Add warning to sdtype text

rdt/transformers/__init__.py

@@ -2,6 +2,7 @@
 
 import importlib
 import inspect
+import warnings
 from collections import defaultdict
 from copy import deepcopy
 from functools import lru_cache
@@ -32,6 +33,7 @@
     AnonymizedFaker,
     PseudoAnonymizedFaker,
 )
+from rdt.transformers.utils import WarnDict
 
 __all__ = [
     'BaseTransformer',
@@ -88,16 +90,15 @@ def get_transformer_name(transformer):
     for transformer in BaseTransformer.get_subclasses()
 }
 
-
-DEFAULT_TRANSFORMERS = {
-    'boolean': UniformEncoder(),
-    'categorical': UniformEncoder(),
-    'datetime': UnixTimestampEncoder(),
-    'id': RegexGenerator(),
-    'numerical': FloatFormatter(),
-    'pii': AnonymizedFaker(),
-    'text': RegexGenerator(),
-}
+DEFAULT_TRANSFORMERS = WarnDict(
+    boolean=UniformEncoder(),
+    categorical=UniformEncoder(),
+    datetime=UnixTimestampEncoder(),
+    id=RegexGenerator(),
+    numerical=FloatFormatter(),
+    pii=AnonymizedFaker(),
+    text=RegexGenerator(),
+)
 
 
 @lru_cache()

rdt/transformers/utils.py

@@ -274,3 +274,37 @@ def learn_rounding_digits(data):
         name,
     )
     return None
+
+
+class WarnDict(dict):
+    """Custom dictionary to raise a deprecation warning."""
+
+    def get(self, key):
+        """Retrun the value for key if key is in the dictionary, else default.
+
+        If the key is `text` raises a `DeprecationWarning` stating that it will be
+        phased out.
+        """
+        if key == 'text':
+            warnings.warn(
+                "The sdtype 'text' is deprecated and will be phased out. "
+                "Please use 'id' instead.",
+                DeprecationWarning,
+            )
+
+        return super().get(key)
+
+    def __getitem__(self, key):
+        """Retrun the value for key if key is in the dictionary.
+
+        If the key is `text` raises a `DeprecationWarning` stating that it will be
+        phased out.
+        """
+        if key == 'text':
+            warnings.warn(
+                "The sdtype 'text' is deprecated and will be phased out. "
+                "Please use 'id' instead.",
+                DeprecationWarning,
+            )
+
+        return super().__getitem__(key)

tests/unit/transformers/test_utils.py

@@ -7,6 +7,7 @@
 import pytest
 
 from rdt.transformers.utils import (
+    WarnDict,
     _any,
     _max_repeat,
     check_nan_in_transform,
@@ -295,3 +296,25 @@ def test_learn_rounding_digits_nullable_numerical_pandas_dtypes():
     for column in data.columns:
         output = learn_rounding_digits(data[column])
         assert output == expected_output[column]
+
+
+def test_warn_dict():
+    """Test that ``WarnDict`` will raise a warning when called with `text`."""
+    # Setup
+    instance = WarnDict()
+    instance['text'] = 'text_transformer'
+
+    # Run
+    warning_msg = (
+        "The sdtype 'text' is deprecated and will be phased out. Please use 'id' instead."
+    )
+
+    with pytest.warns(DeprecationWarning, match=warning_msg):
+        result_access = instance['text']
+
+    with pytest.warns(DeprecationWarning, match=warning_msg):
+        result_get = instance.get('text')
+
+    # Assert
+    assert result_access == 'text_transformer'
+    assert result_get == 'text_transformer'