Add uuid field to the return value of reader

backend/data_import/pipeline/data.py

@@ -1,5 +1,4 @@
 import abc
-import uuid
 from typing import Any, Dict
 
 from pydantic import UUID4, BaseModel, validator
@@ -15,7 +14,6 @@ class BaseData(BaseModel, abc.ABC):
     meta: Dict[Any, Any] = {}
 
     def __init__(self, **data):
-        data["uuid"] = uuid.uuid4()
         super().__init__(**data)
 
     @classmethod

backend/data_import/pipeline/readers.py

@@ -1,6 +1,7 @@
 import abc
 import collections.abc
 import dataclasses
+import uuid
 from typing import Any, Dict, Iterator, List
 
 import pandas as pd
@@ -13,6 +14,7 @@ DEFAULT_LABEL_COLUMN = "label"
 LINE_NUM_COLUMN = "#line_num"
 FILE_NAME_COLUMN = "filename"
 UPLOAD_NAME_COLUMN = "upload_name"
+UUID_COLUMN = "uuid"
 
 
 class BaseReader(collections.abc.Iterable):
@@ -80,6 +82,7 @@ class Reader(BaseReader):
                 try:
                     yield {
                         LINE_NUM_COLUMN: line_num,
+                        UUID_COLUMN: uuid.uuid4(),
                         FILE_NAME_COLUMN: filename.generated_name,
                         UPLOAD_NAME_COLUMN: filename.upload_name,
                         **row,

backend/data_import/tests/test_reader.py

@@ -1,10 +1,16 @@
 import unittest
-from unittest.mock import MagicMock
+from unittest.mock import MagicMock, patch
 
 import pandas as pd
 from pandas.testing import assert_frame_equal
 
-from data_import.pipeline.readers import FILE_NAME_COLUMN, LINE_NUM_COLUMN, Reader
+from data_import.pipeline.readers import (
+    FILE_NAME_COLUMN,
+    LINE_NUM_COLUMN,
+    UPLOAD_NAME_COLUMN,
+    UUID_COLUMN,
+    Reader,
+)
 
 
 class TestReader(unittest.TestCase):
@@ -12,18 +18,36 @@ class TestReader(unittest.TestCase):
         self.parser = MagicMock()
         self.parser.parse.return_value = [{"a": 1}, {"a": 2}]
         filename = MagicMock()
+        filename.generated_name = "filename"
+        filename.upload_name = "upload_name"
         self.filenames = MagicMock()
         self.filenames.__iter__.return_value = [filename]
         self.rows = [
-            {LINE_NUM_COLUMN: 1, FILE_NAME_COLUMN: filename, "a": 1},
-            {LINE_NUM_COLUMN: 2, FILE_NAME_COLUMN: filename, "a": 2},
+            {
+                LINE_NUM_COLUMN: 1,
+                UUID_COLUMN: "uuid",
+                FILE_NAME_COLUMN: filename.generated_name,
+                UPLOAD_NAME_COLUMN: filename.upload_name,
+                "a": 1,
+            },
+            {
+                LINE_NUM_COLUMN: 2,
+                UUID_COLUMN: "uuid",
+                FILE_NAME_COLUMN: filename.generated_name,
+                UPLOAD_NAME_COLUMN: filename.upload_name,
+                "a": 2,
+            },
         ]
 
-    def test_iter_method(self):
+    @patch("data_import.pipeline.readers.uuid.uuid4")
+    def test_iter_method(self, mock):
+        mock.return_value = "uuid"
         reader = Reader(self.filenames, self.parser)
         self.assertEqual(list(reader), self.rows)
 
-    def test_batch(self):
+    @patch("data_import.pipeline.readers.uuid.uuid4")
+    def test_batch(self, mock):
+        mock.return_value = "uuid"
         reader = Reader(self.filenames, self.parser)
         batch = next(reader.batch(2))
         expected_df = pd.DataFrame(self.rows)