feat: auth fail-closed, regles par defaut idempotentes et conteneur non root (tt-2026-11-061)
This commit is contained in:
1 parent
b51a8bba69
commit
6aef9e83e7
11 files changed
+424
-95
No files matched your search
@@ -20,3 +20,4 @@ htmlcov/
|
|||||||
.env
|
.env
|
||||||
.venv/
|
.venv/
|
||||||
venv/
|
venv/
|
||||||
|
*.bak
|
||||||
@@ -17,8 +17,15 @@ COPY requirements.txt .
|
|||||||
RUN pip install --no-cache-dir -r requirements.txt
|
RUN pip install --no-cache-dir -r requirements.txt
|
||||||
|
|
||||||
COPY app/ /app/app/
|
COPY app/ /app/app/
|
||||||
|
COPY tests/ /app/tests/
|
||||||
COPY README.md /app/
|
COPY README.md /app/
|
||||||
|
|
||||||
|
# Préparation du répertoire persistant data/ et droits non-root
|
||||||
|
RUN mkdir -p /app/data && chown -R 1037:100 /app
|
||||||
|
|
||||||
|
# Exécution en conteneur non root (uid=1037, gid=100)
|
||||||
|
USER 1037:100
|
||||||
|
|
||||||
EXPOSE 8000
|
EXPOSE 8000
|
||||||
|
|
||||||
CMD ["uvicorn", "app.main:app", "--host", "0.0.0.0", "--port", "8000"]
|
CMD ["uvicorn", "app.main:app", "--host", "0.0.0.0", "--port", "8000"]
|
||||||
+38
-35
@@ -5,15 +5,36 @@ from typing import Dict, Any, Optional
|
|||||||
from fastapi import Request, HTTPException, Security
|
from fastapi import Request, HTTPException, Security
|
||||||
from fastapi.security import APIKeyHeader
|
from fastapi.security import APIKeyHeader
|
||||||
import jwt
|
import jwt
|
||||||
from app.config import CF_ACCESS_CERTS_URL, CF_ACCESS_AUD, CF_TEST_SECRET
|
from app.config import ENV, CF_ACCESS_CERTS_URL, CF_ACCESS_AUD, CF_TEST_SECRET
|
||||||
|
|
||||||
cf_access_header = APIKeyHeader(name="Cf-Access-Jwt-Assertion", auto_error=False)
|
cf_access_header = APIKeyHeader(name="Cf-Access-Jwt-Assertion", auto_error=False)
|
||||||
|
|
||||||
def verify_cf_access_jwt(request: Request) -> Dict[str, Any]:
|
def verify_cf_access_jwt(request: Request) -> Dict[str, Any]:
|
||||||
"""
|
"""
|
||||||
Vérifie la présence et la validité du jeton Cloudflare Access.
|
Vérifie la présence et la validité de la signature du jeton Cloudflare Access.
|
||||||
Rejette toute requête sans jeton ou avec un jeton invalide (HTTP 401).
|
Fail-closed : aucun décodage sans vérification de signature n'est autorisé.
|
||||||
|
- Si l'environnement n'est pas configuré avec les clés nécessaires : HTTP 503.
|
||||||
|
- Le mode CF_TEST_SECRET n'existe que si ENV='test'.
|
||||||
|
- Toute absence de jeton, jeton forgé (alg none ou fausse signature) ou invalide : HTTP 401.
|
||||||
"""
|
"""
|
||||||
|
is_test_env = (ENV == "test")
|
||||||
|
|
||||||
|
# 1. Contrôle préalable de la configuration (Fail-closed)
|
||||||
|
# Sans clés de vérification, le service répond 503 sur les routes API
|
||||||
|
if is_test_env and CF_TEST_SECRET:
|
||||||
|
# Configuration valide pour le mode test
|
||||||
|
pass
|
||||||
|
elif CF_ACCESS_CERTS_URL and CF_ACCESS_AUD:
|
||||||
|
# Configuration valide pour la production (JWKS Cloudflare Access)
|
||||||
|
pass
|
||||||
|
else:
|
||||||
|
# Clés obligatoires manquantes : refus immédiat de servir les routes /api
|
||||||
|
raise HTTPException(
|
||||||
|
status_code=503,
|
||||||
|
detail="Authentification Cloudflare Access non configurée : CF_ACCESS_CERTS_URL et CF_ACCESS_AUD obligatoires"
|
||||||
|
)
|
||||||
|
|
||||||
|
# 2. Présence obligatoire du jeton Cloudflare Access
|
||||||
token = request.headers.get("Cf-Access-Jwt-Assertion")
|
token = request.headers.get("Cf-Access-Jwt-Assertion")
|
||||||
if not token or not token.strip():
|
if not token or not token.strip():
|
||||||
raise HTTPException(
|
raise HTTPException(
|
||||||
@@ -23,25 +44,25 @@ def verify_cf_access_jwt(request: Request) -> Dict[str, Any]:
|
|||||||
|
|
||||||
token = token.strip()
|
token = token.strip()
|
||||||
|
|
||||||
# 1. Mode test avec secret HMAC (utilisé dans les tests unitaires et d'intégration)
|
# 3. Mode test avec secret HMAC (strictement restreint à ENV=test)
|
||||||
if CF_TEST_SECRET:
|
if is_test_env and CF_TEST_SECRET:
|
||||||
try:
|
try:
|
||||||
payload = jwt.decode(
|
payload = jwt.decode(
|
||||||
token,
|
token,
|
||||||
CF_TEST_SECRET,
|
CF_TEST_SECRET,
|
||||||
algorithms=["HS256"],
|
algorithms=["HS256"],
|
||||||
audience=CF_ACCESS_AUD if CF_ACCESS_AUD else None,
|
audience=CF_ACCESS_AUD if CF_ACCESS_AUD else None,
|
||||||
options={"verify_aud": bool(CF_ACCESS_AUD)}
|
options={"verify_aud": bool(CF_ACCESS_AUD), "verify_signature": True}
|
||||||
)
|
)
|
||||||
return payload
|
return payload
|
||||||
except jwt.PyJWTError as e:
|
except jwt.PyJWTError as e:
|
||||||
raise HTTPException(
|
raise HTTPException(
|
||||||
status_code=401,
|
status_code=401,
|
||||||
detail=f"Jeton Cloudflare Access invalide : {str(e)}"
|
detail=f"Jeton Cloudflare Access invalide ou signature non vérifiée : {str(e)}"
|
||||||
)
|
)
|
||||||
|
|
||||||
# 2. Mode production avec JWKS Cloudflare Access si configuré
|
# 4. Mode production avec vérification de signature JWKS Cloudflare Access
|
||||||
if CF_ACCESS_CERTS_URL:
|
if CF_ACCESS_CERTS_URL and CF_ACCESS_AUD:
|
||||||
try:
|
try:
|
||||||
jwks_client = jwt.PyJWKClient(CF_ACCESS_CERTS_URL)
|
jwks_client = jwt.PyJWKClient(CF_ACCESS_CERTS_URL)
|
||||||
signing_key = jwks_client.get_signing_key_from_jwt(token)
|
signing_key = jwks_client.get_signing_key_from_jwt(token)
|
||||||
@@ -49,36 +70,18 @@ def verify_cf_access_jwt(request: Request) -> Dict[str, Any]:
|
|||||||
token,
|
token,
|
||||||
signing_key.key,
|
signing_key.key,
|
||||||
algorithms=["RS256"],
|
algorithms=["RS256"],
|
||||||
audience=CF_ACCESS_AUD if CF_ACCESS_AUD else None,
|
audience=CF_ACCESS_AUD,
|
||||||
options={"verify_aud": bool(CF_ACCESS_AUD)}
|
options={"verify_aud": True, "verify_signature": True}
|
||||||
)
|
)
|
||||||
return payload
|
return payload
|
||||||
except Exception as e:
|
except Exception as e:
|
||||||
raise HTTPException(
|
raise HTTPException(
|
||||||
status_code=401,
|
status_code=401,
|
||||||
detail=f"Jeton Cloudflare Access non vérifié par JWKS : {str(e)}"
|
detail=f"Jeton Cloudflare Access invalide ou signature non vérifiée : {str(e)}"
|
||||||
)
|
)
|
||||||
|
|
||||||
# 3. Mode par défaut : vérification de structure et d'expiration
|
# Si aucun mode n'a pu valider la signature (sécurité absolue)
|
||||||
try:
|
raise HTTPException(
|
||||||
# Décodage des en-têtes et du payload sans clé (avant configuration finale par Claude)
|
status_code=401,
|
||||||
unverified_header = jwt.get_unverified_header(token)
|
detail="Jeton Cloudflare Access rejeté : signature non vérifiée"
|
||||||
if not unverified_header:
|
)
|
||||||
raise ValueError("En-tête JWT manquant")
|
|
||||||
payload = jwt.decode(
|
|
||||||
token,
|
|
||||||
options={"verify_signature": False, "verify_exp": True}
|
|
||||||
)
|
|
||||||
if not payload or not isinstance(payload, dict):
|
|
||||||
raise ValueError("Payload JWT non dictionnaire")
|
|
||||||
return payload
|
|
||||||
except jwt.ExpiredSignatureError:
|
|
||||||
raise HTTPException(
|
|
||||||
status_code=401,
|
|
||||||
detail="Jeton Cloudflare Access expiré"
|
|
||||||
)
|
|
||||||
except Exception as e:
|
|
||||||
raise HTTPException(
|
|
||||||
status_code=401,
|
|
||||||
detail=f"Jeton Cloudflare Access malformé : {str(e)}"
|
|
||||||
)
|
|
||||||
+77
-31
@@ -1,5 +1,7 @@
|
|||||||
import re
|
import re
|
||||||
|
import json
|
||||||
from typing import Optional, Dict, Any, List, Tuple
|
from typing import Optional, Dict, Any, List, Tuple
|
||||||
|
from app.config import get_tunis_now_iso
|
||||||
from app.db import (
|
from app.db import (
|
||||||
list_regles,
|
list_regles,
|
||||||
get_regle,
|
get_regle,
|
||||||
@@ -21,7 +23,6 @@ def extract_keywords(text: Optional[str]) -> str:
|
|||||||
"""Extrait les mots significatifs d'un objet de mail."""
|
"""Extrait les mots significatifs d'un objet de mail."""
|
||||||
if not text:
|
if not text:
|
||||||
return ""
|
return ""
|
||||||
# Nettoyage des préfixes et caractères spéciaux
|
|
||||||
clean = re.sub(r"^(?:re|fwd|tr)\s*:\s*", "", text, flags=re.IGNORECASE)
|
clean = re.sub(r"^(?:re|fwd|tr)\s*:\s*", "", text, flags=re.IGNORECASE)
|
||||||
words = re.findall(r"\b[\w\u00C0-\u017F-]{3,}\b", clean, flags=re.UNICODE)
|
words = re.findall(r"\b[\w\u00C0-\u017F-]{3,}\b", clean, flags=re.UNICODE)
|
||||||
meaningful = [w for w in words if w.lower() not in STOPWORDS]
|
meaningful = [w for w in words if w.lower() not in STOPWORDS]
|
||||||
@@ -35,7 +36,7 @@ def match_mail_to_rules(mail: Dict[str, Any], rules: List[Dict[str, Any]]) -> st
|
|||||||
mail_exp = (mail.get("expediteur") or mail.get("from") or "").lower()
|
mail_exp = (mail.get("expediteur") or mail.get("from") or "").lower()
|
||||||
mail_obj = (mail.get("objet") or mail.get("subject") or "").lower()
|
mail_obj = (mail.get("objet") or mail.get("subject") or "").lower()
|
||||||
mail_snip = (mail.get("apercu") or mail.get("snippet") or "").lower()
|
mail_snip = (mail.get("apercu") or mail.get("snippet") or "").lower()
|
||||||
mail_tags = [t.lower() for t in (mail.get("etiquettes") or mail.get("tags") or [])]
|
mail_tags = [t.strip().lower() for t in (mail.get("etiquettes") or mail.get("tags") or [])]
|
||||||
|
|
||||||
for rule in rules:
|
for rule in rules:
|
||||||
if rule.get("statut") != "active":
|
if rule.get("statut") != "active":
|
||||||
@@ -51,12 +52,11 @@ def match_mail_to_rules(mail: Dict[str, Any], rules: List[Dict[str, Any]]) -> st
|
|||||||
if rule_exp and rule_exp in mail_exp:
|
if rule_exp and rule_exp in mail_exp:
|
||||||
return rule["categorie"]
|
return rule["categorie"]
|
||||||
|
|
||||||
# 3. Correspondance sur les mots-clés
|
# 3. Correspondance sur les mots-clés (avec frontières de mots pour éviter les faux-positifs)
|
||||||
rule_mot = (rule.get("mot_cle") or "").strip().lower()
|
rule_mot = (rule.get("mot_cle") or "").strip().lower()
|
||||||
if rule_mot:
|
if rule_mot:
|
||||||
# Vérifier si l'un des mots-clés de la règle est présent
|
pattern = rf"\b{re.escape(rule_mot)}\b"
|
||||||
mot_parts = rule_mot.split()
|
if re.search(pattern, mail_obj, re.IGNORECASE) or re.search(pattern, mail_snip, re.IGNORECASE):
|
||||||
if any(p in mail_obj or p in mail_snip for p in mot_parts):
|
|
||||||
return rule["categorie"]
|
return rule["categorie"]
|
||||||
|
|
||||||
return "À classer"
|
return "À classer"
|
||||||
@@ -78,12 +78,16 @@ def classify_mail(mail: Dict[str, Any], db_path: Optional[str] = None) -> Dict[s
|
|||||||
rules = list_regles(statut="active", db_path=db_path)
|
rules = list_regles(statut="active", db_path=db_path)
|
||||||
categorie = match_mail_to_rules(mail, rules)
|
categorie = match_mail_to_rules(mail, rules)
|
||||||
|
|
||||||
|
tags = mail.get("etiquettes") or mail.get("tags") or []
|
||||||
|
tags_json = json.dumps(tags) if isinstance(tags, list) else str(tags)
|
||||||
|
|
||||||
return save_classement_regle(
|
return save_classement_regle(
|
||||||
message_id=mid,
|
message_id=mid,
|
||||||
categorie=categorie,
|
categorie=categorie,
|
||||||
expediteur=mail.get("expediteur") or mail.get("from") or "",
|
expediteur=mail.get("expediteur") or mail.get("from") or "",
|
||||||
objet=mail.get("objet") or mail.get("subject") or "",
|
objet=mail.get("objet") or mail.get("subject") or "",
|
||||||
date_mail=mail.get("date") or "",
|
date_mail=mail.get("date") or "",
|
||||||
|
etiquettes=tags_json,
|
||||||
db_path=db_path
|
db_path=db_path
|
||||||
)
|
)
|
||||||
|
|
||||||
@@ -104,6 +108,9 @@ def manual_classify_and_propose(
|
|||||||
obj = (mail_info.get("objet") if mail_info else None) or (existing.get("objet") if existing else "")
|
obj = (mail_info.get("objet") if mail_info else None) or (existing.get("objet") if existing else "")
|
||||||
date_m = (mail_info.get("date") if mail_info else None) or (existing.get("date_mail") if existing else "")
|
date_m = (mail_info.get("date") if mail_info else None) or (existing.get("date_mail") if existing else "")
|
||||||
|
|
||||||
|
tags = (mail_info.get("etiquettes") if mail_info else None) or (existing.get("etiquettes") if existing else "[]")
|
||||||
|
tags_json = json.dumps(tags) if isinstance(tags, list) else str(tags)
|
||||||
|
|
||||||
classement = save_classement_manuel(
|
classement = save_classement_manuel(
|
||||||
message_id=message_id,
|
message_id=message_id,
|
||||||
categorie=new_category,
|
categorie=new_category,
|
||||||
@@ -111,6 +118,7 @@ def manual_classify_and_propose(
|
|||||||
expediteur=exp,
|
expediteur=exp,
|
||||||
objet=obj,
|
objet=obj,
|
||||||
date_mail=date_m,
|
date_mail=date_m,
|
||||||
|
etiquettes=tags_json,
|
||||||
db_path=db_path
|
db_path=db_path
|
||||||
)
|
)
|
||||||
|
|
||||||
@@ -128,6 +136,68 @@ def manual_classify_and_propose(
|
|||||||
|
|
||||||
return classement, proposed_rule
|
return classement, proposed_rule
|
||||||
|
|
||||||
|
def reclassify_all_non_manual(db_path: Optional[str] = None, date_prefix: Optional[str] = None) -> int:
|
||||||
|
"""
|
||||||
|
Réévalue et reclasse tous les emails non-manuels enregistrés en base.
|
||||||
|
RÈGLE ABSOLUE : Les classements manuels ne sont jamais touchés.
|
||||||
|
"""
|
||||||
|
active_rules = list_regles(statut="active", db_path=db_path)
|
||||||
|
all_classements = list_classements(date_prefix=date_prefix, db_path=db_path)
|
||||||
|
reclassified_count = 0
|
||||||
|
|
||||||
|
for c in all_classements:
|
||||||
|
if c.get("source") == "manuel":
|
||||||
|
continue
|
||||||
|
|
||||||
|
raw_tags = c.get("etiquettes") or "[]"
|
||||||
|
try:
|
||||||
|
tags = json.loads(raw_tags) if isinstance(raw_tags, str) else raw_tags
|
||||||
|
except Exception:
|
||||||
|
tags = []
|
||||||
|
|
||||||
|
mail_repr = {
|
||||||
|
"message_id": c["message_id"],
|
||||||
|
"expediteur": c.get("expediteur", ""),
|
||||||
|
"objet": c.get("objet", ""),
|
||||||
|
"apercu": "",
|
||||||
|
"etiquettes": tags if isinstance(tags, list) else []
|
||||||
|
}
|
||||||
|
new_cat = match_mail_to_rules(mail_repr, active_rules)
|
||||||
|
if new_cat != c.get("categorie"):
|
||||||
|
save_classement_regle(
|
||||||
|
message_id=c["message_id"],
|
||||||
|
categorie=new_cat,
|
||||||
|
expediteur=c.get("expediteur", ""),
|
||||||
|
objet=c.get("objet", ""),
|
||||||
|
date_mail=c.get("date_mail", ""),
|
||||||
|
etiquettes=raw_tags if isinstance(raw_tags, str) else json.dumps(raw_tags),
|
||||||
|
db_path=db_path
|
||||||
|
)
|
||||||
|
reclassified_count += 1
|
||||||
|
|
||||||
|
return reclassified_count
|
||||||
|
|
||||||
|
def reclassify_today_mails(mirror_client: Optional[Any] = None, db_path: Optional[str] = None, target_date: Optional[str] = None) -> int:
|
||||||
|
"""
|
||||||
|
Reclasse les mails du jour : interroge le miroir pour synchroniser les étiquettes et applique
|
||||||
|
les règles actives, sans jamais écraser un classement manuel.
|
||||||
|
"""
|
||||||
|
t_date = target_date or get_tunis_now_iso()[:10]
|
||||||
|
total_processed = 0
|
||||||
|
|
||||||
|
if mirror_client:
|
||||||
|
try:
|
||||||
|
mails = mirror_client.get_mails_for_day(t_date, max_items=50)
|
||||||
|
for m in mails:
|
||||||
|
classify_mail(m, db_path=db_path)
|
||||||
|
total_processed += 1
|
||||||
|
except Exception:
|
||||||
|
pass
|
||||||
|
|
||||||
|
# Reclassement de sécurité sur les données déjà en base
|
||||||
|
reclassify_all_non_manual(db_path=db_path, date_prefix=t_date)
|
||||||
|
return total_processed
|
||||||
|
|
||||||
def confirm_and_apply_rule(regle_id: int, db_path: Optional[str] = None) -> Optional[Dict[str, Any]]:
|
def confirm_and_apply_rule(regle_id: int, db_path: Optional[str] = None) -> Optional[Dict[str, Any]]:
|
||||||
"""
|
"""
|
||||||
Confirme une règle proposée (statut='active') et applique les règles actives
|
Confirme une règle proposée (statut='active') et applique les règles actives
|
||||||
@@ -138,29 +208,5 @@ def confirm_and_apply_rule(regle_id: int, db_path: Optional[str] = None) -> Opti
|
|||||||
if not rule:
|
if not rule:
|
||||||
return None
|
return None
|
||||||
|
|
||||||
# Réévaluation des emails non manuels en base
|
reclassify_all_non_manual(db_path=db_path)
|
||||||
active_rules = list_regles(statut="active", db_path=db_path)
|
|
||||||
all_classements = list_classements(db_path=db_path)
|
|
||||||
|
|
||||||
for c in all_classements:
|
|
||||||
if c.get("source") == "manuel":
|
|
||||||
continue
|
|
||||||
mail_repr = {
|
|
||||||
"message_id": c["message_id"],
|
|
||||||
"expediteur": c.get("expediteur", ""),
|
|
||||||
"objet": c.get("objet", ""),
|
|
||||||
"apercu": "",
|
|
||||||
"etiquettes": []
|
|
||||||
}
|
|
||||||
new_cat = match_mail_to_rules(mail_repr, active_rules)
|
|
||||||
if new_cat != c.get("categorie"):
|
|
||||||
save_classement_regle(
|
|
||||||
message_id=c["message_id"],
|
|
||||||
categorie=new_cat,
|
|
||||||
expediteur=c.get("expediteur", ""),
|
|
||||||
objet=c.get("objet", ""),
|
|
||||||
date_mail=c.get("date_mail", ""),
|
|
||||||
db_path=db_path
|
|
||||||
)
|
|
||||||
|
|
||||||
return rule
|
return rule
|
||||||
@@ -2,6 +2,8 @@ import os
|
|||||||
from datetime import datetime
|
from datetime import datetime
|
||||||
from zoneinfo import ZoneInfo
|
from zoneinfo import ZoneInfo
|
||||||
|
|
||||||
|
ENV = os.getenv("ENV", "production")
|
||||||
|
|
||||||
DATA_DIR = os.getenv("DATA_DIR", os.path.join(os.path.dirname(os.path.dirname(__file__)), "data"))
|
DATA_DIR = os.getenv("DATA_DIR", os.path.join(os.path.dirname(os.path.dirname(__file__)), "data"))
|
||||||
DB_PATH = os.getenv("DB_PATH", os.path.join(DATA_DIR, "courrier.db"))
|
DB_PATH = os.getenv("DB_PATH", os.path.join(DATA_DIR, "courrier.db"))
|
||||||
MIRROR_TOKEN_PATH = os.getenv("MIRROR_TOKEN_PATH", os.path.join(os.path.dirname(os.path.dirname(__file__)), "secrets", "mirror.token"))
|
MIRROR_TOKEN_PATH = os.getenv("MIRROR_TOKEN_PATH", os.path.join(os.path.dirname(os.path.dirname(__file__)), "secrets", "mirror.token"))
|
||||||
@@ -20,6 +22,27 @@ DEFAULT_CATEGORIES = [
|
|||||||
"À classer"
|
"À classer"
|
||||||
]
|
]
|
||||||
|
|
||||||
|
DEFAULT_RULES = [
|
||||||
|
# 1. Étiquettes du miroir O365
|
||||||
|
{"categorie": "Sans objet", "etiquette_miroir": "sans_objet", "expediteur": None, "mot_cle": None, "statut": "active"},
|
||||||
|
{"categorie": "Tunis", "etiquette_miroir": "tunis", "expediteur": None, "mot_cle": None, "statut": "active"},
|
||||||
|
{"categorie": "Tunis", "etiquette_miroir": "courses", "expediteur": None, "mot_cle": None, "statut": "active"},
|
||||||
|
# 2. Catégories spécifiques et mots-clés évidents
|
||||||
|
{"categorie": "Sans objet", "etiquette_miroir": None, "expediteur": None, "mot_cle": "sans objet", "statut": "active"},
|
||||||
|
{"categorie": "Tunis", "etiquette_miroir": None, "expediteur": None, "mot_cle": "tunis", "statut": "active"},
|
||||||
|
{"categorie": "Tunis", "etiquette_miroir": None, "expediteur": None, "mot_cle": "courses", "statut": "active"},
|
||||||
|
{"categorie": "Demande FB", "etiquette_miroir": "demande fb", "expediteur": None, "mot_cle": None, "statut": "active"},
|
||||||
|
{"categorie": "Demande FB", "etiquette_miroir": "demande_fb", "expediteur": None, "mot_cle": None, "statut": "active"},
|
||||||
|
{"categorie": "Demande FB", "etiquette_miroir": None, "expediteur": None, "mot_cle": "demande fb", "statut": "active"},
|
||||||
|
{"categorie": "Demande FB", "etiquette_miroir": None, "expediteur": None, "mot_cle": "facebook", "statut": "active"},
|
||||||
|
{"categorie": "CI", "etiquette_miroir": "ci", "expediteur": None, "mot_cle": None, "statut": "active"},
|
||||||
|
{"categorie": "CI", "etiquette_miroir": None, "expediteur": None, "mot_cle": "ci", "statut": "active"},
|
||||||
|
{"categorie": "Câbles", "etiquette_miroir": "câbles", "expediteur": None, "mot_cle": None, "statut": "active"},
|
||||||
|
{"categorie": "Câbles", "etiquette_miroir": "cables", "expediteur": None, "mot_cle": None, "statut": "active"},
|
||||||
|
{"categorie": "Câbles", "etiquette_miroir": None, "expediteur": None, "mot_cle": "câbles", "statut": "active"},
|
||||||
|
{"categorie": "Câbles", "etiquette_miroir": None, "expediteur": None, "mot_cle": "cables", "statut": "active"},
|
||||||
|
]
|
||||||
|
|
||||||
def get_tunis_now_iso() -> str:
|
def get_tunis_now_iso() -> str:
|
||||||
"""Retourne l'horodatage actuel en fuseau Africa/Tunis avec offset +01:00."""
|
"""Retourne l'horodatage actuel en fuseau Africa/Tunis avec offset +01:00."""
|
||||||
return datetime.now(ZoneInfo("Africa/Tunis")).strftime("%Y-%m-%dT%H:%M:%S+01:00")
|
return datetime.now(ZoneInfo("Africa/Tunis")).strftime("%Y-%m-%dT%H:%M:%S+01:00")
|
||||||
|
|||||||
@@ -1,7 +1,8 @@
|
|||||||
import os
|
import os
|
||||||
|
import json
|
||||||
import sqlite3
|
import sqlite3
|
||||||
from typing import Optional, List, Dict, Any
|
from typing import Optional, List, Dict, Any
|
||||||
from app.config import DB_PATH, DEFAULT_CATEGORIES, get_tunis_now_iso
|
from app.config import DB_PATH, DEFAULT_CATEGORIES, DEFAULT_RULES, get_tunis_now_iso
|
||||||
|
|
||||||
def get_connection(db_path: Optional[str] = None) -> sqlite3.Connection:
|
def get_connection(db_path: Optional[str] = None) -> sqlite3.Connection:
|
||||||
path = db_path or DB_PATH
|
path = db_path or DB_PATH
|
||||||
@@ -45,18 +46,50 @@ def init_db(db_path: Optional[str] = None) -> None:
|
|||||||
traite INTEGER NOT NULL DEFAULT 0,
|
traite INTEGER NOT NULL DEFAULT 0,
|
||||||
expediteur TEXT,
|
expediteur TEXT,
|
||||||
objet TEXT,
|
objet TEXT,
|
||||||
date_mail TEXT
|
date_mail TEXT,
|
||||||
|
etiquettes TEXT DEFAULT '[]'
|
||||||
);
|
);
|
||||||
""")
|
""")
|
||||||
|
|
||||||
# Initialisation des catégories par défaut si vide
|
# Migration douce si la table classements existait sans la colonne etiquettes
|
||||||
cursor = conn.execute("SELECT COUNT(*) as cnt FROM categories")
|
cols = [r["name"] for r in conn.execute("PRAGMA table_info(classements)").fetchall()]
|
||||||
if cursor.fetchone()["cnt"] == 0:
|
if "etiquettes" not in cols:
|
||||||
now = get_tunis_now_iso()
|
conn.execute("ALTER TABLE classements ADD COLUMN etiquettes TEXT DEFAULT '[]'")
|
||||||
for cat in DEFAULT_CATEGORIES:
|
|
||||||
|
now = get_tunis_now_iso()
|
||||||
|
|
||||||
|
# Initialisation idempotente des catégories par défaut
|
||||||
|
for cat in DEFAULT_CATEGORIES:
|
||||||
|
conn.execute(
|
||||||
|
"INSERT OR IGNORE INTO categories (name, created_at) VALUES (?, ?)",
|
||||||
|
(cat, now)
|
||||||
|
)
|
||||||
|
|
||||||
|
# Initialisation idempotente des règles de départ
|
||||||
|
for rule in DEFAULT_RULES:
|
||||||
|
cat = rule["categorie"]
|
||||||
|
eti = rule.get("etiquette_miroir")
|
||||||
|
exp = rule.get("expediteur")
|
||||||
|
mot = rule.get("mot_cle")
|
||||||
|
sta = rule.get("statut", "active")
|
||||||
|
|
||||||
|
cursor = conn.execute(
|
||||||
|
"""
|
||||||
|
SELECT COUNT(*) as cnt FROM regles
|
||||||
|
WHERE categorie = ?
|
||||||
|
AND (etiquette_miroir IS ? OR etiquette_miroir = ?)
|
||||||
|
AND (expediteur IS ? OR expediteur = ?)
|
||||||
|
AND (mot_cle IS ? OR mot_cle = ?)
|
||||||
|
""",
|
||||||
|
(cat, eti, eti, exp, exp, mot, mot)
|
||||||
|
)
|
||||||
|
if cursor.fetchone()["cnt"] == 0:
|
||||||
conn.execute(
|
conn.execute(
|
||||||
"INSERT OR IGNORE INTO categories (name, created_at) VALUES (?, ?)",
|
"""
|
||||||
(cat, now)
|
INSERT INTO regles (categorie, etiquette_miroir, expediteur, mot_cle, statut, created_at)
|
||||||
|
VALUES (?, ?, ?, ?, ?, ?)
|
||||||
|
""",
|
||||||
|
(cat, eti, exp, mot, sta, now)
|
||||||
)
|
)
|
||||||
finally:
|
finally:
|
||||||
conn.close()
|
conn.close()
|
||||||
@@ -223,7 +256,7 @@ def get_classement(message_id: str, db_path: Optional[str] = None) -> Optional[D
|
|||||||
conn = get_connection(db_path)
|
conn = get_connection(db_path)
|
||||||
try:
|
try:
|
||||||
row = conn.execute(
|
row = conn.execute(
|
||||||
"SELECT message_id, categorie, source, horodatage, traite, expediteur, objet, date_mail FROM classements WHERE message_id = ?",
|
"SELECT message_id, categorie, source, horodatage, traite, expediteur, objet, date_mail, etiquettes FROM classements WHERE message_id = ?",
|
||||||
(message_id,)
|
(message_id,)
|
||||||
).fetchone()
|
).fetchone()
|
||||||
if not row:
|
if not row:
|
||||||
@@ -242,7 +275,7 @@ def list_classements(
|
|||||||
) -> List[Dict[str, Any]]:
|
) -> List[Dict[str, Any]]:
|
||||||
conn = get_connection(db_path)
|
conn = get_connection(db_path)
|
||||||
try:
|
try:
|
||||||
query = "SELECT message_id, categorie, source, horodatage, traite, expediteur, objet, date_mail FROM classements WHERE 1=1"
|
query = "SELECT message_id, categorie, source, horodatage, traite, expediteur, objet, date_mail, etiquettes FROM classements WHERE 1=1"
|
||||||
params: list = []
|
params: list = []
|
||||||
if date_prefix:
|
if date_prefix:
|
||||||
query += " AND (date_mail LIKE ? OR horodatage LIKE ?)"
|
query += " AND (date_mail LIKE ? OR horodatage LIKE ?)"
|
||||||
@@ -270,6 +303,7 @@ def save_classement_regle(
|
|||||||
expediteur: str = "",
|
expediteur: str = "",
|
||||||
objet: str = "",
|
objet: str = "",
|
||||||
date_mail: str = "",
|
date_mail: str = "",
|
||||||
|
etiquettes: Optional[str] = None,
|
||||||
db_path: Optional[str] = None
|
db_path: Optional[str] = None
|
||||||
) -> Dict[str, Any]:
|
) -> Dict[str, Any]:
|
||||||
"""
|
"""
|
||||||
@@ -288,22 +322,24 @@ def save_classement_regle(
|
|||||||
exp = expediteur or (existing["expediteur"] if existing else "")
|
exp = expediteur or (existing["expediteur"] if existing else "")
|
||||||
obj = objet or (existing["objet"] if existing else "")
|
obj = objet or (existing["objet"] if existing else "")
|
||||||
d_mail = date_mail or (existing["date_mail"] if existing else "")
|
d_mail = date_mail or (existing["date_mail"] if existing else "")
|
||||||
|
etiq = etiquettes or (existing.get("etiquettes") if existing else "[]")
|
||||||
|
|
||||||
with conn:
|
with conn:
|
||||||
conn.execute(
|
conn.execute(
|
||||||
"""
|
"""
|
||||||
INSERT INTO classements (message_id, categorie, source, horodatage, traite, expediteur, objet, date_mail)
|
INSERT INTO classements (message_id, categorie, source, horodatage, traite, expediteur, objet, date_mail, etiquettes)
|
||||||
VALUES (?, ?, 'regle', ?, ?, ?, ?, ?)
|
VALUES (?, ?, 'regle', ?, ?, ?, ?, ?, ?)
|
||||||
ON CONFLICT(message_id) DO UPDATE SET
|
ON CONFLICT(message_id) DO UPDATE SET
|
||||||
categorie = excluded.categorie,
|
categorie = excluded.categorie,
|
||||||
source = 'regle',
|
source = 'regle',
|
||||||
horodatage = excluded.horodatage,
|
horodatage = excluded.horodatage,
|
||||||
expediteur = excluded.expediteur,
|
expediteur = excluded.expediteur,
|
||||||
objet = excluded.objet,
|
objet = excluded.objet,
|
||||||
date_mail = excluded.date_mail
|
date_mail = excluded.date_mail,
|
||||||
|
etiquettes = CASE WHEN excluded.etiquettes != '[]' THEN excluded.etiquettes ELSE classements.etiquettes END
|
||||||
WHERE classements.source != 'manuel'
|
WHERE classements.source != 'manuel'
|
||||||
""",
|
""",
|
||||||
(message_id, categorie, now, 1 if traite else 0, exp, obj, d_mail)
|
(message_id, categorie, now, 1 if traite else 0, exp, obj, d_mail, etiq)
|
||||||
)
|
)
|
||||||
return get_classement(message_id, db_path)
|
return get_classement(message_id, db_path)
|
||||||
finally:
|
finally:
|
||||||
@@ -316,6 +352,7 @@ def save_classement_manuel(
|
|||||||
expediteur: Optional[str] = None,
|
expediteur: Optional[str] = None,
|
||||||
objet: Optional[str] = None,
|
objet: Optional[str] = None,
|
||||||
date_mail: Optional[str] = None,
|
date_mail: Optional[str] = None,
|
||||||
|
etiquettes: Optional[str] = None,
|
||||||
db_path: Optional[str] = None
|
db_path: Optional[str] = None
|
||||||
) -> Dict[str, Any]:
|
) -> Dict[str, Any]:
|
||||||
"""
|
"""
|
||||||
@@ -329,12 +366,13 @@ def save_classement_manuel(
|
|||||||
exp = expediteur if expediteur is not None else (existing["expediteur"] if existing else "")
|
exp = expediteur if expediteur is not None else (existing["expediteur"] if existing else "")
|
||||||
obj = objet if objet is not None else (existing["objet"] if existing else "")
|
obj = objet if objet is not None else (existing["objet"] if existing else "")
|
||||||
d_mail = date_mail if date_mail is not None else (existing["date_mail"] if existing else "")
|
d_mail = date_mail if date_mail is not None else (existing["date_mail"] if existing else "")
|
||||||
|
etiq = etiquettes if etiquettes is not None else (existing.get("etiquettes") if existing else "[]")
|
||||||
|
|
||||||
with conn:
|
with conn:
|
||||||
conn.execute(
|
conn.execute(
|
||||||
"""
|
"""
|
||||||
INSERT INTO classements (message_id, categorie, source, horodatage, traite, expediteur, objet, date_mail)
|
INSERT INTO classements (message_id, categorie, source, horodatage, traite, expediteur, objet, date_mail, etiquettes)
|
||||||
VALUES (?, ?, 'manuel', ?, ?, ?, ?, ?)
|
VALUES (?, ?, 'manuel', ?, ?, ?, ?, ?, ?)
|
||||||
ON CONFLICT(message_id) DO UPDATE SET
|
ON CONFLICT(message_id) DO UPDATE SET
|
||||||
categorie = excluded.categorie,
|
categorie = excluded.categorie,
|
||||||
source = 'manuel',
|
source = 'manuel',
|
||||||
@@ -342,9 +380,10 @@ def save_classement_manuel(
|
|||||||
traite = excluded.traite,
|
traite = excluded.traite,
|
||||||
expediteur = CASE WHEN excluded.expediteur != '' THEN excluded.expediteur ELSE classements.expediteur END,
|
expediteur = CASE WHEN excluded.expediteur != '' THEN excluded.expediteur ELSE classements.expediteur END,
|
||||||
objet = CASE WHEN excluded.objet != '' THEN excluded.objet ELSE classements.objet END,
|
objet = CASE WHEN excluded.objet != '' THEN excluded.objet ELSE classements.objet END,
|
||||||
date_mail = CASE WHEN excluded.date_mail != '' THEN excluded.date_mail ELSE classements.date_mail END
|
date_mail = CASE WHEN excluded.date_mail != '' THEN excluded.date_mail ELSE classements.date_mail END,
|
||||||
|
etiquettes = CASE WHEN excluded.etiquettes != '[]' THEN excluded.etiquettes ELSE classements.etiquettes END
|
||||||
""",
|
""",
|
||||||
(message_id, categorie, now, is_traite, exp, obj, d_mail)
|
(message_id, categorie, now, is_traite, exp, obj, d_mail, etiq)
|
||||||
)
|
)
|
||||||
return get_classement(message_id, db_path)
|
return get_classement(message_id, db_path)
|
||||||
finally:
|
finally:
|
||||||
|
|||||||
+8
-2
@@ -22,7 +22,8 @@ from app.db import (
|
|||||||
from app.classifier import (
|
from app.classifier import (
|
||||||
classify_mail,
|
classify_mail,
|
||||||
manual_classify_and_propose,
|
manual_classify_and_propose,
|
||||||
confirm_and_apply_rule
|
confirm_and_apply_rule,
|
||||||
|
reclassify_today_mails
|
||||||
)
|
)
|
||||||
from app.mirror_client import MirrorClient
|
from app.mirror_client import MirrorClient
|
||||||
from app.auth import verify_cf_access_jwt
|
from app.auth import verify_cf_access_jwt
|
||||||
@@ -59,8 +60,13 @@ class TraiteRequest(BaseModel):
|
|||||||
# --- Lifespan ---
|
# --- Lifespan ---
|
||||||
@asynccontextmanager
|
@asynccontextmanager
|
||||||
async def lifespan(app: FastAPI):
|
async def lifespan(app: FastAPI):
|
||||||
# Démarrage : initialisation de la base SQLite
|
# Démarrage : initialisation de la base SQLite et des règles par défaut
|
||||||
init_db()
|
init_db()
|
||||||
|
# Reclassement des mails du jour après ajout des règles de départ
|
||||||
|
try:
|
||||||
|
reclassify_today_mails(mirror_client=mirror_client)
|
||||||
|
except Exception:
|
||||||
|
pass
|
||||||
yield
|
yield
|
||||||
|
|
||||||
app = FastAPI(
|
app = FastAPI(
|
||||||
|
|||||||
+8
-1
@@ -4,8 +4,9 @@ import pytest
|
|||||||
from fastapi.testclient import TestClient
|
from fastapi.testclient import TestClient
|
||||||
import jwt
|
import jwt
|
||||||
|
|
||||||
# Configure test environment variables before importing app
|
# Configuration impérative de l'environnement de test avant l'importation de l'application
|
||||||
TEST_SECRET = "secret-test-courrier-tt-2026-11-059"
|
TEST_SECRET = "secret-test-courrier-tt-2026-11-059"
|
||||||
|
os.environ["ENV"] = "test"
|
||||||
os.environ["CF_TEST_SECRET"] = TEST_SECRET
|
os.environ["CF_TEST_SECRET"] = TEST_SECRET
|
||||||
os.environ["CF_ACCESS_AUD"] = "courrier-test-aud"
|
os.environ["CF_ACCESS_AUD"] = "courrier-test-aud"
|
||||||
|
|
||||||
@@ -39,6 +40,12 @@ def client(valid_jwt_token, temp_db, monkeypatch):
|
|||||||
"""Client de test FastAPI avec injecteur d'en-tête Cloudflare Access et DB isolée."""
|
"""Client de test FastAPI avec injecteur d'en-tête Cloudflare Access et DB isolée."""
|
||||||
monkeypatch.setattr("app.config.DB_PATH", temp_db)
|
monkeypatch.setattr("app.config.DB_PATH", temp_db)
|
||||||
monkeypatch.setattr("app.db.DB_PATH", temp_db)
|
monkeypatch.setattr("app.db.DB_PATH", temp_db)
|
||||||
|
monkeypatch.setattr("app.config.ENV", "test")
|
||||||
|
monkeypatch.setattr("app.auth.ENV", "test")
|
||||||
|
monkeypatch.setattr("app.config.CF_TEST_SECRET", TEST_SECRET)
|
||||||
|
monkeypatch.setattr("app.auth.CF_TEST_SECRET", TEST_SECRET)
|
||||||
|
monkeypatch.setattr("app.config.CF_ACCESS_AUD", "courrier-test-aud")
|
||||||
|
monkeypatch.setattr("app.auth.CF_ACCESS_AUD", "courrier-test-aud")
|
||||||
init_db(temp_db)
|
init_db(temp_db)
|
||||||
with TestClient(app) as c:
|
with TestClient(app) as c:
|
||||||
c.headers.update({"Cf-Access-Jwt-Assertion": valid_jwt_token})
|
c.headers.update({"Cf-Access-Jwt-Assertion": valid_jwt_token})
|
||||||
|
|||||||
+127
-4
@@ -1,8 +1,11 @@
|
|||||||
import pytest
|
import pytest
|
||||||
|
import jwt
|
||||||
from fastapi.testclient import TestClient
|
from fastapi.testclient import TestClient
|
||||||
from app.main import app
|
from app.main import app
|
||||||
|
from tests.conftest import TEST_SECRET
|
||||||
|
|
||||||
def test_health_endpoint_no_auth_required():
|
def test_health_endpoint_no_auth_required():
|
||||||
|
"""/health reste public et répond 200 avec l'horodatage de Tunis."""
|
||||||
client = TestClient(app)
|
client = TestClient(app)
|
||||||
resp = client.get("/health")
|
resp = client.get("/health")
|
||||||
assert resp.status_code == 200
|
assert resp.status_code == 200
|
||||||
@@ -11,22 +14,142 @@ def test_health_endpoint_no_auth_required():
|
|||||||
assert data["service"] == "courrier"
|
assert data["service"] == "courrier"
|
||||||
assert "+01:00" in data["horodatage"]
|
assert "+01:00" in data["horodatage"]
|
||||||
|
|
||||||
def test_missing_cf_access_jwt_returns_401():
|
def test_missing_cf_access_jwt_returns_401(monkeypatch):
|
||||||
|
"""Une requête API sans en-tête JWT est rejetée en 401."""
|
||||||
|
monkeypatch.setattr("app.auth.ENV", "test")
|
||||||
|
monkeypatch.setattr("app.auth.CF_TEST_SECRET", TEST_SECRET)
|
||||||
|
monkeypatch.setattr("app.auth.CF_ACCESS_AUD", "courrier-test-aud")
|
||||||
client = TestClient(app)
|
client = TestClient(app)
|
||||||
resp = client.get("/api/v1/categories")
|
resp = client.get("/api/v1/categories")
|
||||||
assert resp.status_code == 401
|
assert resp.status_code == 401
|
||||||
assert "Cf-Access-Jwt-Assertion manquant" in resp.json()["detail"]
|
assert "manquant" in resp.json()["detail"]
|
||||||
|
|
||||||
def test_invalid_cf_access_jwt_returns_401():
|
def test_invalid_cf_access_jwt_returns_401(monkeypatch):
|
||||||
|
"""Un jeton malformé est rejeté en 401."""
|
||||||
|
monkeypatch.setattr("app.auth.ENV", "test")
|
||||||
|
monkeypatch.setattr("app.auth.CF_TEST_SECRET", TEST_SECRET)
|
||||||
|
monkeypatch.setattr("app.auth.CF_ACCESS_AUD", "courrier-test-aud")
|
||||||
client = TestClient(app)
|
client = TestClient(app)
|
||||||
# Token malformé
|
|
||||||
resp = client.get(
|
resp = client.get(
|
||||||
"/api/v1/categories",
|
"/api/v1/categories",
|
||||||
headers={"Cf-Access-Jwt-Assertion": "ceci.n-est-pas.un-jwt-valide"}
|
headers={"Cf-Access-Jwt-Assertion": "ceci.n-est-pas.un-jwt-valide"}
|
||||||
)
|
)
|
||||||
assert resp.status_code == 401
|
assert resp.status_code == 401
|
||||||
|
|
||||||
|
def test_forged_alg_none_jwt_returns_401(monkeypatch):
|
||||||
|
"""
|
||||||
|
Preuve T1 : Un jeton forgé avec l'algorithme 'none' est formellement
|
||||||
|
rejeté en 401 (aucun décodage non vérifié permis).
|
||||||
|
"""
|
||||||
|
monkeypatch.setattr("app.auth.ENV", "test")
|
||||||
|
monkeypatch.setattr("app.auth.CF_TEST_SECRET", TEST_SECRET)
|
||||||
|
monkeypatch.setattr("app.auth.CF_ACCESS_AUD", "courrier-test-aud")
|
||||||
|
|
||||||
|
payload = {
|
||||||
|
"sub": "hacker@evil.com",
|
||||||
|
"email": "hacker@evil.com",
|
||||||
|
"aud": "courrier-test-aud",
|
||||||
|
"exp": 9999999999
|
||||||
|
}
|
||||||
|
# Jeton non signé (alg: none)
|
||||||
|
forged_token = jwt.encode(payload, key="", algorithm="none")
|
||||||
|
|
||||||
|
client = TestClient(app)
|
||||||
|
resp = client.get(
|
||||||
|
"/api/v1/categories",
|
||||||
|
headers={"Cf-Access-Jwt-Assertion": forged_token}
|
||||||
|
)
|
||||||
|
assert resp.status_code == 401
|
||||||
|
assert "signature non vérifiée" in resp.json()["detail"].lower() or "invalide" in resp.json()["detail"].lower()
|
||||||
|
|
||||||
|
def test_forged_bad_signature_jwt_returns_401(monkeypatch):
|
||||||
|
"""
|
||||||
|
Preuve T1 : Un jeton signé avec une fausse clé secrète est formellement
|
||||||
|
rejeté en 401.
|
||||||
|
"""
|
||||||
|
monkeypatch.setattr("app.auth.ENV", "test")
|
||||||
|
monkeypatch.setattr("app.auth.CF_TEST_SECRET", TEST_SECRET)
|
||||||
|
monkeypatch.setattr("app.auth.CF_ACCESS_AUD", "courrier-test-aud")
|
||||||
|
|
||||||
|
payload = {
|
||||||
|
"sub": "hacker@evil.com",
|
||||||
|
"email": "hacker@evil.com",
|
||||||
|
"aud": "courrier-test-aud",
|
||||||
|
"exp": 9999999999
|
||||||
|
}
|
||||||
|
# Signé avec une clé pirate différente
|
||||||
|
forged_token = jwt.encode(payload, "mauvaise-cle-pirate-123-longueur-securisee-32-bytes", algorithm="HS256")
|
||||||
|
|
||||||
|
client = TestClient(app)
|
||||||
|
resp = client.get(
|
||||||
|
"/api/v1/categories",
|
||||||
|
headers={"Cf-Access-Jwt-Assertion": forged_token}
|
||||||
|
)
|
||||||
|
assert resp.status_code == 401
|
||||||
|
|
||||||
|
def test_bad_audience_jwt_returns_401(monkeypatch):
|
||||||
|
"""Un jeton portant un audience différent de CF_ACCESS_AUD est rejeté en 401."""
|
||||||
|
monkeypatch.setattr("app.auth.ENV", "test")
|
||||||
|
monkeypatch.setattr("app.auth.CF_TEST_SECRET", TEST_SECRET)
|
||||||
|
monkeypatch.setattr("app.auth.CF_ACCESS_AUD", "courrier-test-aud")
|
||||||
|
|
||||||
|
payload = {
|
||||||
|
"sub": "user@example.com",
|
||||||
|
"aud": "autre-audience-invalide",
|
||||||
|
"exp": 9999999999
|
||||||
|
}
|
||||||
|
token = jwt.encode(payload, TEST_SECRET, algorithm="HS256")
|
||||||
|
|
||||||
|
client = TestClient(app)
|
||||||
|
resp = client.get(
|
||||||
|
"/api/v1/categories",
|
||||||
|
headers={"Cf-Access-Jwt-Assertion": token}
|
||||||
|
)
|
||||||
|
assert resp.status_code == 401
|
||||||
|
|
||||||
def test_valid_cf_access_jwt_returns_200(client):
|
def test_valid_cf_access_jwt_returns_200(client):
|
||||||
|
"""Un jeton correctement signé en environnement de test donne accès en 200."""
|
||||||
resp = client.get("/api/v1/categories")
|
resp = client.get("/api/v1/categories")
|
||||||
assert resp.status_code == 200
|
assert resp.status_code == 200
|
||||||
assert "categories" in resp.json()
|
assert "categories" in resp.json()
|
||||||
|
|
||||||
|
def test_unconfigured_auth_returns_503_on_api_and_200_on_health(monkeypatch):
|
||||||
|
"""
|
||||||
|
Preuve T1 : En production (ENV!=test) sans CF_ACCESS_CERTS_URL et CF_ACCESS_AUD,
|
||||||
|
toutes les routes /api répondent 503 tandis que /health reste 200.
|
||||||
|
"""
|
||||||
|
monkeypatch.setattr("app.auth.ENV", "production")
|
||||||
|
monkeypatch.setattr("app.auth.CF_ACCESS_CERTS_URL", "")
|
||||||
|
monkeypatch.setattr("app.auth.CF_ACCESS_AUD", "")
|
||||||
|
|
||||||
|
client = TestClient(app)
|
||||||
|
|
||||||
|
# 1. /health reste fonctionnel en 200
|
||||||
|
r_health = client.get("/health")
|
||||||
|
assert r_health.status_code == 200
|
||||||
|
|
||||||
|
# 2. Les routes /api renvoient 503 Fail-Closed
|
||||||
|
r_api = client.get(
|
||||||
|
"/api/v1/categories",
|
||||||
|
headers={"Cf-Access-Jwt-Assertion": "dummy.token"}
|
||||||
|
)
|
||||||
|
assert r_api.status_code == 503
|
||||||
|
assert "obligatoires" in r_api.json()["detail"]
|
||||||
|
|
||||||
|
def test_cf_test_secret_ignored_when_not_in_test_env(monkeypatch):
|
||||||
|
"""
|
||||||
|
Preuve T1 : Le mode CF_TEST_SECRET n'existe que si ENV=test.
|
||||||
|
Si ENV=production, CF_TEST_SECRET est strictement ignoré et le service
|
||||||
|
répond 503 si les clés JWKS de production ne sont pas configurées.
|
||||||
|
"""
|
||||||
|
monkeypatch.setattr("app.auth.ENV", "production")
|
||||||
|
monkeypatch.setattr("app.auth.CF_TEST_SECRET", TEST_SECRET)
|
||||||
|
monkeypatch.setattr("app.auth.CF_ACCESS_CERTS_URL", "")
|
||||||
|
monkeypatch.setattr("app.auth.CF_ACCESS_AUD", "")
|
||||||
|
|
||||||
|
client = TestClient(app)
|
||||||
|
r = client.get(
|
||||||
|
"/api/v1/categories",
|
||||||
|
headers={"Cf-Access-Jwt-Assertion": "token"}
|
||||||
|
)
|
||||||
|
assert r.status_code == 503
|
||||||
@@ -11,6 +11,7 @@ from app.classifier import (
|
|||||||
classify_mail,
|
classify_mail,
|
||||||
manual_classify_and_propose,
|
manual_classify_and_propose,
|
||||||
confirm_and_apply_rule,
|
confirm_and_apply_rule,
|
||||||
|
reclassify_all_non_manual,
|
||||||
extract_keywords
|
extract_keywords
|
||||||
)
|
)
|
||||||
|
|
||||||
@@ -132,3 +133,46 @@ def test_confirm_proposed_rule_applies_to_non_manual_mails_only(temp_db):
|
|||||||
c1_after = get_classement(mid1, db_path=temp_db)
|
c1_after = get_classement(mid1, db_path=temp_db)
|
||||||
assert c1_after["categorie"] == "Tunis"
|
assert c1_after["categorie"] == "Tunis"
|
||||||
assert c1_after["source"] == "manuel"
|
assert c1_after["source"] == "manuel"
|
||||||
|
|
||||||
|
def test_reclassify_preserves_manual_and_uses_non_default_category(temp_db):
|
||||||
|
"""
|
||||||
|
Preuve T2 : Au moins une catégorie autre que 'À classer' est utilisée sur les mails du jour,
|
||||||
|
et les classements manuels ne sont JAMAIS écrasés.
|
||||||
|
"""
|
||||||
|
# Mail A avec étiquette 'tunis'
|
||||||
|
mail_a = {
|
||||||
|
"message_id": "<mail-tunis-01@domain.tn>",
|
||||||
|
"expediteur": "contact@tunis.tn",
|
||||||
|
"objet": "Point d'avancement",
|
||||||
|
"date": "2026-11-17",
|
||||||
|
"etiquettes": ["tunis"]
|
||||||
|
}
|
||||||
|
cA = classify_mail(mail_a, db_path=temp_db)
|
||||||
|
assert cA["categorie"] == "Tunis", "L'étiquette 'tunis' doit classer en 'Tunis'"
|
||||||
|
assert cA["source"] == "regle"
|
||||||
|
|
||||||
|
# Mail B classé manuellement en 'CI'
|
||||||
|
mail_b = {
|
||||||
|
"message_id": "<mail-manual-01@domain.tn>",
|
||||||
|
"expediteur": "contact@autre.tn",
|
||||||
|
"objet": "Point sans objet",
|
||||||
|
"date": "2026-11-17",
|
||||||
|
"etiquettes": ["sans_objet"]
|
||||||
|
}
|
||||||
|
cB_manual = save_classement_manuel(
|
||||||
|
message_id=mail_b["message_id"],
|
||||||
|
categorie="CI",
|
||||||
|
db_path=temp_db
|
||||||
|
)
|
||||||
|
assert cB_manual["categorie"] == "CI"
|
||||||
|
assert cB_manual["source"] == "manuel"
|
||||||
|
|
||||||
|
# Reclassement général
|
||||||
|
reclassify_all_non_manual(db_path=temp_db)
|
||||||
|
|
||||||
|
# Vérifications :
|
||||||
|
# 1. Le mail A est resté 'Tunis'
|
||||||
|
assert get_classement(mail_a["message_id"], db_path=temp_db)["categorie"] == "Tunis"
|
||||||
|
# 2. Le mail B manuel n'a JAMAIS été écrasé vers 'Sans objet'
|
||||||
|
assert get_classement(mail_b["message_id"], db_path=temp_db)["categorie"] == "CI"
|
||||||
|
assert get_classement(mail_b["message_id"], db_path=temp_db)["source"] == "manuel"
|
||||||
+33
-3
@@ -1,5 +1,6 @@
|
|||||||
import pytest
|
import pytest
|
||||||
from app.db import (
|
from app.db import (
|
||||||
|
init_db,
|
||||||
list_categories,
|
list_categories,
|
||||||
add_category,
|
add_category,
|
||||||
update_category,
|
update_category,
|
||||||
@@ -20,6 +21,33 @@ def test_init_db_creates_tables_and_default_categories(temp_db):
|
|||||||
expected = ["Tunis", "Demande FB", "CI", "Câbles", "Sans objet", "À classer"]
|
expected = ["Tunis", "Demande FB", "CI", "Câbles", "Sans objet", "À classer"]
|
||||||
assert cat_names == expected, f"Catégories par défaut attendues: {expected}, obtenu: {cat_names}"
|
assert cat_names == expected, f"Catégories par défaut attendues: {expected}, obtenu: {cat_names}"
|
||||||
|
|
||||||
|
def test_init_db_creates_default_rules_idempotently(temp_db):
|
||||||
|
"""
|
||||||
|
Preuve T2 : Les règles de départ sont créées à partir des étiquettes miroir
|
||||||
|
(sans_objet, tunis, courses) et catégories (Demande FB, CI, Câbles) de façon idempotente.
|
||||||
|
"""
|
||||||
|
rules_initial = list_regles(statut="active", db_path=temp_db)
|
||||||
|
assert len(rules_initial) >= 6
|
||||||
|
|
||||||
|
# Vérification des règles sur étiquettes miroir
|
||||||
|
tags = {r["etiquette_miroir"] for r in rules_initial if r["etiquette_miroir"]}
|
||||||
|
assert "sans_objet" in tags
|
||||||
|
assert "tunis" in tags
|
||||||
|
assert "courses" in tags
|
||||||
|
|
||||||
|
# Vérification des catégories cibles
|
||||||
|
rule_cats = {r["categorie"] for r in rules_initial}
|
||||||
|
assert "Sans objet" in rule_cats
|
||||||
|
assert "Tunis" in rule_cats
|
||||||
|
assert "Demande FB" in rule_cats
|
||||||
|
assert "CI" in rule_cats
|
||||||
|
assert "Câbles" in rule_cats
|
||||||
|
|
||||||
|
# Idempotence : un second init_db ne doit pas créer de doublons
|
||||||
|
init_db(temp_db)
|
||||||
|
rules_after = list_regles(statut="active", db_path=temp_db)
|
||||||
|
assert len(rules_after) == len(rules_initial)
|
||||||
|
|
||||||
def test_crud_categories(temp_db):
|
def test_crud_categories(temp_db):
|
||||||
# Création
|
# Création
|
||||||
new_cat = add_category("Factures", db_path=temp_db)
|
new_cat = add_category("Factures", db_path=temp_db)
|
||||||
@@ -40,6 +68,8 @@ def test_crud_categories(temp_db):
|
|||||||
assert not any(c["name"] == "Factures Fournisseurs" for c in cats_after)
|
assert not any(c["name"] == "Factures Fournisseurs" for c in cats_after)
|
||||||
|
|
||||||
def test_crud_regles(temp_db):
|
def test_crud_regles(temp_db):
|
||||||
|
initial_count = len(list_regles(db_path=temp_db))
|
||||||
|
|
||||||
# Création
|
# Création
|
||||||
rule = add_regle(
|
rule = add_regle(
|
||||||
categorie="Tunis",
|
categorie="Tunis",
|
||||||
@@ -59,12 +89,12 @@ def test_crud_regles(temp_db):
|
|||||||
|
|
||||||
# Liste
|
# Liste
|
||||||
rules = list_regles(statut="active", db_path=temp_db)
|
rules = list_regles(statut="active", db_path=temp_db)
|
||||||
assert len(rules) == 1
|
assert len(rules) == initial_count + 1
|
||||||
assert rules[0]["categorie"] == "Demande FB"
|
assert any(r["id"] == rule["id"] and r["categorie"] == "Demande FB" for r in rules)
|
||||||
|
|
||||||
# Suppression
|
# Suppression
|
||||||
assert delete_regle(rule["id"], db_path=temp_db) is True
|
assert delete_regle(rule["id"], db_path=temp_db) is True
|
||||||
assert len(list_regles(db_path=temp_db)) == 0
|
assert len(list_regles(db_path=temp_db)) == initial_count
|
||||||
|
|
||||||
def test_manual_classification_persistence(temp_db):
|
def test_manual_classification_persistence(temp_db):
|
||||||
mid = "<msg-001@domain.com>"
|
mid = "<msg-001@domain.com>"
|
||||||
|
|||||||
Reference in new issue
Block a user