feat: grounding CPAM — tags DP/DAS/ANT/COMPL + fuzzy matching CIM-10 + prompt renforcé
Cause racine du Tier C : le LLM inventait des tags ([C83.3], [Antécédents]) car _build_tagged_context() ne taguait que bio/img/trt/actes. Le DP, les DAS, antécédents et complications n'avaient aucun tag citable. - cpam_context: 4 nouveaux types de tags [DP], [DAS-N], [ANT-N], [COMPL-N] - cpam_validation: fuzzy matching — résout les refs CIM-10 nues vers le tag contenant ce code - templates: liste explicite des tags valides, interdiction d'inventer des tags - tests: 18 nouveaux tests (tags, fuzzy match, grounding DAS/DP) Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
This commit is contained in:
dom
@@ -6,7 +6,9 @@ import pytest
|
||||
|
||||
from src.config import (
|
||||
ActeCCAM,
|
||||
Antecedent,
|
||||
BiologieCle,
|
||||
Complication,
|
||||
ControleCPAM,
|
||||
Diagnostic,
|
||||
DossierMedical,
|
||||
@@ -23,6 +25,7 @@ from src.control.cpam_response import (
|
||||
_check_das_bio_coherence,
|
||||
_extraction_pass,
|
||||
_format_response,
|
||||
_fuzzy_match_ref,
|
||||
_get_cim10_definitions,
|
||||
_get_code_label,
|
||||
_search_rag_for_control,
|
||||
@@ -854,15 +857,21 @@ class TestBuildTaggedContext:
|
||||
assert "BAS" in tag_map.get("BIO-3", "")
|
||||
|
||||
def test_tagged_context_empty_dossier(self):
|
||||
"""Dossier sans données cliniques → texte vide, tag_map vide."""
|
||||
"""Dossier sans aucune donnée clinique → texte vide, tag_map vide."""
|
||||
dossier = DossierMedical(source_file="test.pdf")
|
||||
text, tag_map = _build_tagged_context(dossier)
|
||||
assert text == ""
|
||||
assert tag_map == {}
|
||||
|
||||
def test_tagged_context_dp_only_dossier(self):
|
||||
"""Dossier avec DP mais sans bio/img/trt → tag [DP] généré."""
|
||||
dossier = DossierMedical(
|
||||
source_file="test.pdf",
|
||||
diagnostic_principal=Diagnostic(texte="Test", cim10_suggestion="Z00"),
|
||||
)
|
||||
text, tag_map = _build_tagged_context(dossier)
|
||||
|
||||
assert text == ""
|
||||
assert tag_map == {}
|
||||
assert "DP" in tag_map
|
||||
assert "[DP]" in text
|
||||
|
||||
def test_tagged_context_in_prompt(self):
|
||||
"""Le contexte tagué apparaît dans le prompt généré."""
|
||||
@@ -876,10 +885,9 @@ class TestBuildTaggedContext:
|
||||
assert len(tag_map) > 0
|
||||
|
||||
def test_poor_dossier_warning_in_prompt(self):
|
||||
"""Dossier sans bio/imagerie → avertissement dans le prompt."""
|
||||
"""Dossier totalement vide → avertissement DOSSIER PAUVRE dans le prompt."""
|
||||
dossier = DossierMedical(
|
||||
source_file="test.pdf",
|
||||
diagnostic_principal=Diagnostic(texte="Test", cim10_suggestion="Z00"),
|
||||
sejour=Sejour(sexe="M", age=70),
|
||||
)
|
||||
controle = _make_controle()
|
||||
@@ -889,6 +897,19 @@ class TestBuildTaggedContext:
|
||||
assert "Ne spécule PAS" in prompt
|
||||
assert len(tag_map) == 0
|
||||
|
||||
def test_dp_only_dossier_not_poor(self):
|
||||
"""Dossier avec DP mais sans bio/img → PAS de warning DOSSIER PAUVRE (DP génère un tag)."""
|
||||
dossier = DossierMedical(
|
||||
source_file="test.pdf",
|
||||
diagnostic_principal=Diagnostic(texte="Test", cim10_suggestion="Z00"),
|
||||
sejour=Sejour(sexe="M", age=70),
|
||||
)
|
||||
controle = _make_controle()
|
||||
prompt, tag_map = _build_cpam_prompt(dossier, controle, [])
|
||||
|
||||
assert "DOSSIER PAUVRE" not in prompt
|
||||
assert "DP" in tag_map
|
||||
|
||||
|
||||
class TestValidateGrounding:
|
||||
"""Tests pour la validation des preuves grounded."""
|
||||
@@ -1996,3 +2017,175 @@ class TestCodesAutorisesWhitelist:
|
||||
controle = _make_controle()
|
||||
prompt, _ = _build_cpam_prompt(dossier, controle, [])
|
||||
assert "Ne mentionne AUCUN code CIM-10 qui ne figure pas" in prompt
|
||||
|
||||
|
||||
class TestTaggedContextNewTags:
|
||||
"""Tests pour les tags DP, DAS-N, ANT-N, COMPL-N dans _build_tagged_context()."""
|
||||
|
||||
def test_dp_tag_generated(self):
|
||||
"""Le tag [DP] est généré pour le diagnostic principal."""
|
||||
dossier = DossierMedical(
|
||||
source_file="test.pdf",
|
||||
diagnostic_principal=Diagnostic(texte="Cholécystite aiguë", cim10_suggestion="K81.0"),
|
||||
biologie_cle=[BiologieCle(test="CRP", valeur="180 mg/L")],
|
||||
)
|
||||
text, tag_map = _build_tagged_context(dossier)
|
||||
assert "DP" in tag_map
|
||||
assert "Cholécystite aiguë (K81.0)" in tag_map["DP"]
|
||||
assert "[DP]" in text
|
||||
|
||||
def test_dp_without_code(self):
|
||||
"""Le tag [DP] fonctionne même sans code CIM-10."""
|
||||
dossier = DossierMedical(
|
||||
source_file="test.pdf",
|
||||
diagnostic_principal=Diagnostic(texte="Infection urinaire"),
|
||||
biologie_cle=[BiologieCle(test="CRP", valeur="50 mg/L")],
|
||||
)
|
||||
text, tag_map = _build_tagged_context(dossier)
|
||||
assert "DP" in tag_map
|
||||
assert "Infection urinaire" in tag_map["DP"]
|
||||
assert "()" not in tag_map["DP"]
|
||||
|
||||
def test_das_tags_generated(self):
|
||||
"""Les tags [DAS-1], [DAS-2] sont générés pour les diagnostics associés."""
|
||||
dossier = DossierMedical(
|
||||
source_file="test.pdf",
|
||||
diagnostic_principal=Diagnostic(texte="DP test"),
|
||||
diagnostics_associes=[
|
||||
Diagnostic(texte="Iléus réflexe", cim10_suggestion="K56.0"),
|
||||
Diagnostic(texte="HTA", cim10_suggestion="I10"),
|
||||
],
|
||||
)
|
||||
text, tag_map = _build_tagged_context(dossier)
|
||||
assert "DAS-1" in tag_map
|
||||
assert "DAS-2" in tag_map
|
||||
assert "K56.0" in tag_map["DAS-1"]
|
||||
assert "[DAS-1]" in text
|
||||
assert "[DAS-2]" in text
|
||||
|
||||
def test_ant_tags_generated(self):
|
||||
"""Les tags [ANT-1], [ANT-2] sont générés pour les antécédents."""
|
||||
dossier = DossierMedical(
|
||||
source_file="test.pdf",
|
||||
diagnostic_principal=Diagnostic(texte="DP test"),
|
||||
antecedents=[
|
||||
Antecedent(texte="Diabète type 2"),
|
||||
Antecedent(texte="HTA"),
|
||||
],
|
||||
)
|
||||
text, tag_map = _build_tagged_context(dossier)
|
||||
assert "ANT-1" in tag_map
|
||||
assert "ANT-2" in tag_map
|
||||
assert "Diabète type 2" in tag_map["ANT-1"]
|
||||
assert "[ANT-1]" in text
|
||||
assert "[ANT-2]" in text
|
||||
|
||||
def test_ant_tags_capped_at_10(self):
|
||||
"""Les antécédents sont limités à 10 tags maximum."""
|
||||
dossier = DossierMedical(
|
||||
source_file="test.pdf",
|
||||
diagnostic_principal=Diagnostic(texte="DP test"),
|
||||
antecedents=[Antecedent(texte=f"Antécédent {i}") for i in range(15)],
|
||||
)
|
||||
_, tag_map = _build_tagged_context(dossier)
|
||||
assert "ANT-10" in tag_map
|
||||
assert "ANT-11" not in tag_map
|
||||
|
||||
def test_compl_tags_generated(self):
|
||||
"""Les tags [COMPL-1] sont générés pour les complications."""
|
||||
dossier = DossierMedical(
|
||||
source_file="test.pdf",
|
||||
diagnostic_principal=Diagnostic(texte="DP test"),
|
||||
complications=[
|
||||
Complication(texte="Infection de paroi"),
|
||||
Complication(texte="Hémorragie post-op"),
|
||||
],
|
||||
)
|
||||
text, tag_map = _build_tagged_context(dossier)
|
||||
assert "COMPL-1" in tag_map
|
||||
assert "COMPL-2" in tag_map
|
||||
assert "Infection de paroi" in tag_map["COMPL-1"]
|
||||
assert "[COMPL-1]" in text
|
||||
|
||||
def test_all_new_tags_in_complet_dossier(self):
|
||||
"""Un dossier complet génère tous les types de tags."""
|
||||
dossier = DossierMedical(
|
||||
source_file="test.pdf",
|
||||
diagnostic_principal=Diagnostic(texte="Cholécystite", cim10_suggestion="K81.0"),
|
||||
diagnostics_associes=[Diagnostic(texte="Iléus", cim10_suggestion="K56.0")],
|
||||
biologie_cle=[BiologieCle(test="CRP", valeur="180 mg/L")],
|
||||
imagerie=[Imagerie(type="Scanner")],
|
||||
traitements_sortie=[Traitement(medicament="Augmentin")],
|
||||
actes_ccam=[ActeCCAM(texte="Cholécystectomie")],
|
||||
antecedents=[Antecedent(texte="HTA")],
|
||||
complications=[Complication(texte="Hémorragie")],
|
||||
)
|
||||
text, tag_map = _build_tagged_context(dossier)
|
||||
for expected_tag in ["BIO-1", "IMG-1", "TRT-1", "ACTE-1", "DP", "DAS-1", "ANT-1", "COMPL-1"]:
|
||||
assert expected_tag in tag_map, f"Tag {expected_tag} manquant"
|
||||
|
||||
def test_grounding_das_ref_valid(self):
|
||||
"""Ref DAS-1 dans preuves_dossier → pas de warning."""
|
||||
tag_map = {"DAS-1": "Iléus réflexe (K56.0)", "DP": "Cholécystite (K81.0)"}
|
||||
response_data = {
|
||||
"preuves_dossier": [
|
||||
{"ref": "DAS-1", "element": "diagnostic", "valeur": "Iléus réflexe", "signification": "DAS justifié"},
|
||||
{"ref": "DP", "element": "diagnostic", "valeur": "Cholécystite", "signification": "DP confirmé"},
|
||||
]
|
||||
}
|
||||
warnings = _validate_grounding(response_data, tag_map)
|
||||
assert len(warnings) == 0
|
||||
|
||||
|
||||
class TestFuzzyMatchRef:
|
||||
"""Tests pour le fuzzy matching de refs CIM-10 dans _fuzzy_match_ref()."""
|
||||
|
||||
def test_cim10_code_matches_das_content(self):
|
||||
"""Un code CIM-10 nu (C83.3) est résolu vers le DAS qui le contient."""
|
||||
tag_map = {
|
||||
"DAS-1": "Lymphome folliculaire (C83.3)",
|
||||
"BIO-1": "CRP: 180 mg/L",
|
||||
}
|
||||
result = _fuzzy_match_ref("C83.3", tag_map)
|
||||
assert result == "DAS-1"
|
||||
|
||||
def test_cim10_code_matches_dp(self):
|
||||
"""Un code CIM-10 résolu vers le tag DP."""
|
||||
tag_map = {"DP": "Cholécystite aiguë (K81.0)"}
|
||||
result = _fuzzy_match_ref("K81.0", tag_map)
|
||||
assert result == "DP"
|
||||
|
||||
def test_cim10_code_no_match(self):
|
||||
"""Un code CIM-10 absent du tag_map → None."""
|
||||
tag_map = {"BIO-1": "CRP: 180 mg/L", "DAS-1": "Iléus (K56.0)"}
|
||||
result = _fuzzy_match_ref("Z45.8", tag_map)
|
||||
assert result is None
|
||||
|
||||
def test_non_cim10_ref_no_match(self):
|
||||
"""Une ref non-CIM-10 (ex: 'Antécédents') → None."""
|
||||
tag_map = {"ANT-1": "HTA", "DP": "Test (K81.0)"}
|
||||
result = _fuzzy_match_ref("Antécédents", tag_map)
|
||||
assert result is None
|
||||
|
||||
def test_grounding_fuzzy_resolves_cim10(self):
|
||||
"""_validate_grounding résout une ref CIM-10 via fuzzy matching → pas de warning."""
|
||||
tag_map = {"DAS-1": "Lymphome (C83.3)", "BIO-1": "CRP: 180"}
|
||||
response_data = {
|
||||
"preuves_dossier": [
|
||||
{"ref": "C83.3", "element": "clinique", "valeur": "Lymphome", "signification": "onco"},
|
||||
]
|
||||
}
|
||||
warnings = _validate_grounding(response_data, tag_map)
|
||||
assert len(warnings) == 0
|
||||
|
||||
def test_grounding_category_name_still_warns(self):
|
||||
"""Une ref catégorielle ('Antécédents') n'est pas résolue → warning maintenu."""
|
||||
tag_map = {"ANT-1": "HTA", "BIO-1": "CRP: 5"}
|
||||
response_data = {
|
||||
"preuves_dossier": [
|
||||
{"ref": "Antécédents", "element": "clinique", "valeur": "HTA", "signification": "contexte"},
|
||||
]
|
||||
}
|
||||
warnings = _validate_grounding(response_data, tag_map)
|
||||
assert len(warnings) == 1
|
||||
assert "Antécédents" in warnings[0]
|
||||
|
||||
Reference in New Issue
Block a user