Stop combining text runs which are connected by 'COMMON' blocks.

ui/gfx/text_elider.cc

 // 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.
 //
 // This file implements utility functions for eliding and formatting UI text.
 //
 // Note that several of the functions declared in text_elider.h are implemented
 // in this file using helper classes in an unnamed namespace.
 
 #include "ui/gfx/text_elider.h"
 
 #include <stdint.h>
 
 #include <string>
 #include <vector>
 
 #include "base/files/file_path.h"
 #include "base/i18n/break_iterator.h"
 #include "base/i18n/char_iterator.h"
 #include "base/i18n/rtl.h"
 #include "base/memory/scoped_ptr.h"
 #include "base/numerics/safe_conversions.h"
 #include "base/strings/string_split.h"
 #include "base/strings/string_util.h"
 #include "base/strings/sys_string_conversions.h"
 #include "base/strings/utf_string_conversions.h"
 #include "third_party/icu/source/common/unicode/rbbi.h"
 #include "third_party/icu/source/common/unicode/uchar.h"

[msw, 2015/06/03 00:32:03]

nit: can we remove some includes, like this one, now?

[xdai1, 2015/06/06 00:05:19]

Done.

 #include "third_party/icu/source/common/unicode/uloc.h"
 #include "third_party/icu/source/common/unicode/umachine.h"
 #include "third_party/icu/source/common/unicode/utf16.h"

[msw, 2015/06/03 00:32:03]

ditto

[xdai1, 2015/06/06 00:05:19]

Done.

 #include "ui/gfx/font_list.h"
 #include "ui/gfx/geometry/rect_conversions.h"
 #include "ui/gfx/render_text.h"
 #include "ui/gfx/text_utils.h"
 
 using base::ASCIIToUTF16;
 using base::UTF8ToUTF16;
 using base::WideToUTF16;
 
 namespace gfx {
@@ skipping 55 matching lines @@
 
   // Fit the username in the remaining width (at this point the elided username
   // is guaranteed to fit with at least one character remaining given all the
   // precautions taken earlier).
   available_pixel_width -= GetStringWidthF(domain, font_list);
   username = ElideText(username, font_list, available_pixel_width, ELIDE_TAIL);
   return username + kAtSignUTF16 + domain;
 }
 #endif
 
-// Returns true if the code point |c| is a combining mark character in Unicode.
-bool CharIsMark(UChar32 c) {
-  int8_t char_type = u_charType(c);
-  return char_type == U_NON_SPACING_MARK || char_type == U_ENCLOSING_MARK ||
-         char_type == U_COMBINING_SPACING_MARK;
-}
-
-// Gets the code point of |str| at the given code unit position |index|. If
-// |index| is a surrogate code unit, returns the whole code point (unless the
-// code unit is unpaired, in which case it just returns the surrogate value).
-UChar32 GetCodePointAt(const base::string16& str, size_t index) {
-  UChar32 c;
-  U16_GET(str.data(), 0, index, str.size(), c);
-  return c;
-}
-
 }  // namespace
 
 // U+2026 in utf8
 const char kEllipsis[] = "\xE2\x80\xA6";
 const base::char16 kEllipsisUTF16[] = { 0x2026, 0 };
 const base::char16 kForwardSlash = '/';
 
 StringSlicer::StringSlicer(const base::string16& text,
                            const base::string16& ellipsis,
                            bool elide_in_middle,
                            bool elide_at_beginning)
     : text_(text),
       ellipsis_(ellipsis),
       elide_in_middle_(elide_in_middle),
       elide_at_beginning_(elide_at_beginning) {
 }
 
 base::string16 StringSlicer::CutString(size_t length,
                                        bool insert_ellipsis) const {
   const base::string16 ellipsis_text = insert_ellipsis ? ellipsis_
                                                        : base::string16();
 
   if (elide_at_beginning_)
     return ellipsis_text +
-           text_.substr(FindValidBoundaryBefore(text_.length() - length));
+           text_.substr(
+               FindValidBoundaryBefore(text_, text_.length() - length));
 
   if (!elide_in_middle_)
-    return text_.substr(0, FindValidBoundaryBefore(length)) + ellipsis_text;
+    return text_.substr(0, FindValidBoundaryBefore(text_, length)) +
+           ellipsis_text;
 
   // We put the extra character, if any, before the cut.
   const size_t half_length = length / 2;
-  const size_t prefix_length = FindValidBoundaryBefore(length - half_length);
+  const size_t prefix_length =
+      FindValidBoundaryBefore(text_, length - half_length);
   const size_t suffix_start =
-      FindValidBoundaryAfter(text_.length() - half_length);
+      FindValidBoundaryAfter(text_, text_.length() - half_length);
   return text_.substr(0, prefix_length) + ellipsis_text +
          text_.substr(suffix_start);
 }
 
-size_t StringSlicer::FindValidBoundaryBefore(size_t index) const {
-  size_t length = text_.length();
-  DCHECK_LE(index, length);
-  if (index == length)
-    return index;
-
-  // If |index| straddles a combining character sequence, go back until we find
-  // a base character.
-  while (index > 0 && CharIsMark(GetCodePointAt(text_, index)))
-    --index;
-
-  // If |index| straddles a UTF-16 surrogate pair, go back.
-  U16_SET_CP_START(text_.data(), 0, index);
-  return index;
-}
-
-size_t StringSlicer::FindValidBoundaryAfter(size_t index) const {
-  DCHECK_LE(index, text_.length());
-  if (index == text_.length())
-    return index;
-
-  int32_t text_index = base::checked_cast<int32_t>(index);
-  int32_t text_length = base::checked_cast<int32_t>(text_.length());
-
-  // If |index| straddles a combining character sequence, go forward until we
-  // find a base character.
-  while (text_index < text_length &&
-         CharIsMark(GetCodePointAt(text_, text_index))) {
-    ++text_index;
-  }
-
-  // If |index| straddles a UTF-16 surrogate pair, go forward.
-  U16_SET_CP_LIMIT(text_.data(), 0, text_index, text_length);
-  return static_cast<size_t>(text_index);
-}
-
 base::string16 ElideFilename(const base::FilePath& filename,
                              const FontList& font_list,
                              float available_pixel_width) {
 #if defined(OS_WIN)
   base::string16 filename_utf16 = filename.value();
   base::string16 extension = filename.Extension();
   base::string16 rootname = filename.BaseName().RemoveExtension().value();
 #elif defined(OS_POSIX)
   base::string16 filename_utf16 = WideToUTF16(base::SysNativeMBToWide(
       filename.value()));
@@ skipping 654 matching lines @@
     index = char_iterator.getIndex();
   } else {
     // String has leading whitespace, return the elide string.
     return kElideString;
   }
 
   return string.substr(0, index) + kElideString;
 }
 
 }  // namespace gfx