Add 13 German country-specific predefined PII PatternRecognizers and…

docs/supported_entities.md

@@ -103,6 +103,23 @@ For more information, refer to the [adding new recognizers documentation](analyz
 | KR_DRIVER_LICENSE    |  The Korean driver license number is a 12-digit number. | Pattern match, context and custom logic. |
 | KR_PASSPORT| The Korean Passport Number  | Pattern match, context. |
 
+### Germany
+| FieldType  | Description                                                                                             | Detection Method                         |
+|------------|---------------------------------------------------------------------------------------------------------|------------------------------------------|
+| DE_BSNR | The German Betriebsstättennummer (BSNR) is a 9-digit facility identifier for healthcare providers. | Pattern match, context and regional code validation. |
+| DE_COMMERCIAL_REGISTER | German commercial register number (Handelsregisternummer) in formats like HRA/HRB 12345 for registered businesses. | Pattern match and context. |
+| DE_DRIVER_LICENSE | German driver license number (Führerscheinnummer), an 11-character alphanumeric identifier. | Pattern match and context. |
+| DE_KVNR | The German Krankenversichertennummer (KVNR) is a 10-character health insurance number starting with a letter. | Pattern match, context and checksum. |
+| DE_LANR | The German Lebenslange Arztnummer (LANR) is a 9-digit lifelong physician identifier. | Pattern match, context and checksum. |
+| DE_LICENSE_PLATE | German vehicle license plate (Kfz-Kennzeichen) in standard format like B-AB 1234. | Pattern match and context. |
+| DE_PASSPORT | German passport number (Reisepassnummer), a 9-character alphanumeric identifier. | Pattern match, context and checksum. |
+| DE_PERSONAL_ID | German personal identity card number (Personalausweisnummer), a 9 or 10-character alphanumeric identifier. | Pattern match, context and checksum. |
+| DE_POSTAL_CODE | German postal code (Postleitzahl/PLZ), a 5-digit number from 01001 to 99998. | Pattern match and context. |
+| DE_SOCIAL_SECURITY | German social security number (Sozialversicherungsnummer/SVNR), a 12-character identifier with area code and birthdate. | Pattern match, context and checksum. |
+| DE_TAX_ID | German tax identification number (Steuer-Identifikationsnummer), an 11-digit unique tax identifier. | Pattern match, context and checksum. |
+| DE_TELEMATIK_ID | German Telematik-ID for healthcare IT infrastructure, identifying patients and healthcare professionals. | Pattern match and context. |
+| DE_VAT_CODE | German VAT identification number (Umsatzsteuer-Identifikationsnummer), format DE followed by 9 digits. | Pattern match and context. |
+
 ### Thai
 | FieldType  | Description                                                                                             | Detection Method                         |
 |------------|---------------------------------------------------------------------------------------------------------|------------------------------------------|

presidio-analyzer/presidio_analyzer/conf/default_recognizers.yaml

@@ -208,6 +208,84 @@ recognizers:
     type: predefined
     enabled: false
 
+  - name: DeBsnrRecognizer
+    supported_languages:
+    - de
+    type: predefined
+    enabled: false
+
+  - name: DeCommercialRegisterRecognizer
+    supported_languages:
+    - de
+    type: predefined
+    enabled: false
+
+  - name: DeDriverLicenseRecognizer
+    supported_languages:
+    - de
+    type: predefined
+    enabled: false
+
+  - name: DeKvnrRecognizer
+    supported_languages:
+    - de
+    type: predefined
+    enabled: false
+
+  - name: DeLanrRecognizer
+    supported_languages:
+    - de
+    type: predefined
+    enabled: false
+
+  - name: DeLicensePlateRecognizer
+    supported_languages:
+    - de
+    type: predefined
+    enabled: false
+
+  - name: DePassportRecognizer
+    supported_languages:
+    - de
+    type: predefined
+    enabled: false
+
+  - name: DePersonalIdRecognizer
+    supported_languages:
+    - de
+    type: predefined
+    enabled: false
+
+  - name: DePostalCodeRecognizer
+    supported_languages:
+    - de
+    type: predefined
+    enabled: false
+
+  - name: DeSocialSecurityRecognizer
+    supported_languages:
+    - de
+    type: predefined
+    enabled: false
+
+  - name: DeTaxIdRecognizer
+    supported_languages:
+    - de
+    type: predefined
+    enabled: false
+
+  - name: DeTelematikIdRecognizer
+    supported_languages:
+    - de
+    type: predefined
+    enabled: false
+
+  - name: DeVatCodeRecognizer
+    supported_languages:
+    - de
+    type: predefined
+    enabled: false
+
   - name: OllamaRecognizer
     class_name: OllamaLangExtractRecognizer
     supported_languages:

presidio-analyzer/presidio_analyzer/predefined_recognizers/__init__.py

@@ -14,11 +14,34 @@
 from .country_specific.finland.fi_personal_identity_code_recognizer import (
     FiPersonalIdentityCodeRecognizer,
 )
-from .country_specific.india import (
-    InVehicleRegistrationRecognizer,
+
+# Germany recognizers
+from .country_specific.germany.de_bsnr_recognizer import DeBsnrRecognizer
+from .country_specific.germany.de_commercial_register_recognizer import (
+    DeCommercialRegisterRecognizer,
+)
+from .country_specific.germany.de_driver_license_recognizer import (
+    DeDriverLicenseRecognizer,
+)
+from .country_specific.germany.de_kvnr_recognizer import DeKvnrRecognizer
+from .country_specific.germany.de_lanr_recognizer import DeLanrRecognizer
+from .country_specific.germany.de_license_plate_recognizer import (
+    DeLicensePlateRecognizer,
 )
+from .country_specific.germany.de_passport_recognizer import DePassportRecognizer
+from .country_specific.germany.de_personal_id_recognizer import DePersonalIdRecognizer
+from .country_specific.germany.de_postal_code_recognizer import DePostalCodeRecognizer
+from .country_specific.germany.de_social_security_recognizer import (
+    DeSocialSecurityRecognizer,
+)
+from .country_specific.germany.de_tax_id_recognizer import DeTaxIdRecognizer
+from .country_specific.germany.de_telematik_id_recognizer import DeTelematikIdRecognizer
+from .country_specific.germany.de_vat_code_recognizer import DeVatCodeRecognizer
 
 # India recognizers
+from .country_specific.india import (
+    InVehicleRegistrationRecognizer,
+)
 from .country_specific.india.in_aadhaar_recognizer import InAadhaarRecognizer
 from .country_specific.india.in_gstin_recognizer import InGstinRecognizer
 from .country_specific.india.in_pan_recognizer import InPanRecognizer
@@ -115,6 +138,19 @@
 __all__ = [
     "AbaRoutingRecognizer",
     "CreditCardRecognizer",
+    "DeBsnrRecognizer",
+    "DeCommercialRegisterRecognizer",
+    "DeDriverLicenseRecognizer",
+    "DeKvnrRecognizer",
+    "DeLanrRecognizer",
+    "DeLicensePlateRecognizer",
+    "DePassportRecognizer",
+    "DePersonalIdRecognizer",
+    "DePostalCodeRecognizer",
+    "DeSocialSecurityRecognizer",
+    "DeTaxIdRecognizer",
+    "DeTelematikIdRecognizer",
+    "DeVatCodeRecognizer",
     "CryptoRecognizer",
     "DateRecognizer",
     "EmailRecognizer",

presidio-analyzer/presidio_analyzer/predefined_recognizers/country_specific/germany/__init__.py

@@ -0,0 +1,31 @@
+"""Germany-specific recognizers."""
+
+from .de_bsnr_recognizer import DeBsnrRecognizer
+from .de_commercial_register_recognizer import DeCommercialRegisterRecognizer
+from .de_driver_license_recognizer import DeDriverLicenseRecognizer
+from .de_kvnr_recognizer import DeKvnrRecognizer
+from .de_lanr_recognizer import DeLanrRecognizer
+from .de_license_plate_recognizer import DeLicensePlateRecognizer
+from .de_passport_recognizer import DePassportRecognizer
+from .de_personal_id_recognizer import DePersonalIdRecognizer
+from .de_postal_code_recognizer import DePostalCodeRecognizer
+from .de_social_security_recognizer import DeSocialSecurityRecognizer
+from .de_tax_id_recognizer import DeTaxIdRecognizer
+from .de_telematik_id_recognizer import DeTelematikIdRecognizer
+from .de_vat_code_recognizer import DeVatCodeRecognizer
+
+__all__ = [
+    "DeBsnrRecognizer",
+    "DeCommercialRegisterRecognizer",
+    "DeDriverLicenseRecognizer",
+    "DeKvnrRecognizer",
+    "DeLanrRecognizer",
+    "DeLicensePlateRecognizer",
+    "DePassportRecognizer",
+    "DePersonalIdRecognizer",
+    "DePostalCodeRecognizer",
+    "DeSocialSecurityRecognizer",
+    "DeTaxIdRecognizer",
+    "DeTelematikIdRecognizer",
+    "DeVatCodeRecognizer",
+]

presidio-analyzer/presidio_analyzer/predefined_recognizers/country_specific/germany/de_bsnr_recognizer.py

@@ -0,0 +1,142 @@
+from typing import List, Optional, Tuple
+
+from presidio_analyzer import EntityRecognizer, Pattern, PatternRecognizer
+
+
+class DeBsnrRecognizer(PatternRecognizer):
+    """
+    Recognize German BSNR (Betriebsstättennummer) using regex and validation.
+
+    The BSNR is a facility number that uniquely identifies the location of
+    service provision in the German statutory health insurance system:
+    - 9 digits total
+    - Digits 1-2: KV state/regional association code (see VALID_KV_CODES)
+    - Digits 3-7: Facility identifier (assigned by KV)
+    - Digits 8-9: Additional digits (often "00" for older BSNRs)
+
+    The BSNR appears in prescriptions, discharge letters, and billing documents,
+    identifying the treatment facility. This is quasi-PII as it can narrow down
+    where a patient received treatment.
+
+    Legal basis: §75 Abs. 7 SGB V
+    Issuing authority: Kassenärztliche Vereinigungen (KV)
+    Source: KBV Arztnummern-Richtlinie Anlage 1
+
+    :param patterns: List of patterns to be used by this recognizer
+    :param context: List of context words to increase confidence in detection
+    :param supported_language: Language this recognizer supports
+    :param supported_entity: The entity this recognizer can detect
+    :param replacement_pairs: List of tuples with potential replacement values
+    """
+
+    # Pattern source: https://wiki.hl7.de/index.php/LANR_und_BSNR
+
+    # Valid KV region codes per KBV Arztnummern-Richtlinie Anlage 1
+    # Standard KV regions
+    VALID_KV_CODES = {
+        "01",  # Schleswig-Holstein
+        "02",  # Hamburg
+        "03",  # Bremen
+        "17",  # Niedersachsen
+        "20",  # Westfalen-Lippe
+        "38",  # Nordrhein
+        "46",  # Hessen
+        "51",  # Rheinland-Pfalz
+        "52",  # Baden-Württemberg
+        "71",  # Bayern
+        "72",  # Berlin
+        "73",  # Saarland
+        "74",  # KBV (Kassenärztliche Bundesvereinigung)
+        "78",  # Mecklenburg-Vorpommern
+        "83",  # Brandenburg
+        "88",  # Sachsen-Anhalt
+        "93",  # Thüringen
+        "98",  # Sachsen
+        # Special codes for hospitals (Anlage 8 BMV-Ä)
+        "35",  # Krankenhäuser
+    }
+
+    PATTERNS = [
+        Pattern(
+            "BSNR (9 digits)",
+            r"\b[0-9]{9}\b",
+            0.05,  # Very low score - requires context or valid KV code
+        ),
+        Pattern(
+            "BSNR (with context)",
+            r"(?i)(?:bsnr|betriebsstättennummer|betriebsstaetten-nr|betriebsstätten-nr)[\s:]*([0-9]{9})\b",
+            0.5,
+        ),
+    ]
+
+    CONTEXT = [
+        "bsnr",
+        "betriebsstättennummer",
+        "betriebsst\u00e4ttennummer",  # With umlaut
+        "betriebsstaetten-nr",
+        "betriebsst\u00e4tten-nr",  # With umlaut
+        "facility number",
+        "praxis",
+        "praxisnummer",
+        "behandlungsort",
+        "einrichtung",
+        "klinik",
+        "krankenhaus",
+        "behandlungsstelle",
+    ]
+
+    def __init__(
+        self,
+        patterns: Optional[List[Pattern]] = None,
+        context: Optional[List[str]] = None,
+        supported_language: str = "de",
+        supported_entity: str = "DE_BSNR",
+        replacement_pairs: Optional[List[Tuple[str, str]]] = None,
+        name: Optional[str] = None,
+    ):
+        self.replacement_pairs = (
+            replacement_pairs if replacement_pairs else [("-", ""), (" ", ""), (".", "")]
+        )
+        patterns = patterns if patterns else self.PATTERNS
+        context = context if context else self.CONTEXT
+        super().__init__(
+            supported_entity=supported_entity,
+            patterns=patterns,
+            context=context,
+            supported_language=supported_language,
+            name=name,
+        )
+
+    def validate_result(self, pattern_text: str) -> Optional[bool]:
+        """
+        Validate the BSNR format using KV regional code validation.
+
+        Validates that the first 2 digits match a valid KV region code
+        per KBV Arztnummern-Richtlinie Anlage 1.
+
+        :param pattern_text: Text detected as pattern by regex
+        :return: True if valid KV code, False if invalid, None if uncertain
+        """
+        pattern_text = EntityRecognizer.sanitize_value(
+            pattern_text, self.replacement_pairs
+        )
+
+        if len(pattern_text) != 9:
+            return False
+
+        if not pattern_text.isdigit():
+            return False
+
+        # Basic validation: BSNR should not be all zeros
+        if pattern_text == "000000000":
+            return False
+
+        # Validate KV regional code (digits 1-2)
+        kv_code = pattern_text[:2]
+        if kv_code in self.VALID_KV_CODES:
+            # Valid KV code - increase confidence
+            return True
+
+        # Unknown KV code - could be valid (historic or special cases)
+        # but reduce confidence by returning None
+        return None