langchain-experimental: Add allow_list support in experimental/data_anonymizer (#11597)

- **Description:** Add allow_list support in langchain experimental
data-anonymizer package
  - **Issue:** no
  - **Dependencies:** no
  - **Tag maintainer:** @hwchase17
  - **Twitter handle:**

libs/experimental/langchain_experimental/data_anonymizer/base.py

@@ -1,5 +1,5 @@
 from abc import ABC, abstractmethod
-from typing import Callable, Optional
+from typing import Callable, List, Optional
 
 from langchain_experimental.data_anonymizer.deanonymizer_mapping import MappingDataType
 from langchain_experimental.data_anonymizer.deanonymizer_matching_strategies import (
@@ -16,12 +16,19 @@ class AnonymizerBase(ABC):
         wrapping the behavior for all methods in a base class.
     """
 
-    def anonymize(self, text: str, language: Optional[str] = None) -> str:
+    def anonymize(
+        self,
+        text: str,
+        language: Optional[str] = None,
+        allow_list: Optional[List[str]] = None,
+    ) -> str:
         """Anonymize text"""
-        return self._anonymize(text, language)
+        return self._anonymize(text, language, allow_list)
 
     @abstractmethod
-    def _anonymize(self, text: str, language: Optional[str]) -> str:
+    def _anonymize(
+        self, text: str, language: Optional[str], allow_list: Optional[List[str]] = None
+    ) -> str:
         """Abstract method to anonymize text"""
 
 

libs/experimental/langchain_experimental/data_anonymizer/presidio.py

@@ -139,7 +139,12 @@ class PresidioAnonymizerBase(AnonymizerBase):
 
 
 class PresidioAnonymizer(PresidioAnonymizerBase):
-    def _anonymize(self, text: str, language: Optional[str] = None) -> str:
+    def _anonymize(
+        self,
+        text: str,
+        language: Optional[str] = None,
+        allow_list: Optional[List[str]] = None,
+    ) -> str:
         """Anonymize text.
         Each PII entity is replaced with a fake value.
         Each time fake values will be different, as they are generated randomly.
@@ -172,6 +177,7 @@ class PresidioAnonymizer(PresidioAnonymizerBase):
             text,
             entities=self.analyzed_fields,
             language=language,
+            allow_list=allow_list,
         )
 
         filtered_analyzer_results = (
@@ -226,7 +232,12 @@ class PresidioReversibleAnonymizer(PresidioAnonymizerBase, ReversibleAnonymizerB
             for key, inner_dict in self.deanonymizer_mapping.items()
         }
 
-    def _anonymize(self, text: str, language: Optional[str] = None) -> str:
+    def _anonymize(
+        self,
+        text: str,
+        language: Optional[str] = None,
+        allow_list: Optional[List[str]] = None,
+    ) -> str:
         """Anonymize text.
         Each PII entity is replaced with a fake value.
         Each time fake values will be different, as they are generated randomly.
@@ -261,6 +272,7 @@ class PresidioReversibleAnonymizer(PresidioAnonymizerBase, ReversibleAnonymizerB
             text,
             entities=self.analyzed_fields,
             language=language,
+            allow_list=allow_list,
         )
 
         filtered_analyzer_results = (

libs/experimental/tests/unit_tests/test_data_anonymizer.py

@@ -36,6 +36,21 @@ def test_anonymize(analyzed_fields: List[str], should_contain: bool) -> None:
     assert ("John Doe" in anonymized_text) == should_contain
 
 
+@pytest.mark.requires("presidio_analyzer", "presidio_anonymizer", "faker")
+@pytest.mark.parametrize(
+    "analyzed_fields,should_contain",
+    [(["PERSON"], True), (["PHONE_NUMBER"], True), (None, True)],
+)
+def test_anonymize_allow_list(analyzed_fields: List[str], should_contain: bool) -> None:
+    """Test anonymizing a name in a simple sentence"""
+    from langchain_experimental.data_anonymizer import PresidioAnonymizer
+
+    text = "Hello, my name is John Doe."
+    anonymizer = PresidioAnonymizer(analyzed_fields=analyzed_fields)
+    anonymized_text = anonymizer.anonymize(text, allow_list=["John Doe"])
+    assert ("John Doe" in anonymized_text) == should_contain
+
+
 @pytest.mark.requires("presidio_analyzer", "presidio_anonymizer", "faker")
 def test_anonymize_multiple() -> None:
     """Test anonymizing multiple items in a sentence"""

libs/experimental/tests/unit_tests/test_reversible_data_anonymizer.py

@@ -37,6 +37,21 @@ def test_anonymize(analyzed_fields: List[str], should_contain: bool) -> None:
     assert ("John Doe" in anonymized_text) == should_contain
 
 
+@pytest.mark.requires("presidio_analyzer", "presidio_anonymizer", "faker")
+@pytest.mark.parametrize(
+    "analyzed_fields,should_contain",
+    [(["PERSON"], True), (["PHONE_NUMBER"], True), (None, True)],
+)
+def test_anonymize_allow_list(analyzed_fields: List[str], should_contain: bool) -> None:
+    """Test anonymizing a name in a simple sentence"""
+    from langchain_experimental.data_anonymizer import PresidioReversibleAnonymizer
+
+    text = "Hello, my name is John Doe."
+    anonymizer = PresidioReversibleAnonymizer(analyzed_fields=analyzed_fields)
+    anonymized_text = anonymizer.anonymize(text, allow_list=["John Doe"])
+    assert ("John Doe" in anonymized_text) == should_contain
+
+
 @pytest.mark.requires("presidio_analyzer", "presidio_anonymizer", "faker")
 def test_anonymize_multiple() -> None:
     """Test anonymizing multiple items in a sentence"""