nacim/Presidio
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Update default.yaml

Browse Source
This commit is contained in:
Nacim
2025-06-24 13:34:12 +02:00
committed by GitHub
parent 7dc1b6b89f
commit 2170fea7ab
1 changed files with 7 additions and 8 deletions
Download Patch File Download Diff File Expand all files Collapse all files

15
conf/default.yaml
View File

@@ -1,5 +1,5 @@
# ===================================================================== # =====================================================================
# CONFIGURATION PRESIDIO - v24 (FINALE, COMPLÈTE ET FIABLE) # CONFIGURATION PRESIDIO - v24 (COMPLÈTE, AVEC VOTRE REGEX TÉLÉPHONE)
# ===================================================================== # =====================================================================
supported_languages: [en, fr] supported_languages: [en, fr]
@@ -52,13 +52,14 @@ recognizer_registry:
patterns: [{name: Email Pattern, regex: "\\b[A-Za-z0-9._%+-]+@[A-Za-z0-9.-]+\\.[A-Z|a-z]{2,}\\b", score: 1.0}] 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"] context: ["email", "courriel", "mail"]
# --- CORRECTION DÉFINITIVE : Création d'une entité personnalisée pour le téléphone --- # --- VOTRE REGEX TÉLÉPHONE, ADAPTÉE POUR PRESIDIO ---
- name: CustomBelgianPhoneRecognizer - name: CustomBelgianPhoneRecognizer
supported_language: fr supported_language: fr
supported_entity: CUSTOM_PHONE_BE # Nouvelle entité pour éviter les conflits supported_entity: CUSTOM_PHONE_BE
patterns: patterns:
- name: Belgian Phone Pattern (+32 + 8 or 9 digits) - name: Belgian Phone Pattern (votre regex adaptée)
regex: "\\b(((\\+|00)32[ ]?(?:\\(0\\)[ ]?)?)|0){1}(4(60|[789]\\d)\\/?(\\s?\\d{2}\\.?){2}(\\s?\\d{2})|(\\d\\/?\\s?\\d{3}|\\d{2}\\/?\\s?\\d{2})(\\.?\\s?\\d{2}){2})\\b" # J'ai doublé les backslashes et remplacé ^/$ par \b
regex: "\\b(((\\+|00)32[ ]?(?:\\(0\\)[ ]?)?)|0){1}(4(60|[789]\\d)\\/(\\s?\\d{2}\\.?){2}(\\s?\\d{2})|(\\d\\/?\\s?\\d{3}|\\d{2}\\/?\\s?\\d{2})(\\.?\\s?\\d{2}){2})\\b"
score: 1.0 score: 1.0
context: ["téléphone", "tel", "gsm", "mobile"] context: ["téléphone", "tel", "gsm", "mobile"]
@@ -113,7 +114,7 @@ recognizer_registry:
supported_entity: FR_SOCIAL_SECURITY_NUMBER supported_entity: FR_SOCIAL_SECURITY_NUMBER
patterns: patterns:
- name: INSEE Pattern with flexible spaces - name: INSEE Pattern with flexible spaces
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" 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 score: 0.95
context: ["sécurité sociale", "insee", "nir"] context: ["sécurité sociale", "insee", "nir"]
@@ -183,7 +184,6 @@ anonymizer_config:
BE_NATIONAL_REGISTER_NUMBER: replace BE_NATIONAL_REGISTER_NUMBER: replace
FR_SOCIAL_SECURITY_NUMBER: replace FR_SOCIAL_SECURITY_NUMBER: replace
SWIFT_CODE: replace SWIFT_CODE: replace
# --- CORRECTION DÉFINITIVE : On anonymise notre nouvelle entité téléphone ---
CUSTOM_PHONE_BE: replace CUSTOM_PHONE_BE: replace
replacements: replacements:
@@ -200,5 +200,4 @@ anonymizer_config:
BE_NATIONAL_REGISTER_NUMBER: "<NRN_BELGE>" BE_NATIONAL_REGISTER_NUMBER: "<NRN_BELGE>"
FR_SOCIAL_SECURITY_NUMBER: "<NUM_SECU_FR>" FR_SOCIAL_SECURITY_NUMBER: "<NUM_SECU_FR>"
SWIFT_CODE: "<CODE_SWIFT>" SWIFT_CODE: "<CODE_SWIFT>"
# --- CORRECTION DÉFINITIVE : On ajoute le token de remplacement ---
CUSTOM_PHONE_BE: "<TELEPHONE>" CUSTOM_PHONE_BE: "<TELEPHONE>"