1 files changed, 86 insertions, 0 deletions
diff --git a/base/utf_string_conversion_utils.h b/base/utf_string_conversion_utils.h
new file mode 100644
index 0000000..a8a76c5
--- /dev/null
+++ b/base/utf_string_conversion_utils.h
@@ -0,0 +1,86 @@
+// Copyright (c) 2009 The Chromium Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style license that can be
+// found in the LICENSE file.
+
+#ifndef BASE_UTF_STRING_CONVERSION_UTILS_H_
+#define BASE_UTF_STRING_CONVERSION_UTILS_H_
+
+// This should only be used by the various UTF string conversion files.
+
+#include "base/string16.h"
+
+namespace base {
+
+inline bool IsValidCodepoint(uint32 code_point) {
+  // Excludes the surrogate code points ([0xD800, 0xDFFF]) and
+  // codepoints larger than 0x10FFFF (the highest codepoint allowed).
+  // Non-characters and unassigned codepoints are allowed.
+  return code_point < 0xD800u ||
+         (code_point >= 0xE000u && code_point <= 0x10FFFFu);
+}
+
+// ReadUnicodeCharacter --------------------------------------------------------
+
+// Reads a UTF-8 stream, placing the next code point into the given output
+// |*code_point|. |src| represents the entire string to read, and |*char_index|
+// is the character offset within the string to start reading at. |*char_index|
+// will be updated to index the last character read, such that incrementing it
+// (as in a for loop) will take the reader to the next character.
+//
+// Returns true on success. On false, |*code_point| will be invalid.
+bool ReadUnicodeCharacter(const char* src,
+                          int32 src_len,
+                          int32* char_index,
+                          uint32* code_point_out);
+
+// Reads a UTF-16 character. The usage is the same as the 8-bit version above.
+bool ReadUnicodeCharacter(const char16* src,
+                          int32 src_len,
+                          int32* char_index,
+                          uint32* code_point);
+
+#if defined(WCHAR_T_IS_UTF32)
+// Reads UTF-32 character. The usage is the same as the 8-bit version above.
+bool ReadUnicodeCharacter(const wchar_t* src,
+                          int32 src_len,
+                          int32* char_index,
+                          uint32* code_point);
+#endif  // defined(WCHAR_T_IS_UTF32)
+
+// WriteUnicodeCharacter -------------------------------------------------------
+
+// Appends a UTF-8 character to the given 8-bit string.  Returns the number of
+// bytes written.
+size_t WriteUnicodeCharacter(uint32 code_point, std::string* output);
+
+// Appends the given code point as a UTF-16 character to the given 16-bit
+// string.  Returns the number of 16-bit values written.
+size_t WriteUnicodeCharacter(uint32 code_point, string16* output);
+
+#if defined(WCHAR_T_IS_UTF32)
+// Appends the given UTF-32 character to the given 32-bit string.  Returns the
+// number of 32-bit values written.
+inline size_t WriteUnicodeCharacter(uint32 code_point, std::wstring* output) {
+  // This is the easy case, just append the character.
+  output->push_back(code_point);
+  return 1;
+}
+#endif  // defined(WCHAR_T_IS_UTF32)
+
+// Generalized Unicode converter -----------------------------------------------
+
+// Guesses the length of the output in UTF-8 in bytes, clears that output
+// string, and reserves that amount of space.  We assume that the input
+// character types are unsigned, which will be true for UTF-16 and -32 on our
+// systems.
+template<typename CHAR>
+void PrepareForUTF8Output(const CHAR* src, size_t src_len, std::string* output);
+
+// Prepares an output buffer (containing either UTF-16 or -32 data) given some
+// UTF-8 input that will be converted to it.  See PrepareForUTF8Output().
+template<typename STRING>
+void PrepareForUTF16Or32Output(const char* src, size_t src_len, STRING* output);
+
+}  // namespace base
+
+#endif  // BASE_UTF_STRING_CONVERSION_UTILS_H_