Make ensure_ascii Dynamic with Default Set to True in JSON Serialization (#168)

* add parameter - ensure_ascii to load non ascii characters when set to false

* add unit test for ensuring ascii characters while loading

* update setup and changelog

Author: sgandhi1311

CHANGELOG.md

@@ -1,5 +1,8 @@
 # Changelog
 
+## 6.1.0
+  * Make ensure_ascii Dynamic with Default Set to True in JSON Serialization. Required to handle the special characters [#168](https://github.com/singer-io/singer-python/pull/168)
+
 ## 6.0.1
   * Pin backoff and simplejson to any version greater than or equal to the previously allowed version, up to the next major version [#167](https://github.com/singer-io/singer-python/pull/167)
 

setup.py

@@ -4,7 +4,7 @@
 import subprocess
 
 setup(name="singer-python",
-      version='6.0.1',
+      version='6.1.0',
       description="Singer.io utility library",
       author="Stitch",
       classifiers=['Programming Language :: Python :: 3 :: Only'],

singer/messages.py

@@ -218,12 +218,12 @@ def parse_message(msg):
         return None
 
 
-def format_message(message):
-    return json.dumps(message.asdict(), use_decimal=True)
+def format_message(message, ensure_ascii=True):
+    return json.dumps(message.asdict(), use_decimal=True, ensure_ascii=ensure_ascii)
 
 
-def write_message(message):
-    sys.stdout.write(format_message(message) + '\n')
+def write_message(message, ensure_ascii=True):
+    sys.stdout.write(format_message(message, ensure_ascii=ensure_ascii) + '\n')
     sys.stdout.flush()
 
 

tests/test_singer.py

@@ -1,5 +1,6 @@
 import singer
 import unittest
+from unittest.mock import patch
 import datetime
 import dateutil
 from decimal import Decimal
@@ -179,6 +180,31 @@ def test_parse_bulk_decs(self):
             value = self.create_record(value_str)
             self.assertEqual(Decimal(value_str), value)
 
+    @patch('sys.stdout')
+    def test_ensure_ascii_false(self, mock_stdout):
+        """
+        Setting ensure_ascii=False will preserve special characters like é 
+        in their original form.
+        """
+        rec = {"name": "José"}
+        expected_output = '{"type": "RECORD", "stream": "test_stream", "record": {"name": "José"}}\n'
+        rec_message = singer.RecordMessage(stream="test_stream", record=rec)
+        result = singer.write_message(rec_message, ensure_ascii=False)
+        mock_stdout.write.assert_called_once_with(expected_output)
+        mock_stdout.flush.assert_called_once()
+
+    @patch('sys.stdout')
+    def test_ensure_ascii_true(self, mock_stdout):
+        """
+        ensure_ascii defaults to True, special characters like é are 
+        escaped into their ASCII representation (e.g., \u00e9)
+        """
+        rec = {"name": "José"}
+        expected_output = '{"type": "RECORD", "stream": "test_stream", "record": {"name": "Jos\\u00e9"}}\n'
+        rec_message = singer.RecordMessage(stream="test_stream", record=rec)
+        result = singer.write_message(rec_message)
+        mock_stdout.write.assert_called_once_with(expected_output)
+        mock_stdout.flush.assert_called_once()
 
 if __name__ == '__main__':
     unittest.main()