raise error for repeat uuid

Author: rllin

labelbox/schema/bulk_import_request.py

@@ -2,8 +2,11 @@
 import logging
 import time
 from pathlib import Path
+from typing import Any
 from typing import BinaryIO
+from typing import Dict
 from typing import Iterable
+from typing import Set
 from typing import Tuple
 from typing import Union
 
@@ -217,6 +220,7 @@ def create_from_objects(cls, client, project_id: str, name: str,
         Returns:
             BulkImportRequest object
         """
+        _validate_ndjson(predictions)
         data_str = ndjson.dumps(predictions)
         data = data_str.encode('utf-8')
         file_name = _make_file_name(project_id, name)
@@ -264,8 +268,7 @@ def create_from_local_file(cls,
                 # by iterating through the file so we only store
                 # each line in memory rather than the entire file
                 try:
-                    for line in reader:
-                        pass
+                    _validate_ndjson(reader)
                 except ValueError:
                     raise ValueError(f"{file} is not a valid ndjson file")
                 else:
@@ -274,3 +277,21 @@ def create_from_local_file(cls,
             response_data = _send_create_file_command(client, request_data,
                                                       file_name, file_data)
         return cls(client, response_data["createBulkImportRequest"])
+
+
+class NdjsonError(Exception):
+    pass
+
+
+class UuidError(NdjsonError):
+    pass
+
+
+def _validate_ndjson(lines: Iterable[Dict[str, Any]]) -> None:
+    uuids: Set[str] = set()
+    for line in lines:
+        uuid = line['uuid']
+        if uuid in uuids:
+            raise UuidError(f'{uuid} already used in this import job, '
+                            'must be unique for the project.')
+        uuids.add(line['uuid'])

tests/integration/test_bulk_import_request.py

@@ -4,7 +4,7 @@
 import pytest
 import requests
 
-from labelbox.schema.bulk_import_request import BulkImportRequest
+from labelbox.schema.bulk_import_request import BulkImportRequest, UuidError
 from labelbox.schema.enums import BulkImportRequestState
 
 PREDICTIONS = [{
@@ -116,6 +116,20 @@ def test_validate_ndjson(tmp_path, project):
         project.upload_annotations(name="name", annotations=str(file_path))
 
 
+def test_validate_ndjson_uuid(tmp_path, project):
+    file_name = f"repeat_uuid.ndjson"
+    file_path = tmp_path / file_name
+    repeat_uuid = PREDICTIONS.copy()
+    repeat_uuid[0]['uuid'] = 'test_uuid'
+    repeat_uuid[1]['uuid'] = 'test_uuid'
+
+    with file_path.open("w") as f:
+        ndjson.dump(repeat_uuid, f)
+
+    with pytest.raises(UuidError):
+        project.upload_annotations(name="name", annotations=str(file_path))
+
+
 @pytest.mark.slow
 def test_wait_till_done(project):
     name = str(uuid.uuid4())