Domi31tls
1c44a26eb3
Anonymise toutes les références à des entités réelles (CHCB, Bayonne, Saint-Denis, Réunion, etc.) dans le code source, les configurations YAML, les scripts/outils, et les tests unitaires. Conserve les tests synthétiques (cases) intentionnels. - profile key chcb_strict → chuxx_strict - CHCB → CHUXX, Bayonne → Chicago, Saint-Denis → Springfield, Réunion → Province Bêta, 64100/97400 → 12345, FINESS → 999999999, préfixe tél 05.59.44 → 0X.XX.XX - renomme tools/test_chcb_leak.py → tools/test_force_term_leak.py Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
110 lines
3.0 KiB
Python
Executable File
110 lines
3.0 KiB
Python
Executable File
#!/usr/bin/env python3
|
|
"""Test complet du workflow GUI."""
|
|
|
|
from pathlib import Path
|
|
import sys
|
|
import time
|
|
|
|
# Ajouter le répertoire parent au path
|
|
sys.path.insert(0, str(Path(__file__).parent.parent))
|
|
|
|
import anonymizer_core_refactored_onnx as core
|
|
from config_defaults import RUNTIME_DICTIONARIES_CONFIG_PATH
|
|
|
|
# Dossier de test
|
|
test_dir = Path("/tmp/test_gui_pdfs")
|
|
out_dir = test_dir / "anonymise"
|
|
out_dir.mkdir(exist_ok=True)
|
|
|
|
# Trouver tous les PDFs
|
|
pdfs = sorted([p for p in test_dir.rglob("*.pdf") if p.is_file()])
|
|
print(f"📁 Dossier: {test_dir}")
|
|
print(f"📄 PDFs trouvés: {len(pdfs)}")
|
|
|
|
if not pdfs:
|
|
print("❌ Aucun PDF trouvé")
|
|
sys.exit(1)
|
|
|
|
# Traiter chaque PDF
|
|
start_time = time.time()
|
|
ok = ko = 0
|
|
total_masked = 0
|
|
|
|
for i, pdf in enumerate(pdfs, start=1):
|
|
print(f"\n[{i}/{len(pdfs)}] {pdf.name}")
|
|
|
|
try:
|
|
# Appel identique au GUI
|
|
outputs = core.process_pdf(
|
|
pdf_path=pdf,
|
|
out_dir=out_dir,
|
|
make_vector_redaction=False,
|
|
also_make_raster_burn=True,
|
|
config_path=RUNTIME_DICTIONARIES_CONFIG_PATH,
|
|
use_hf=False,
|
|
ner_manager=None,
|
|
ner_thresholds=None,
|
|
ogc_label=None,
|
|
vlm_manager=None,
|
|
)
|
|
|
|
print(f" ✅ Succès")
|
|
for k, v in outputs.items():
|
|
print(f" - {k}: {Path(v).name}")
|
|
|
|
# Compter les PII
|
|
audit_path = Path(outputs.get("audit", ""))
|
|
if audit_path.exists():
|
|
import json
|
|
pii_count = 0
|
|
with open(audit_path, 'r', encoding='utf-8') as f:
|
|
for line in f:
|
|
try:
|
|
json.loads(line)
|
|
pii_count += 1
|
|
except:
|
|
pass
|
|
print(f" - PII détectés: {pii_count}")
|
|
total_masked += pii_count
|
|
|
|
ok += 1
|
|
except Exception as e:
|
|
print(f" ❌ Erreur: {e}")
|
|
import traceback
|
|
traceback.print_exc()
|
|
ko += 1
|
|
|
|
total_time = time.time() - start_time
|
|
|
|
# Résumé
|
|
print(f"\n{'='*60}")
|
|
print(f"✅ Succès: {ok}")
|
|
print(f"❌ Erreurs: {ko}")
|
|
print(f"🔒 PII masqués: {total_masked}")
|
|
print(f"⏱️ Temps total: {total_time:.1f}s ({total_time/len(pdfs):.1f}s/doc)")
|
|
|
|
# Vérifier les fuites
|
|
import re
|
|
leak_count = 0
|
|
patterns = {
|
|
"date_naissance": re.compile(r"(?:n[ée]+\s+le|DDN)\s*:?\s*\d{1,2}[/.\-]\d{1,2}[/.\-]\d{2,4}", re.IGNORECASE),
|
|
"chuxx": re.compile(r"\bCHUXX\b", re.IGNORECASE),
|
|
}
|
|
|
|
for txt_file in out_dir.glob("*.pseudonymise.txt"):
|
|
with open(txt_file, 'r', encoding='utf-8') as f:
|
|
content = f.read()
|
|
|
|
for pattern_name, pattern in patterns.items():
|
|
matches = pattern.findall(content)
|
|
if matches:
|
|
print(f"⚠️ Fuite {pattern_name} dans {txt_file.name}: {matches}")
|
|
leak_count += len(matches)
|
|
|
|
if leak_count == 0:
|
|
print("🔒 0 fuite détectée")
|
|
else:
|
|
print(f"⚠️ {leak_count} fuite(s) potentielle(s)")
|
|
|
|
print(f"{'='*60}")
|