Add DLP code samples for custom info types [(#1524)](GoogleCloudPlatf…

…orm/python-docs-samples#1524)

* Add custom info type samples to inspect_content.py

Use flags to indicate dictionary word lists and regex patterns, then parse them into custom info types.

* Make code compatible with python 2.7

* Add missing commas

* Remove bad import

* Add tests for custom info types

* Add info_types parameter to deid.py

* Update deid tests to use info_types parameter

* Fix indentation

* Add blank lines

* Share logic for building custom info types

* Fix line too long

* Fix typo.

* Revert "Fix typo."

This reverts commit b4ffea6eef1fc2ccd2a4f17adb6e9492e54f1b76, so that
the sharing of the custom info type logic can be reverted as well to
make the code samples more readable.

* Revert "Share logic for building custom info types"

This reverts commit 47fc04f74c77db3bd5397459cf9242dc11521c37. This makes
the code samples more readable.

* Switch from indexes to using enumerate.

* Updated help message for custom dictionaries.

* Fix enumerate syntax error.

samples/snippets/deid.py

@@ -20,7 +20,7 @@
 
 
 # [START dlp_deidentify_masking]
-def deidentify_with_mask(project, string, masking_character=None,
+def deidentify_with_mask(project, string, info_types, masking_character=None,
                          number_to_mask=0):
     """Uses the Data Loss Prevention API to deidentify sensitive data in a
     string by masking it with a character.
@@ -44,6 +44,11 @@ def deidentify_with_mask(project, string, masking_character=None,
     # Convert the project id into a full resource id.
     parent = dlp.project_path(project)
 
+    # Construct inspect configuration dictionary
+    inspect_config = {
+        'info_types': [{'name': info_type} for info_type in info_types]
+    }
+
     # Construct deidentify configuration dictionary
     deidentify_config = {
         'info_type_transformations': {
@@ -65,15 +70,16 @@ def deidentify_with_mask(project, string, masking_character=None,
 
     # Call the API
     response = dlp.deidentify_content(
-        parent, deidentify_config=deidentify_config, item=item)
+        parent, inspect_config=inspect_config,
+        deidentify_config=deidentify_config, item=item)
 
     # Print out the results.
     print(response.item.value)
 # [END dlp_deidentify_masking]
 
 
 # [START dlp_deidentify_fpe]
-def deidentify_with_fpe(project, string, alphabet=None,
+def deidentify_with_fpe(project, string, info_types, alphabet=None,
                         surrogate_type=None, key_name=None, wrapped_key=None):
     """Uses the Data Loss Prevention API to deidentify sensitive data in a
     string using Format Preserving Encryption (FPE).
@@ -127,6 +133,11 @@ def deidentify_with_fpe(project, string, alphabet=None,
             'name': surrogate_type
         }
 
+    # Construct inspect configuration dictionary
+    inspect_config = {
+        'info_types': [{'name': info_type} for info_type in info_types]
+    }
+
     # Construct deidentify configuration dictionary
     deidentify_config = {
         'info_type_transformations': {
@@ -146,7 +157,8 @@ def deidentify_with_fpe(project, string, alphabet=None,
 
     # Call the API
     response = dlp.deidentify_content(
-        parent, deidentify_config=deidentify_config, item=item)
+        parent, inspect_config=inspect_config,
+        deidentify_config=deidentify_config, item=item)
 
     # Print results
     print(response.item.value)
@@ -404,6 +416,13 @@ def write_data(data):
         'deid_mask',
         help='Deidentify sensitive data in a string by masking it with a '
              'character.')
+    mask_parser.add_argument(
+        '--info_types', action='append',
+        help='Strings representing info types to look for. A full list of '
+             'info categories and types is available from the API. Examples '
+             'include "FIRST_NAME", "LAST_NAME", "EMAIL_ADDRESS". '
+             'If unspecified, the three above examples will be used.',
+        default=['FIRST_NAME', 'LAST_NAME', 'EMAIL_ADDRESS'])
     mask_parser.add_argument(
         'project',
         help='The Google Cloud project id to use as a parent resource.')
@@ -423,6 +442,13 @@ def write_data(data):
         'deid_fpe',
         help='Deidentify sensitive data in a string using Format Preserving '
              'Encryption (FPE).')
+    fpe_parser.add_argument(
+        '--info_types', action='append',
+        help='Strings representing info types to look for. A full list of '
+             'info categories and types is available from the API. Examples '
+             'include "FIRST_NAME", "LAST_NAME", "EMAIL_ADDRESS". '
+             'If unspecified, the three above examples will be used.',
+        default=['FIRST_NAME', 'LAST_NAME', 'EMAIL_ADDRESS'])
     fpe_parser.add_argument(
          'project',
          help='The Google Cloud project id to use as a parent resource.')
@@ -532,11 +558,12 @@ def write_data(data):
     args = parser.parse_args()
 
     if args.content == 'deid_mask':
-        deidentify_with_mask(args.project, args.item,
+        deidentify_with_mask(args.project, args.item, args.info_types,
                              masking_character=args.masking_character,
                              number_to_mask=args.number_to_mask)
     elif args.content == 'deid_fpe':
-        deidentify_with_fpe(args.project, args.item, alphabet=args.alphabet,
+        deidentify_with_fpe(args.project, args.item, args.info_types,
+                            alphabet=args.alphabet,
                             wrapped_key=args.wrapped_key,
                             key_name=args.key_name,
                             surrogate_type=args.surrogate_type)

samples/snippets/deid_test.py

@@ -43,7 +43,8 @@ def tempdir():
 
 
 def test_deidentify_with_mask(capsys):
-    deid.deidentify_with_mask(GCLOUD_PROJECT, HARMFUL_STRING)
+    deid.deidentify_with_mask(GCLOUD_PROJECT, HARMFUL_STRING,
+                              ['US_SOCIAL_SECURITY_NUMBER'])
 
     out, _ = capsys.readouterr()
     assert 'My SSN is *********' in out
@@ -60,14 +61,17 @@ def test_deidentify_with_mask_masking_character_specified(capsys):
     deid.deidentify_with_mask(
         GCLOUD_PROJECT,
         HARMFUL_STRING,
+        ['US_SOCIAL_SECURITY_NUMBER'],
         masking_character='#')
 
     out, _ = capsys.readouterr()
     assert 'My SSN is #########' in out
 
 
 def test_deidentify_with_mask_masking_number_specified(capsys):
-    deid.deidentify_with_mask(GCLOUD_PROJECT, HARMFUL_STRING, number_to_mask=7)
+    deid.deidentify_with_mask(GCLOUD_PROJECT, HARMFUL_STRING,
+                              ['US_SOCIAL_SECURITY_NUMBER'],
+                              number_to_mask=7)
 
     out, _ = capsys.readouterr()
     assert 'My SSN is *******27' in out
@@ -77,6 +81,7 @@ def test_deidentify_with_fpe(capsys):
     deid.deidentify_with_fpe(
         GCLOUD_PROJECT,
         HARMFUL_STRING,
+        ['US_SOCIAL_SECURITY_NUMBER'],
         alphabet='NUMERIC',
         wrapped_key=WRAPPED_KEY,
         key_name=KEY_NAME)
@@ -90,6 +95,7 @@ def test_deidentify_with_fpe_uses_surrogate_info_types(capsys):
     deid.deidentify_with_fpe(
         GCLOUD_PROJECT,
         HARMFUL_STRING,
+        ['US_SOCIAL_SECURITY_NUMBER'],
         alphabet='NUMERIC',
         wrapped_key=WRAPPED_KEY,
         key_name=KEY_NAME,