[refactor] Remove WebWidget::applyReplacementRange

third_party/WebKit/Source/web/WebFrameWidgetImpl.cpp

Index: third_party/WebKit/Source/web/WebFrameWidgetImpl.cpp
diff --git a/third_party/WebKit/Source/web/WebFrameWidgetImpl.cpp b/third_party/WebKit/Source/web/WebFrameWidgetImpl.cpp
index b015b9b24c3c74bcf88b809e27ac3f92d7c83155..e76461f35025a7a6b7260a18bc2abcb945fb213c 100644
--- a/third_party/WebKit/Source/web/WebFrameWidgetImpl.cpp
+++ b/third_party/WebKit/Source/web/WebFrameWidgetImpl.cpp
@@ -705,16 +705,6 @@ bool WebFrameWidgetImpl::getCompositionCharacterBounds(
   return true;
 }
 
-// TODO(ekaramad):This method is almost duplicated in WebViewImpl as well. This
-// code needs to be refactored  (http://crbug.com/629721).
-void WebFrameWidgetImpl::applyReplacementRange(const WebRange& range) {
-  if (LocalFrame* frame = focusedLocalFrameInWidget()) {

[dcheng, 2017/02/07 07:58:42]

Is the reason it's OK to lose this call is because we're relying on the messages
being correctly routed by IPC now?

[EhsanK, 2017/02/07 17:14:47]

Yes. The IME IPC from RenderWidget of local root ends up at the focused frame's
WebInputMethodController. So we should be fine.
Please see this:
https://cs.chromium.org/chromium/src/content/renderer/render_widget.cc?rcl=41...

And this:
https://cs.chromium.org/chromium/src/third_party/WebKit/Source/web/WebFrameWi...

-    // TODO(dglazkov): Going from LocalFrame to WebLocalFrameImpl seems
-    // silly. What is going on here?
-    WebLocalFrameImpl::fromFrame(frame)->selectRange(range);
-  }
-}
-
 void WebFrameWidgetImpl::setRemoteViewportIntersection(
     const WebRect& viewportIntersection) {
   // Remote viewports are only applicable to local frames with remote ancestors.