Files
Presidio/conf/default.yaml
2025-06-23 17:18:19 +02:00

137 lines
4.6 KiB
YAML

# =====================================================================
# CONFIGURATION PRESIDIO COMPLÈTE - RGPD FRANÇAIS/BELGE
# =====================================================================
# Langues supportées par l'ensemble du moteur
supported_languages: [en, fr]
# 1. CONFIGURATION DU MOTEUR NLP
# =====================================================================
nlp_configuration:
nlp_engine_name: spacy
models:
- lang_code: en
model_name: en_core_web_lg
- lang_code: fr
model_name: fr_core_news_sm
ner_model_configuration:
labels_to_ignore:
- MISC
- CARDINAL
- EVENT
- LANGUAGE
- LAW
- ORDINAL
- PERCENT
- PRODUCT
- QUANTITY
- WORK_OF_ART
confidence_thresholds:
DEFAULT_CONFIDENCE: 0.65
entities:
PERSON: 0.85
LOCATION: 0.75
ORGANIZATION: 0.7
# 2. CONFIGURATION DU REGISTRE DES DÉTECTEURS
# =====================================================================
recognizer_registry:
# A) CHARGER LES DÉTECTEURS PRÉDÉFINIS pour les langues supportées
# Presidio va charger les versions 'en' et 'fr' si elles existent
load_predefined_recognizers: true
# B) AJOUTER VOS DÉTECTEURS PERSONNALISÉS (surtout pour le français)
recognizers:
- name: EmailRecognizer
supported_language: fr
entity_name: EMAIL_ADDRESS
patterns:
- name: Email Pattern
regex: "\\b[A-Za-z0-9._%+-]+@[A-Za-z0-9.-]+\\.[A-Z|a-z]{2,}\\b"
score: 1.0
context: ["email", "courriel", "mail"]
- name: PhoneRecognizer
supported_language: fr
entity_name: PHONE_NUMBER
patterns:
- name: Phone Pattern (FR/BE/LUX)
regex: "\\b(?:(?:\\+|00)?(?:32|33|352)|0)\\s?[1-9](?:[\\s.-]?\\d{2}){3,4}\\b"
score: 0.8
context: ["téléphone", "tel", "mobile", "gsm", "contact"]
- name: IbanRecognizer
supported_language: fr
entity_name: IBAN
patterns:
- name: IBAN Pattern
regex: "\\b[A-Z]{2}[0-9]{2}\\s?(?:[A-Z0-9]{4}\\s?){2,7}[A-Z0-9]{1,4}\\b"
score: 0.95
context: ["iban", "compte", "bancaire"]
- name: CreditCardRecognizer
supported_language: fr
entity_name: CREDIT_CARD_NUMBER
patterns:
- name: Credit Card Pattern
regex: "\\b(?:4[0-9]{12}(?:[0-9]{3})?|5[1-5][0-9]{14}|6(?:011|5[0-9]{2})[0-9]{12}|3[47][0-9]{13}|3(?:0[0-5]|[68][0-9])[0-9]{11}|(?:2131|1800|35\\d{3})\\d{11})\\b"
score: 0.9
context: ["carte bancaire", "visa", "mastercard", "amex"]
- name: MoneyRecognizer
supported_language: fr
entity_name: MONEY
patterns:
- name: Money Pattern EUR
regex: "(?:EUR|€)\\s*\\d{1,3}(?:[.,\\s]\\d{3})*(?:[.,]\\d{2})?|\\d{1,3}(?:[.,\\s]\\d{3})*(?:[.,]\\d{2})?\\s*(?:EUR|€)"
score: 0.85
- name: BelgianNRNRecognizer
supported_language: fr
entity_name: BE_NATIONAL_REGISTER_NUMBER
patterns:
- name: NRN Pattern
regex: "\\b[0-9]{2}\\.[0-9]{2}\\.[0-9]{2}-[0-9]{3}\\.[0-9]{2}\\b"
score: 1.0
context: ["registre national", "nrn", "niss"]
- name: BelgianEnterpriseRecognizer
supported_language: fr
entity_name: BE_ENTERPRISE_NUMBER
patterns:
- name: BE Enterprise Number
regex: "\\bBE\\s?0\\d{3}[\\.\\s]?\\d{3}[\\.\\s]?\\d{3}\\b"
score: 0.95
context: ["numéro d'entreprise", "btw", "tva", "BCE", "KBO"]
- name: FrenchINSEERecognizer
supported_language: fr
entity_name: FR_SOCIAL_SECURITY_NUMBER
patterns:
- name: INSEE Pattern
regex: "\\b[12][\\s]?[0-9]{2}[\\s]?(?:0[1-9]|1[0-2])[\\s]?(?:2[ABab]|[0-9]{2})[\\s]?[0-9]{3}[\\s]?[0-9]{3}[\\s]?[0-9]{2}\\b"
score: 0.95
context: ["sécurité sociale", "insee", "nir"]
- name: FrenchSIRENSIRETRecognizer
supported_language: fr
entity_name: FR_SIREN_SIRET
patterns:
- name: SIRET Pattern
regex: "\\b[0-9]{3}[\\s]?[0-9]{3}[\\s]?[0-9]{3}[\\s]?[0-9]{5}\\b"
score: 0.9
- name: SIREN Pattern
regex: "\\b[0-9]{3}[\\s]?[0-9]{3}[\\s]?[0-9]{3}\\b"
score: 0.85
context: ["siren", "siret"]
# 3. LISTE D'EXCLUSION
# =====================================================================
allow_list:
- Contrat
- Document
# ... (votre liste complète)
# 4. CONFIGURATION D'ANONYMISATION (ne change pas)
# =====================================================================
# ... (votre section anonymizer_config reste ici)