My question is - does CursorAdapter re-uses views by internally
implementing the ViewHolder pattern or it needs to be coded as we
normally do in custom ArrayAdapter?
I'm not sure at what do you refer by the ViewHolder
pattern. If you are referring to having a helper class to cache looking for view each time(and setting it as a tag for the row View
) then the answer is no. If you want to implement this pattern you'll need to setup the holder(look for the views in the row view with findViewById
) in the newView
method and then set it as the tag for the row view. Then in the bindView
method you can call getTag
, retrieve the holder and use it. An example:
// custom CursorAdapter ...
@Override
public View newView(Context context, Cursor cursor, ViewGroup parent) {
View rowView = ((LayoutInflater) context
.getSystemService("layout_inflater")).inflate(
R.layout.row_layout, parent, false);
ViewHolder holder = new ViewHolder();
holder.v1 = rowView.findViewById(R.id.v1);
holder.v2 = rowView.findViewById(R.id.v2);
rowView.setTag(holder);
return rowView;
}
@Override
public void bindView(View view, Context context, Cursor cursor) {
ViewHolder holder = (ViewHolder) view.getTag();
// use the holder filled with views
// hlder.v1.setSomething
}
class ViewHolder {
View v1, v2;
}
// ...
If you are referring to the convertView
being reused(like in non Cursor
based adapters) then the answer is yes, the getView
method implements this pattern, you just need to implement the newView
and bindView
methods and you're guaranteed to get a view which was recycled(if possible at that moment).
CursorAdapter
(from the SDK or the compatibility package). – Stereotype