Allow lists wherever possible in builtin operators

fiftyone/core/dataset.py

@@ -1936,8 +1936,8 @@ def _populate_summary_field(self, field_name, summary_info):
         else:
             _id = "_id"
 
-        if list_fields:
-            pipeline.append({"$unwind": "$" + list_fields[0]})
+        for list_field in list_fields:
+            pipeline.append({"$unwind": "$" + list_field})
 
         if field_type == "categorical":
             if include_counts:

fiftyone/core/odm/mixins.py

@@ -670,6 +670,8 @@ def _clear_fields(cls, sample_collection, paths):
         """
         is_dataset = isinstance(sample_collection, fod.Dataset)
 
+        paths = _remove_nested_paths(paths)
+
         simple_paths = []
         coll_paths = []
 
@@ -716,6 +718,8 @@ def _delete_fields(cls, paths, error_level=0):
         media_type = dataset.media_type
         is_frame_field = cls._is_frames_doc
 
+        paths = _remove_nested_paths(paths)
+
         del_paths = []
         del_schema_paths = []
 
@@ -819,6 +823,8 @@ def _remove_dynamic_fields(cls, paths, error_level=0):
         dataset = cls._dataset
         dataset_doc = dataset._doc
 
+        paths = _remove_nested_paths(paths)
+
         del_paths = []
 
         for path in paths:
@@ -1863,6 +1869,14 @@ def _split_path(path):
     return chunks[0], chunks[1]
 
 
+def _remove_nested_paths(paths):
+    return [
+        path
+        for path in paths
+        if not any(path.startswith(p + ".") for p in paths)
+    ]
+
+
 def _add_field_doc(field_docs, root_doc, field_or_doc):
     if isinstance(field_or_doc, fof.Field):
         new_field_doc = SampleFieldDocument.from_field(field_or_doc)