Thanks for working on this. This is pretty much ok. The only question I have is:
+void RenderWidgetHostView::UpdateCursor(const WebCursor& cursor) {
+ if (GetFormFactorHint() != FORM_FACTOR_DESKTOP) {
+ return; // Cursor only on desktop
+ }
+
Is this check really necessary? Is this done in other parts of the stack that process cursor changes? And do we know whether Chromium calls UpdateCursor() without mouse events?
+ QImage::Format format = QImage::Format_Invalid;
+ switch (cursor_info.custom_image.config()) {
+ case SkBitmap::kRGB_565_Config: format = QImage::Format_RGB16;
+ case SkBitmap::kARGB_4444_Config: format = QImage::Format_ARGB4444_Premultiplied;
+ case SkBitmap::kARGB_8888_Config: format = QImage::Format_ARGB32_Premultiplied;
+ default: ;
+ }
I was going to say that this bit is missing an indent for each of the case statements, but I've just noticed that this rule is already applied quite inconsistently in this file, so it doesn't matter for now :)
Thanks for working on this. This is pretty much ok. The only question I have is:
+void RenderWidgetHos tView:: UpdateCursor( const WebCursor& cursor) { int() != FORM_FACTOR_ DESKTOP) {
+ if (GetFormFactorH
+ return; // Cursor only on desktop
+ }
+
Is this check really necessary? Is this done in other parts of the stack that process cursor changes? And do we know whether Chromium calls UpdateCursor() without mouse events?
+ QImage::Format format = QImage: :Format_ Invalid; info.custom_ image.config( )) { :kRGB_565_ Config: format = QImage: :Format_ RGB16; :kARGB_ 4444_Config: format = QImage: :Format_ ARGB4444_ Premultiplied; :kARGB_ 8888_Config: format = QImage: :Format_ ARGB32_ Premultiplied;
+ switch (cursor_
+ case SkBitmap:
+ case SkBitmap:
+ case SkBitmap:
+ default: ;
+ }
I was going to say that this bit is missing an indent for each of the case statements, but I've just noticed that this rule is already applied quite inconsistently in this file, so it doesn't matter for now :)