Files
anonymisation/tools/quick_test_date_correction.py

60 lines
1.9 KiB
Python

#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""Test rapide de la correction DATE"""
import sys
from pathlib import Path
sys.path.insert(0, str(Path(__file__).parent.parent))
from config_defaults import RUNTIME_DICTIONARIES_CONFIG_PATH
from anonymizer_core_refactored_onnx import process_pdf
# Test sur 3 documents du test dataset
test_docs = [
"tests/ground_truth/pdfs/001_simple_compte_rendu_460_23153652_CR_COLOSCOPIE.pdf",
"tests/ground_truth/pdfs/008_moyen_compte_rendu_195_23144210_ANAPATH.pdf",
"tests/ground_truth/pdfs/013_moyen_compte_rendu_363_23085243_CRO.pdf",
]
print("Test correction DATE (Phase 1)")
print("=" * 80)
out_dir = Path("tests/phase1_test_output")
out_dir.mkdir(exist_ok=True)
for doc in test_docs:
pdf_path = Path(doc)
if not pdf_path.exists():
print(f"⚠️ {pdf_path.name}: non trouvé")
continue
try:
result = process_pdf(
pdf_path=pdf_path,
out_dir=out_dir,
make_vector_redaction=False,
also_make_raster_burn=False,
config_path=RUNTIME_DICTIONARIES_CONFIG_PATH,
use_hf=False,
ner_manager=None,
vlm_manager=None,
)
# Lire le fichier texte anonymisé
text_file = out_dir / f"{pdf_path.stem}.pseudonymise.txt"
if text_file.exists():
text = text_file.read_text(encoding='utf-8')
date_count = text.count("[DATE]")
date_naissance_count = text.count("[DATE_NAISSANCE]")
status = "" if date_count == 0 else ""
print(f"{status} {pdf_path.name}")
print(f" [DATE]: {date_count} (attendu: 0)")
print(f" [DATE_NAISSANCE]: {date_naissance_count}")
else:
print(f"⚠️ {pdf_path.name}: fichier texte non trouvé")
except Exception as e:
print(f"{pdf_path.name}: Erreur - {e}")
print("\n✅ Test terminé")