Merge branch 'develop' of https://github.com/Labelbox/labelbox-python into ms/coco-invalid-polygons

Author: Matt Sokoloff

CHANGELOG.md

@@ -1,9 +1,18 @@
 # Changelog
 
-# Version 0.0.0 (YYYY-MM-DD) - In Progress
+# Version 3.28.0 (2022-10-14)
+
+### Added
+* Added warning for upcoming change in default project queue_mode setting
+* Added notebook example for importing Conversational Text annotations using Model-Assisted Labeling
 
 ### Changed
-* Update QueueMode enum to support new value for QueueMode.Batch = `BATCH`.
+* Updated QueueMode enum to support new value for QueueMode.Batch = `BATCH`.
+* Task.failed_data_rows is now a property
+
+### Fixed
+* Fixed Task.wait_till_done() showing warning message for every completed task, instead of only warning when task has errors
+* Fixed error on dataset creation step in examples/annotation_import/video.ipynb notebook
 
 # Version 3.27.2 (2022-10-04)
 

docs/source/conf.py

@@ -21,7 +21,7 @@
 copyright = '2021, Labelbox'
 author = 'Labelbox'
 
-release = '3.27.2'
+release = '3.28.0'
 
 # -- General configuration ---------------------------------------------------
 

labelbox/__init__.py

@@ -1,5 +1,5 @@
 name = "labelbox"
-__version__ = "3.27.2"
+__version__ = "3.28.0"
 
 from labelbox.client import Client
 from labelbox.schema.project import Project

labelbox/client.py

@@ -1185,9 +1185,13 @@ def _format_failed_rows(rows: List[str],
                 errors.extend(
                     _format_failed_rows(data['deletedDataRowGlobalKeys'],
                                         "Data Row deleted"))
+
+                # Invalid results may contain empty string, so we must filter
+                # them prior to checking for PARTIAL_SUCCESS
+                filtered_results = list(filter(lambda r: r != '', results))
                 if not errors:
                     status = CollectionJobStatus.SUCCESS.value
-                elif errors and results:
+                elif errors and len(filtered_results) > 0:
                     status = CollectionJobStatus.PARTIAL_SUCCESS.value
                 else:
                     status = CollectionJobStatus.FAILURE.value

labelbox/schema/task.py

@@ -63,9 +63,7 @@ def wait_till_done(self, timeout_seconds=300) -> None:
         check_frequency = 2  # frequency of checking, in seconds
         while True:
             if self.status != "IN_PROGRESS":
-                if self.status == "FAILED" or (self.status == "COMPLETE" and
-                                               self.failed_data_rows
-                                               is not None):
+                if self.errors is not None:
                     logger.warning(
                         "There are errors present. Please look at `task.errors` for more details"
                     )
@@ -87,7 +85,7 @@ def errors(self) -> Optional[Dict[str, Any]]:
             result = self._fetch_remote_json()
             return result["error"]
         elif self.status == "COMPLETE":
-            return self.failed_data_rows()
+            return self.failed_data_rows
         return None
 
     @property
@@ -105,11 +103,12 @@ def result(self) -> List[Dict[str, Any]]:
                 'global_key': data_row.get('globalKey'),
             } for data_row in result['createdDataRows']]
 
+    @property
     def failed_data_rows(self) -> Optional[Dict[str, Any]]:
         """ Fetch data rows which failed to be created for an import task.
         """
         result = self._fetch_remote_json()
-        if result.get("errors") is not None:
+        if len(result.get("errors", [])) > 0:
             return result["errors"]
         else:
             return None

tests/integration/test_data_row_media_attributes.py

@@ -1,5 +1,10 @@
+from time import sleep
+
+
 def test_export_empty_media_attributes(configured_project_with_label):
     project, _, _, _ = configured_project_with_label
+    # Wait for exporter to retrieve latest labels
+    sleep(10)
     labels = project.label_generator()
     label = next(labels)
     assert label.data.media_attributes == {}

tests/integration/test_data_row_metadata.py

@@ -1,4 +1,5 @@
 from datetime import datetime
+from time import sleep
 
 import pytest
 import uuid
@@ -70,6 +71,8 @@ def make_metadata(dr_id) -> DataRowMetadata:
 
 def test_export_empty_metadata(configured_project_with_label):
     project, _, _, _ = configured_project_with_label
+    # Wait for exporter to retrieve latest labels
+    sleep(10)
     labels = project.label_generator()
     label = next(labels)
     assert label.data.metadata == []

tests/integration/test_data_rows.py

@@ -352,7 +352,8 @@ def test_create_data_rows_with_invalid_metadata(dataset, image_url):
         DataRow.metadata_fields: fields
     }])
     task.wait_till_done()
-    assert task.status == "FAILED"
+    assert task.status == "COMPLETE"
+    assert len(task.failed_data_rows) > 0
 
 
 def test_create_data_rows_with_metadata_missing_value(dataset, image_url):
@@ -632,7 +633,8 @@ def test_data_row_bulk_creation_with_same_global_keys(dataset, sample_image):
     }])
 
     task.wait_till_done()
-    assert task.status == "FAILED"
+    assert task.status == "COMPLETE"
+    assert len(task.failed_data_rows) > 0
     assert len(list(dataset.data_rows())) == 0
 
     task = dataset.create_data_rows([{

tests/integration/test_export.py

@@ -1,3 +1,4 @@
+from time import sleep
 import uuid
 
 from labelbox.data.annotation_types.annotation import ObjectAnnotation
@@ -46,6 +47,8 @@ def test_export_annotations_nested_checklist(
     task = LabelImport.create_from_objects(client, project.uid,
                                            f'label-import-{uuid.uuid4()}', data)
     task.wait_until_done()
+    # Wait for exporter to retrieve latest labels
+    sleep(10)
     labels = project.label_generator().as_list()
     object_annotation = [
         annot for annot in next(labels).annotations

tests/integration/test_global_keys.py

@@ -251,9 +251,10 @@ def test_get_data_row_ids_for_invalid_global_keys(client, dataset, image_url):
     assert res['status'] == "PARTIAL SUCCESS"
 
     assert len(res['errors']) == 1
-    assert len(res['results']) == 1
+    assert len(res['results']) == 2
 
     assert res['errors'][0]['error'] == "Data Row not found"
     assert res['errors'][0]['global_key'] == gk_1
 
-    assert res['results'][0] == dr_2.uid
+    assert res['results'][0] == ''
+    assert res['results'][1] == dr_2.uid