT_TRUE(IsUnicodeWhitespace(L' '));
+  EXPECT_TRUE(IsUnicodeWhitespace(L'\xa0'));
+  EXPECT_TRUE(IsUnicodeWhitespace(L'\x3000'));
+  EXPECT_TRUE(IsUnicodeWhitespace(L'\t'));
+  EXPECT_TRUE(IsUnicodeWhitespace(L'\r'));
+  EXPECT_TRUE(IsUnicodeWhitespace(L'\v'));
+  EXPECT_TRUE(IsUnicodeWhitespace(L'\f'));
+  EXPECT_TRUE(IsUnicodeWhitespace(L'\n'));
+}
+
 class WriteIntoTest : public testing::Test {
  protected:
   static void WritesCorrectly(size_t num_chars) {