more robust handling of exports

Author: Matt Sokoloff

labelbox/data/serialization/labelbox_v1/converter.py

@@ -29,8 +29,16 @@ def deserialize_video(json_data: Iterable[Dict[str, Any]],
         Returns:
             LabelGenerator containing the video data.
         """
-        label_generator = (LBV1Label(**example).to_common()
-                           for example in LBV1VideoIterator(json_data, client))
+
+        def label_generator():
+            for example in LBV1VideoIterator(json_data, client):
+                if example['Label']:
+                    if 'frames' not in example['Label']:
+                        raise ValueError(
+                            "Use `LBV1Converter.deserialize` to process non-video data"
+                        )
+                    yield LBV1Label(**example).to_common()
+
         return LabelGenerator(data=label_generator)
 
     @staticmethod
@@ -50,7 +58,9 @@ def label_generator():
                     raise ValueError(
                         "Use `LBV1Converter.deserialize_video` to process video"
                     )
-                yield LBV1Label(**example).to_common()
+                if example['Label']:
+                    # Don't construct empty dict
+                    yield LBV1Label(**example).to_common()
 
         return LabelGenerator(data=label_generator())
 

labelbox/schema/project.py

@@ -173,13 +173,17 @@ def video_label_generator(self, timeout_seconds=60):
         Returns:
             LabelGenerator for accessing labels for each video
         """
+        _check_converter_import()
         json_data = self.export_labels(download=True,
                                        timeout_seconds=timeout_seconds)
-        if 'frames' not in json_data[0]['Label']:
+        is_video = [
+            'frames' in row['Label'] for row in json_data if row['Label']
+        ]
+        if len(is_video) and not all(is_video):
             raise ValueError(
-                "frames key not found in the first label. Cannot export video data."
-            )
-        _check_converter_import()
+                "Found non-video data rows in export. "
+                "Use project.export_labels() to export projects with mixed data types. "
+                "Or use project.label_generator() for text and imagery data.")
         return LBV1Converter.deserialize_video(json_data, self.client)
 
     def label_generator(self, timeout_seconds=60):
@@ -189,9 +193,17 @@ def label_generator(self, timeout_seconds=60):
         Returns:
             LabelGenerator for accessing labels for each text or image
         """
+        _check_converter_import()
         json_data = self.export_labels(download=True,
                                        timeout_seconds=timeout_seconds)
-        _check_converter_import()
+        is_video = [
+            'frames' in row['Label'] for row in json_data if row['Label']
+        ]
+        if len(is_video) and not any(is_video):
+            raise ValueError(
+                "Found video data rows in export. "
+                "Use project.export_labels() to export projects with mixed data types. "
+                "Or use project.video_label_generator() for video data.")
         return LBV1Converter.deserialize(json_data)
 
     def export_labels(self, download=False, timeout_seconds=60):