fix know issues in RenderText

ui/gfx/render_text.cc

Index: ui/gfx/render_text.cc
===================================================================
--- ui/gfx/render_text.cc	(revision 100008)
+++ ui/gfx/render_text.cc	(working copy)
@@ -247,6 +247,12 @@
     sel.set_caret_pos(end.caret_pos());
     sel.set_caret_placement(end.caret_placement());
   }
+
+  if (!IsCursorablePosition(sel.selection_start()) ||
+      !IsCursorablePosition(sel.selection_end()) ||
+      !IsCursorablePosition(sel.caret_pos()))
+    return false;
+
   bool changed = !sel.Equals(selection_model_);
   SetSelectionModel(sel);
   return changed;
@@ -281,41 +287,28 @@
   SetSelectionModel(sel);
 }
 
-// TODO(xji): it does not work for languages do not use space as word breaker,
-// such as Chinese. Should use BreakIterator.
 void RenderText::SelectWord() {
-  size_t selection_start = GetSelectionStart();
   size_t cursor_position = GetCursorPosition();
-  // First we setup selection_start_ and selection_end_. There are so many cases
-  // because we try to emulate what select-word looks like in a gtk textfield.
-  // See associated testcase for different cases.
-  if (cursor_position > 0 && cursor_position < text().length()) {
-    if (u_isalnum(text()[cursor_position])) {
-      selection_start = cursor_position;
-      cursor_position++;
-    } else
-      selection_start = cursor_position - 1;
-  } else if (cursor_position == 0) {
-    selection_start = cursor_position;
-    if (text().length() > 0)
-      cursor_position++;
-  } else {
-    selection_start = cursor_position - 1;
-  }
 
-  // Now we move selection_start_ to beginning of selection. Selection boundary
-  // is defined as the position where we have alpha-num character on one side
-  // and non-alpha-num char on the other side.
-  for (; selection_start > 0; selection_start--) {
-    if (IsPositionAtWordSelectionBoundary(selection_start))
+  base::i18n::BreakIterator iter(text(), base::i18n::BreakIterator::BREAK_WORD);
+  bool success = iter.Init();
+  DCHECK(success);
+  if (!success)
+    return;
+
+  size_t selection_start = cursor_position;
+  for (; selection_start != 0; --selection_start) {
+    if (iter.IsStartOfWord(selection_start) ||
+        iter.IsEndOfWord(selection_start))
       break;
   }
 
-  // Now we move selection_end_ to end of selection. Selection boundary
-  // is defined as the position where we have alpha-num character on one side
-  // and non-alpha-num char on the other side.
-  for (; cursor_position < text().length(); cursor_position++) {
-    if (IsPositionAtWordSelectionBoundary(cursor_position))
+  if (selection_start == cursor_position)
+    ++cursor_position;
+
+  for (; cursor_position < text().length(); ++cursor_position) {
+    if (iter.IsEndOfWord(cursor_position) ||
+        iter.IsStartOfWord(cursor_position))
       break;
   }
 
@@ -462,6 +455,12 @@
   return cursor_bounds_;
 }
 
+size_t RenderText::GetIndexOfNextGrapheme(size_t position) {
+  // TODO(xji): Handle complex script.
+  return std::min(static_cast<long>(position + 1),

[msw, 2011/09/14 02:43:48]

Both of these static casts have lint errors:
"Use int16/int64/etc, rather than the C type long"
If it's an easy fix, please also fix RenderText::GetIndexOfPreviousGrapheme().

Or, you could take this opportunity to merge the win/linux impls that just call
GetIndexOfAdjacentGrapheme (with true/false and left/right) into these base
impls. I think you can make RenderText::GetIndexOfAdjacentGrapheme pure virtual
and just leave the implementations to RenderText[Win/Linux].

[xji, 2011/09/15 23:38:13]

Merged win/linux impls.

+                  static_cast<long>(text_.length()));
+}
+
 RenderText::RenderText()
     : text_(),
       selection_model_(),
@@ -575,6 +574,10 @@
   return std::vector<Rect>(1, rect);
 }
 
+bool RenderText::IsCursorablePosition(size_t position) {
+  return true;

[msw, 2011/09/14 02:43:48]

I think this can potentially be pure virtual, and leave the impls to win/linux.

[xji, 2011/09/15 23:38:13]

Done.

+}
+
 void RenderText::ApplyCompositionAndSelectionStyles(
     StyleRanges* style_ranges) const {
   // TODO(msw): This pattern ought to be reconsidered; what about composition
@@ -632,15 +635,12 @@
   SelectionModel::CaretPlacement placement = (caret_pos == cursor) ?
       SelectionModel::LEADING : SelectionModel::TRAILING;
   size_t selection_start = select ? GetSelectionStart() : cursor;
-  SelectionModel sel(selection_start, cursor, caret_pos, placement);
-  SetSelectionModel(sel);
+  if (IsCursorablePosition(cursor)) {
+    SelectionModel sel(selection_start, cursor, caret_pos, placement);
+    SetSelectionModel(sel);
+  }
 }
 
-bool RenderText::IsPositionAtWordSelectionBoundary(size_t pos) {
-  return pos == 0 || (u_isalnum(text()[pos - 1]) && !u_isalnum(text()[pos])) ||
-      (!u_isalnum(text()[pos - 1]) && u_isalnum(text()[pos]));
-}
-
 void RenderText::UpdateCachedBoundsAndOffset() {
   if (cached_bounds_and_offset_valid_)
     return;