refactor: split cim10_extractor → bio_normals, bio_extraction, diagnostic_extraction, validation_pipeline

Découpe le monolithe cim10_extractor.py (1356L) en 4 modules spécialisés :
- bio_normals.py : constante BIO_NORMALS + _is_abnormal() (feuille)
- bio_extraction.py : extraction biologie structurée
- diagnostic_extraction.py : extraction DP/DAS/actes CCAM
- validation_pipeline.py : validation CIM-10/CCAM + règles métier

Le cim10_extractor.py reste orchestrateur (~450L) avec re-exports
backward-compat. Imports mis à jour dans clinical_context, rag_search,
fusion. 748 tests passent.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
This commit is contained in:
dom
2026-02-20 10:06:18 +01:00
parent 5823eb6b53
commit e760b12961
8 changed files with 957 additions and 889 deletions

View File

@@ -20,7 +20,7 @@ from ..config import (
Traitement,
)
from ..medical.das_filter import is_valid_diagnostic_text, apply_semantic_dedup
from ..medical.cim10_extractor import _is_dp_family_redundant
from ..medical.validation_pipeline import _is_dp_family_redundant
logger = logging.getLogger(__name__)