Move unusual character block logic to a helper function

BUG=
NOTRY=true
R=msw

Review URL: https://codereview.chromium.org/324983002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@276776 0039d316-1c4b-4281-b951-d872f2087c98

ui/gfx/render_text_harfbuzz.cc

@@ -281,6 +281,26 @@ bool IsUnusualBlockCode(UBlockCode block_code) {
          block_code == UBLOCK_MISCELLANEOUS_SYMBOLS;
 }
 
+// Returns the index of the first unusual character after a usual character or
+// vice versa. Unusual characters are defined by |IsUnusualBlockCode|.
+size_t FindUnusualCharacter(const base::string16& text,
+                            size_t run_start,
+                            size_t run_break) {
+  const int32 run_length = static_cast<int32>(run_break - run_start);
+  base::i18n::UTF16CharIterator iter(text.c_str() + run_start,
+                                     run_length);
+  const UBlockCode first_block_code = ublock_getCode(iter.get());
+  const bool first_block_unusual = IsUnusualBlockCode(first_block_code);
+  while (iter.Advance() && iter.array_pos() < run_length) {
+    const UBlockCode current_block_code = ublock_getCode(iter.get());
+    if (current_block_code != first_block_code &&
+        (first_block_unusual || IsUnusualBlockCode(current_block_code))) {
+      return run_start + iter.array_pos();
+    }
+  }
+  return run_break;
+}
+
 // If the given scripts match, returns the one that isn't USCRIPT_COMMON or
 // USCRIPT_INHERITED, i.e. the more specific one. Otherwise returns
 // USCRIPT_INVALID_CODE.
@@ -912,22 +932,8 @@ void RenderTextHarfBuzz::ItemizeText() {
     // Break runs adjacent to character substrings in certain code blocks.
     // This avoids using their fallback fonts for more characters than needed,
     // in cases like "\x25B6 Media Title", etc. http://crbug.com/278913
-    if (run_break > run->range.start()) {
-      const size_t run_start = run->range.start();
-      const int32 run_length = static_cast<int32>(run_break - run_start);
-      base::i18n::UTF16CharIterator iter(text.c_str() + run_start,
-                                         run_length);
-      const UBlockCode first_block_code = ublock_getCode(iter.get());
-      const bool first_block_unusual = IsUnusualBlockCode(first_block_code);
-      while (iter.Advance() && iter.array_pos() < run_length) {
-        const UBlockCode current_block_code = ublock_getCode(iter.get());
-        if (current_block_code != first_block_code &&
-            (first_block_unusual || IsUnusualBlockCode(current_block_code))) {
-          run_break = run_start + iter.array_pos();
-          break;
-        }
-      }
-    }
+    if (run_break > run->range.start())
+      run_break = FindUnusualCharacter(text, run->range.start(), run_break);
 
     DCHECK(IsValidCodePointIndex(text, run_break));
     style.UpdatePosition(LayoutIndexToTextIndex(run_break));

ui/gfx/render_text_harfbuzz.h

@@ -92,6 +92,7 @@ class GFX_EXPORT RenderTextHarfBuzz : public RenderText {
  private:
   friend class RenderTextTest;
   FRIEND_TEST_ALL_PREFIXES(RenderTextTest, HarfBuzz_RunDirection);
+  FRIEND_TEST_ALL_PREFIXES(RenderTextTest, HarfBuzz_BreakRunsByUnicodeBlocks);
 
   // Return the run index that contains the argument; or the length of the
   // |runs_| vector if argument exceeds the text length or width.

ui/gfx/render_text_unittest.cc

@@ -1903,7 +1903,6 @@ TEST_F(RenderTextTest, Multiline_Newline) {
   }
 }
 
-
 TEST_F(RenderTextTest, Win_BreakRunsByUnicodeBlocks) {
   scoped_ptr<RenderTextWin> render_text(
       static_cast<RenderTextWin*>(RenderText::CreateInstance()));
@@ -1974,7 +1973,6 @@ TEST_F(RenderTextTest, HarfBuzz_CharToGlyph) {
                 run.CharRangeToGlyphRange(Range(j, j + 1)));
     }
   }
-
 }
 
 TEST_F(RenderTextTest, HarfBuzz_RunDirection) {
@@ -1989,4 +1987,23 @@ TEST_F(RenderTextTest, HarfBuzz_RunDirection) {
   EXPECT_TRUE(render_text.runs_[2]->is_rtl);
 }
 
+TEST_F(RenderTextTest, HarfBuzz_BreakRunsByUnicodeBlocks) {
+  RenderTextHarfBuzz render_text;
+
+  // The '\x25B6' "play character" should break runs. http://crbug.com/278913
+  render_text.SetText(WideToUTF16(L"x\x25B6y"));
+  render_text.EnsureLayout();
+  ASSERT_EQ(3U, render_text.runs_.size());
+  EXPECT_EQ(Range(0, 1), render_text.runs_[0]->range);
+  EXPECT_EQ(Range(1, 2), render_text.runs_[1]->range);
+  EXPECT_EQ(Range(2, 3), render_text.runs_[2]->range);
+
+  render_text.SetText(WideToUTF16(L"x \x25B6 y"));
+  render_text.EnsureLayout();
+  ASSERT_EQ(3U, render_text.runs_.size());
+  EXPECT_EQ(Range(0, 2), render_text.runs_[0]->range);
+  EXPECT_EQ(Range(2, 3), render_text.runs_[1]->range);
+  EXPECT_EQ(Range(3, 5), render_text.runs_[2]->range);
+}
+
 }  // namespace gfx

ui/gfx/render_text_win.cc

@@ -264,6 +264,26 @@ bool IsUnusualBlockCode(const UBlockCode block_code) {
          block_code == UBLOCK_MISCELLANEOUS_SYMBOLS;
 }
 
+// Returns the index of the first unusual character after a usual character or
+// vice versa. Unusual characters are defined by |IsUnusualBlockCode|.
+size_t FindUnusualCharacter(const base::string16& text,
+                            size_t run_start,
+                            size_t run_break) {
+  const int32 run_length = static_cast<int32>(run_break - run_start);
+  base::i18n::UTF16CharIterator iter(text.c_str() + run_start,
+                                     run_length);
+  const UBlockCode first_block_code = ublock_getCode(iter.get());
+  const bool first_block_unusual = IsUnusualBlockCode(first_block_code);
+  while (iter.Advance() && iter.array_pos() < run_length) {
+    const UBlockCode current_block_code = ublock_getCode(iter.get());
+    if (current_block_code != first_block_code &&
+        (first_block_unusual || IsUnusualBlockCode(current_block_code))) {
+      return run_start + iter.array_pos();
+    }
+  }
+  return run_break;
+}
+
 }  // namespace
 
 namespace internal {
@@ -955,20 +975,8 @@ void RenderTextWin::ItemizeLogicalText() {
     // This avoids using their fallback fonts for more characters than needed,
     // in cases like "\x25B6 Media Title", etc. http://crbug.com/278913
     if (run_break > run->range.start()) {
-      const size_t run_start = run->range.start();
-      const int32 run_length = static_cast<int32>(run_break - run_start);
-      base::i18n::UTF16CharIterator iter(layout_text.c_str() + run_start,
-                                         run_length);
-      const UBlockCode first_block_code = ublock_getCode(iter.get());
-      const bool first_block_unusual = IsUnusualBlockCode(first_block_code);
-      while (iter.Advance() && iter.array_pos() < run_length) {
-        const UBlockCode current_block_code = ublock_getCode(iter.get());
-        if (current_block_code != first_block_code &&
-            (first_block_unusual || IsUnusualBlockCode(current_block_code))) {
-          run_break = run_start + iter.array_pos();
-          break;
-        }
-      }
+      run_break =
+          FindUnusualCharacter(layout_text, run->range.start(), run_break);
     }
 
     DCHECK(IsValidCodePointIndex(layout_text, run_break));