Index: base/utf_offset_string_conversions.cc |
diff --git a/base/utf_offset_string_conversions.cc b/base/utf_offset_string_conversions.cc |
index 8405303fd9d95bd5fb8fe3ef0a0bcdb70102cd80..3e468b554ab29a89114f756c5ed8a4666fb40a04 100644 |
--- a/base/utf_offset_string_conversions.cc |
+++ b/base/utf_offset_string_conversions.cc |
@@ -1,4 +1,4 @@ |
-// Copyright (c) 2011 The Chromium Authors. All rights reserved. |
+// Copyright (c) 2012 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. |
@@ -6,8 +6,10 @@ |
#include <algorithm> |
+#include "base/logging.h" |
#include "base/memory/scoped_ptr.h" |
#include "base/string_piece.h" |
+#include "base/third_party/icu/icu_utf.h" |
#include "base/utf_string_conversion_utils.h" |
using base::PrepareForUTF16Or32Output; |
@@ -15,6 +17,8 @@ using base::PrepareForUTF8Output; |
using base::ReadUnicodeCharacter; |
using base::WriteUnicodeCharacter; |
+namespace { |
+ |
// Converts the given source Unicode character type to the given destination |
// Unicode character type as a STL string. The given input buffer and size |
// determine the source, and the given output STL string will be replaced by |
@@ -56,6 +60,41 @@ bool ConvertUnicode(const SrcChar* src, |
return success; |
} |
+} // namespace |
+ |
+bool IsValidCodePointIndex(const string16& s, size_t index) { |
+ return index == 0 || index == s.length() || |
+ !(CBU16_IS_TRAIL(s[index]) && CBU16_IS_LEAD(s[index - 1])); |
+} |
+ |
+ptrdiff_t Utf16IndexToOffset(const string16& s, size_t base, size_t pos) { |
+ DCHECK_LE(base, s.length()); |
+ DCHECK_LE(pos, s.length()); |
+ ptrdiff_t delta = 0; |
+ while (base < pos) |
+ delta += IsValidCodePointIndex(s, base++) ? 1 : 0; |
+ while (pos < base) |
+ delta -= IsValidCodePointIndex(s, pos++) ? 1 : 0; |
+ return delta; |
+} |
+ |
+size_t Utf16OffsetToIndex(const string16& s, size_t pos, ptrdiff_t offset) { |
+ DCHECK_LE(pos, s.length()); |
+ while (offset > 0 && pos < s.length()) |
+ offset -= IsValidCodePointIndex(s, pos++) ? 1 : 0; |
+ while (offset < 0 && pos > 0) |
+ offset += IsValidCodePointIndex(s, --pos) ? 1 : 0; |
+ // If offset != 0 then we ran off the edge of the string, which shouldn't |
+ // happen but is handled anyway for safety. |
+ DCHECK_EQ(offset, 0); |
+ // Since the second half of a surrogate pair has "length" zero, there is an |
+ // ambiguity in the returned position. Resolve it by always returning a valid |
+ // index. |
+ if (!IsValidCodePointIndex(s, pos)) |
+ ++pos; |
+ return pos; |
+} |
+ |
bool UTF8ToUTF16AndAdjustOffset(const char* src, |
size_t src_len, |
string16* output, |