Update app.py

This commit is contained in:
Nacim
2025-06-23 15:14:29 +02:00
committed by GitHub
parent f2b3cf40fe
commit 45a2c7a092

28
app.py
View File

@@ -3,15 +3,13 @@ import logging
import yaml import yaml
from flask import Flask, request, jsonify, make_response from flask import Flask, request, jsonify, make_response
# Import des classes nécessaires de Presidio
from presidio_analyzer import AnalyzerEngine, RecognizerRegistry, PatternRecognizer, Pattern from presidio_analyzer import AnalyzerEngine, RecognizerRegistry, PatternRecognizer, Pattern
from presidio_analyzer.nlp_engine import NlpEngineProvider from presidio_analyzer.nlp_engine import NlpEngineProvider
from presidio_analyzer.deny_list_recognizer import DenyListRecognizer
# Configuration du logging pour un meilleur débogage
logging.basicConfig(level=logging.INFO, format='%(asctime)s - %(name)s - %(levelname)s - %(message)s') logging.basicConfig(level=logging.INFO, format='%(asctime)s - %(name)s - %(levelname)s - %(message)s')
logger = logging.getLogger(__name__) logger = logging.getLogger(__name__)
# --- CHARGEMENT MANUEL ET EXPLICITE DE LA CONFIGURATION ---
CONFIG_FILE_PATH = os.environ.get("PRESIDIO_ANALYZER_CONFIG_FILE", "conf/default.yaml") CONFIG_FILE_PATH = os.environ.get("PRESIDIO_ANALYZER_CONFIG_FILE", "conf/default.yaml")
logger.info(f"Loading configuration from: {CONFIG_FILE_PATH}") logger.info(f"Loading configuration from: {CONFIG_FILE_PATH}")
@@ -25,18 +23,22 @@ except FileNotFoundError:
except yaml.YAMLError as e: except yaml.YAMLError as e:
logger.error(f"Error parsing YAML configuration file: {e}. Using default empty config.") logger.error(f"Error parsing YAML configuration file: {e}. Using default empty config.")
# Création du fournisseur de moteur NLP supported_languages_from_config = config.get("supported_languages", ["en"])
logger.info(f"Languages supported according to config: {supported_languages_from_config}")
logger.info("Creating NLP engine provider...") logger.info("Creating NLP engine provider...")
nlp_engine_provider = NlpEngineProvider(nlp_configuration=config.get("nlp_engine_configuration")) nlp_engine_provider = NlpEngineProvider(nlp_configuration=config.get("nlp_engine_configuration"))
nlp_engine = nlp_engine_provider.create_engine() nlp_engine = nlp_engine_provider.create_engine()
logger.info(f"NLP engine created with models for: {nlp_engine.get_supported_languages()}") logger.info(f"NLP engine created with models for: {nlp_engine.get_supported_languages()}")
# Création du registre de recognizers # --- CORRECTION ICI ---
# On s'assure que le registry est initialisé avec les mêmes langues que le reste
logger.info("Creating and populating recognizer registry...") logger.info("Creating and populating recognizer registry...")
registry = RecognizerRegistry() registry = RecognizerRegistry()
registry.load_predefined_recognizers(languages=config.get("supported_languages", ["en"])) registry.load_predefined_recognizers(languages=supported_languages_from_config) # On utilise la variable
logger.info(f"Recognizer registry loaded for languages: {supported_languages_from_config}")
# Ajout des recognizers personnalisés définis dans le YAML # Ajout des recognizers personnalisés
custom_recognizers_conf = config.get("recognizers", []) custom_recognizers_conf = config.get("recognizers", [])
for recognizer_conf in custom_recognizers_conf: for recognizer_conf in custom_recognizers_conf:
patterns = [Pattern(name=p['name'], regex=p['regex'], score=p['score']) for p in recognizer_conf['patterns']] patterns = [Pattern(name=p['name'], regex=p['regex'], score=p['score']) for p in recognizer_conf['patterns']]
@@ -52,18 +54,12 @@ for recognizer_conf in custom_recognizers_conf:
registry.add_recognizer(custom_recognizer) registry.add_recognizer(custom_recognizer)
logger.info(f"Loaded custom recognizer: {custom_recognizer.name}") logger.info(f"Loaded custom recognizer: {custom_recognizer.name}")
# --- CORRECTION DE LA GESTION DE L'ALLOW LIST --- # Préparation de l'allow_list
# On prépare la liste de mots à ignorer. Pas besoin d'importer DenyListRecognizer.
allow_list_config = config.get("allow_list", []) allow_list_config = config.get("allow_list", [])
# On extrait juste les chaînes de caractères de la liste.
allow_list_terms = [item if isinstance(item, str) else item.get('text') for item in allow_list_config if item] allow_list_terms = [item if isinstance(item, str) else item.get('text') for item in allow_list_config if item]
if allow_list_terms: if allow_list_terms:
logger.info(f"Prepared {len(allow_list_terms)} terms for the allow list.") logger.info(f"Prepared {len(allow_list_terms)} terms for the allow list.")
else:
logger.info("No allow list terms found in configuration.")
# Initialisation de l'application Flask
app = Flask(__name__) app = Flask(__name__)
# Initialisation du moteur Presidio Analyzer # Initialisation du moteur Presidio Analyzer
@@ -71,7 +67,7 @@ logger.info("Initializing AnalyzerEngine with custom configuration...")
analyzer = AnalyzerEngine( analyzer = AnalyzerEngine(
nlp_engine=nlp_engine, nlp_engine=nlp_engine,
registry=registry, registry=registry,
supported_languages=config.get("supported_languages", ["en"]), supported_languages=supported_languages_from_config, # On utilise la variable ici aussi
default_score_threshold=config.get("ner_model_configuration", {}).get("confidence_threshold", {}).get("default", 0.35) default_score_threshold=config.get("ner_model_configuration", {}).get("confidence_threshold", {}).get("default", 0.35)
) )
logger.info("AnalyzerEngine initialized successfully.") logger.info("AnalyzerEngine initialized successfully.")
@@ -86,8 +82,6 @@ def analyze_text():
if not text_to_analyze: if not text_to_analyze:
return jsonify({"error": "text field is missing or empty"}), 400 return jsonify({"error": "text field is missing or empty"}), 400
# --- CORRECTION DE L'APPEL A ANALYZE ---
# On passe directement la liste de mots à ignorer au paramètre 'allow_list'
results = analyzer.analyze( results = analyzer.analyze(
text=text_to_analyze, text=text_to_analyze,
language=language, language=language,