Implement fast path in UTF8ToUTF16 for pure ASCII strings

base/strings/string_util_unittest.cc

Index: base/strings/string_util_unittest.cc
diff --git a/base/strings/string_util_unittest.cc b/base/strings/string_util_unittest.cc
index 0928c0f9d2237e6339b5489017888093022b9911..16448467901e442c11966d92c3dae421f9bc48c3 100644
--- a/base/strings/string_util_unittest.cc
+++ b/base/strings/string_util_unittest.cc
@@ -386,6 +386,55 @@ TEST(StringUtilTest, IsStringUTF8) {
   EXPECT_FALSE(IsStringUTF8("embedded\xc0\x80U+0000"));
 }
 
+TEST(StringUtilTest, IsStringASCII) {
+  static char char_ascii[] =
+      "0123456789ABCDEF0123456789ABCDEF0123456789ABCDEF";
+  static char16 char16_ascii[] = {
+      '0', '1', '2', '3', '4', '5', '6', '7', '8', '9', '0', 'A',
+      'B', 'C', 'D', 'E', 'F', '0', '1', '2', '3', '4', '5', '6',
+      '7', '8', '9', '0', 'A', 'B', 'C', 'D', 'E', 'F', 0 };
+
+  // Test a variety of the fragment start positions and lengths in order to make
+  // sure that bit masking in IsStringASCII works correctly.
+  // Also, test that a non-ASCII character will be detected regardless of its
+  // position inside the string.
+  {
+    const size_t string_length = arraysize(char_ascii) - 1;
+    for (size_t offset = 0; offset < 8; ++offset) {
+      for (size_t len = 0, max_len = string_length - offset; len < max_len;
+           ++len) {
+        EXPECT_TRUE(IsStringASCII(StringPiece(char_ascii + offset, len)));
+        for (size_t char_pos = offset; char_pos < len; ++char_pos) {
+          char_ascii[char_pos] |= '\x80';
+          EXPECT_FALSE(IsStringASCII(StringPiece(char_ascii + offset, len)));
+          char_ascii[char_pos] &= ~'\x80';
+        }
+      }
+    }
+  }
+
+  {
+    const size_t string_length = arraysize(char16_ascii) - 1;
+    for (size_t offset = 0; offset < 4; ++offset) {
+      for (size_t len = 0, max_len = string_length - offset; len < max_len;
+           ++len) {
+        EXPECT_TRUE(IsStringASCII(StringPiece16(char16_ascii + offset, len)));
+        for (size_t char_pos = offset; char_pos < len; ++char_pos) {
+          char16_ascii[char_pos] |= 0x80;
+          EXPECT_FALSE(
+              IsStringASCII(StringPiece16(char16_ascii + offset, len)));
+          char16_ascii[char_pos] &= ~0x80;
+          // Also test when the upper half is non-zero.
+          char16_ascii[char_pos] |= 0x100;
+          EXPECT_FALSE(
+              IsStringASCII(StringPiece16(char16_ascii + offset, len)));
+          char16_ascii[char_pos] &= ~0x100;
+        }
+      }
+    }
+  }
+}
+
 TEST(StringUtilTest, ConvertASCII) {
   static const char* char_cases[] = {
     "Google Video",