Pseudonymizer late error on non existing regexmapping

CHANGELOG.md

@@ -17,6 +17,7 @@ This release limits the maximum python version to `3.12.3` because of the issue
 
 * fixes a bug where it could happen that a config value could be overwritten by a default in a later configuration in a multi source config scenario
 * fixes a bug in the `field_manager` where extending a non list target leads to a processing failure
+* fixes a bug in `pseudonymizer` where a missing regex_mapping from an existing config_file causes logprep to crash continuously
 
 ## 12.0.0
 

logprep/processor/pseudonymizer/processor.py

@@ -58,6 +58,7 @@
 from urlextract import URLExtract
 
 from logprep.abc.processor import Processor
+from logprep.factory_error import InvalidConfigurationError
 from logprep.metrics.metrics import CounterMetric, GaugeMetric
 from logprep.processor.field_manager.processor import FieldManager
 from logprep.processor.pseudonymizer.rule import PseudonymizerRule
@@ -241,6 +242,10 @@ def _replace_regex_keywords_by_regex_expression(self):
             for dotted_field, regex_keyword in rule.pseudonyms.items():
                 if regex_keyword in self._regex_mapping:
                     rule.pseudonyms[dotted_field] = re.compile(self._regex_mapping[regex_keyword])
+                elif isinstance(regex_keyword, str):  # after the first run, the regex is compiled
+                    raise InvalidConfigurationError(
+                        f"Regex keyword '{regex_keyword}' not found in regex_mapping '{self._config.regex_mapping}'"
+                    )
 
     def _apply_rules(self, event: dict, rule: PseudonymizerRule):
         source_dict = {}

tests/unit/processor/pseudonymizer/test_pseudonymizer.py

@@ -10,6 +10,7 @@
 import pytest
 
 from logprep.factory import Factory
+from logprep.factory_error import InvalidConfigurationError
 from logprep.util.pseudo.encrypter import (
     DualPKCS1HybridCTREncrypter,
     DualPKCS1HybridGCMEncrypter,
@@ -1065,3 +1066,20 @@ def test_uses_encrypter(self, mode, encrypter_class):
         config["mode"] = mode
         object_with_encrypter = Factory.create({"pseudonymizer": config})
         assert isinstance(object_with_encrypter._encrypter, encrypter_class)
+
+    def test_setup_raises_invalid_configuration_on_missing_regex_mapping(self):
+        rule_dict = {
+            "filter": "winlog.event_id: 1234 AND winlog.provider_name: Test456",
+            "pseudonymizer": {
+                "mapping": {
+                    "winlog.event_data.param2": "RE_WHOLE_FIELD",
+                }
+            },
+        }
+        self._load_specific_rule(rule_dict)
+        self.object._specific_rules[0].mapping["winlog.event_data.param2"] = "RE_DOES_NOT_EXIST"
+        error_message = (
+            r"Regex keyword 'RE_DOES_NOT_EXIST' not found in regex_mapping '.*\/regex_mapping.yml'"
+        )
+        with pytest.raises(InvalidConfigurationError, match=error_message):
+            self.object.setup()