Mouigni/regwatch-backend
0
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Restauration : fichiers Python depuis commit 4080a308ea

Browse Source
This commit is contained in:
Mouigni
2026-08-11 23:33:24 +02:00
parent 8df04dd679
commit 1f7d06c26d
14 changed files with 1366 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files
tests/__init__.py
View File
tests/test_regwatch_complet.py
+531
View File
@@ -0,0 +1,531 @@
import pytest
import sys
import os
import time
sys.path.insert(0, os.path.abspath(os.path.join(os.path.dirname(__file__), "..")))
# ── Import app FastAPI ────────────────────────────────────────
try:
from fastapi.testclient import TestClient
from api.main import app
client = TestClient(app)
APP_OK = True
except Exception as e:
APP_OK = False
def skip_no_app():
if not APP_OK:
pytest.skip("App / BDD non disponible")
# ══════════════════════════════════════════════════════════════
# PARTIE 1 — TESTS UNITAIRES (scraper CIR)
# ══════════════════════════════════════════════════════════════
class TestUnitairesExtractPagesText:
"""
Tests unitaires de extract_pages_text()
Réf. cir_scraper.py l.46 — extraction texte PDF
"""
def test_bytes_vide_retourne_chaine_vide(self):
"""Bytes vides → '' sans crash"""
from scrapers.cir_scraper import extract_pages_text
assert extract_pages_text(b"") == ""
def test_pdf_invalide_retourne_chaine_vide(self):
"""Contenu non-PDF → '' (le except l.57 fonctionne)"""
from scrapers.cir_scraper import extract_pages_text
assert extract_pages_text(b"ceci n est pas un pdf") == ""
def test_retour_toujours_une_str(self):
"""La fonction retourne toujours str, jamais None"""
from scrapers.cir_scraper import extract_pages_text
result = extract_pages_text(b"")
assert isinstance(result, str)
def test_limite_5000_caracteres(self):
"""Le texte est tronqué à 5000 chars max (l.55 : [:5000])"""
long_text = "A" * 10000
assert len(long_text[:5000]) == 5000
class TestUnitairesFallbackOllama:
"""
Tests unitaires du fallback Ollama (cir_scraper.py l.65)
Quand le texte extrait est vide, on utilise le nom du fichier PDF
"""
def test_fallback_utilise_nom_fichier(self):
"""pdf_url → nom du fichier sans extension (l.66-68)"""
pdf_url = "https://cir-safety.org/sites/files/Retinol_2024.pdf"
filename = pdf_url.split("/")[-1].replace("_", " ").replace(".pdf", "")
assert filename == "Retinol 2024"
def test_fallback_retourne_4_cles(self):
"""Le fallback retourne bien title, ingredient, document_type, date"""
fallback = {
"title": "Mon Document",
"ingredient": None,
"document_type": "document",
"date": None,
}
assert all(k in fallback for k in ["title", "ingredient", "document_type", "date"])
assert fallback["ingredient"] is None
assert fallback["document_type"] == "document"
def test_nettoyage_ingredient_null_string(self):
"""Ollama retourne 'null'/'none'/'n/a' → converti en None (l.112-113)"""
for val in ["null", "none", "n/a", "", "NULL", "None"]:
result = None if str(val).strip().lower() in ("null", "none", "n/a", "") else val
assert result is None, f"'{val}' aurait dû être None"
def test_titre_tronque_200_chars(self):
"""Le titre est limité à 200 chars (l.117 : [:200])"""
long = "A" * 300
assert len(str(long).strip()[:200]) == 200
class TestUnitairesTemplatesURL:
"""
Tests unitaires des templates d'URL CIR (cir_scraper.py l.31-40)
"""
def test_tous_les_templates_contiennent_num(self):
"""Chaque template contient {num}"""
from scrapers.cir_scraper import url_templates
for t in url_templates:
assert "{num}" in t
def test_tous_les_templates_sont_https(self):
"""Tous les templates utilisent HTTPS"""
from scrapers.cir_scraper import url_templates
for t in url_templates:
assert t.startswith("https://")
def test_format_url_remplace_num(self):
"""Le format {num} est bien remplacé"""
from scrapers.cir_scraper import url_templates
url = url_templates[0].format(num=150)
assert "150" in url
assert "{num}" not in url
def test_meeting_numbers_couvre_115_a_174(self):
"""MEETING_NUMBERS = range(115, 175) → 60 réunions"""
from scrapers.cir_scraper import MEETING_NUMBERS
nums = list(MEETING_NUMBERS)
assert nums[0] == 115
assert nums[-1] == 174
assert len(nums) == 60
def test_pdf_limit_est_100(self):
"""PDF_LIMIT = 100 (l.32)"""
from scrapers.cir_scraper import PDF_LIMIT
assert PDF_LIMIT == 100
class TestUnitairesFiltragePDF:
"""
Tests unitaires du filtrage des liens PDF (cir_scraper.py l.148)
"""
def test_skip_keywords_contient_les_3_valeurs(self):
"""SKIP_KEYWORDS = ['Agenda', 'Minutes', 'Status Report']"""
from scrapers.cir_scraper import SKIP_KEYWORDS
assert "Agenda" in SKIP_KEYWORDS
assert "Minutes" in SKIP_KEYWORDS
assert "Status Report" in SKIP_KEYWORDS
def test_regex_pdf_detecte_extension_pdf(self):
"""PDF_RE détecte .pdf / .PDF / .Pdf"""
from scrapers.cir_scraper import PDF_RE
assert PDF_RE.search("document.pdf")
assert PDF_RE.search("document.PDF")
assert PDF_RE.search("rapport.Pdf")
def test_regex_pdf_ne_detecte_pas_autres(self):
"""PDF_RE ne détecte pas .docx, .png"""
from scrapers.cir_scraper import PDF_RE
assert not PDF_RE.search("document.docx")
assert not PDF_RE.search("image.png")
def test_agenda_est_filtre(self):
"""Un contexte contenant 'Agenda' → filtré"""
from scrapers.cir_scraper import SKIP_KEYWORDS
context = "116th Expert Panel Meeting Agenda"
assert any(kw in context for kw in SKIP_KEYWORDS)
def test_final_report_non_filtre(self):
"""Un 'Final Report' passe le filtre"""
from scrapers.cir_scraper import SKIP_KEYWORDS
context = "Final Report on the Safety Assessment of Retinol"
assert not any(kw in context for kw in SKIP_KEYWORDS)
class TestUnitairesDeduplication:
"""
Tests unitaires de la logique de déduplication (cir_scraper.py l.172)
"""
def test_nouvelle_url_acceptee(self):
"""Une URL absente du set → à insérer"""
existing = {"https://cir.org/doc1.pdf", "https://cir.org/doc2.pdf"}
assert "https://cir.org/doc3.pdf" not in existing
def test_url_existante_rejetee(self):
"""Une URL déjà présente → doublons ignoré"""
existing = {"https://cir.org/doc1.pdf"}
assert "https://cir.org/doc1.pdf" in existing
def test_structure_doc_6_champs(self):
"""Un doc à sauvegarder contient les 6 champs requis"""
doc = {
"title": "Final Report on Retinol",
"ingredient": "Retinol",
"source": "CIR",
"document_type": "final report",
"meeting_date": "2024",
"pdf_url": "https://cir.org/retinol_2024.pdf",
}
for champ in ["title", "ingredient", "source", "document_type", "meeting_date", "pdf_url"]:
assert champ in doc
def test_source_est_cir(self):
"""La source est toujours 'CIR' pour ce scraper"""
assert "CIR" == "CIR"
class TestUnitairesHeadersHTTP:
"""
Tests unitaires des headers HTTP anti-blocage (cir_scraper.py l.22)
"""
def test_user_agent_simule_chrome(self):
"""User-Agent contient Mozilla et Chrome"""
from scrapers.cir_scraper import headers
assert "User-Agent" in headers
assert "Mozilla" in headers["User-Agent"]
assert "Chrome" in headers["User-Agent"]
def test_session_utilise_les_headers(self):
"""La session requests hérite des headers"""
from scrapers.cir_scraper import session
assert "User-Agent" in session.headers
assert "Mozilla" in session.headers["User-Agent"]
# ══════════════════════════════════════════════════════════════
# PARTIE 2 — TESTS D'INTÉGRATION (endpoints FastAPI)
# ══════════════════════════════════════════════════════════════
class TestIntegrationRacine:
"""Tests de GET / (main.py l.21)"""
def test_racine_200(self):
skip_no_app()
assert client.get("/").status_code == 200
def test_racine_retourne_message_running(self):
skip_no_app()
data = client.get("/").json()
assert "message" in data
assert "running" in data["message"].lower() or "RegWatch" in data["message"]
def test_racine_retourne_json(self):
skip_no_app()
r = client.get("/")
assert r.headers["content-type"].startswith("application/json")
class TestIntegrationSignup:
"""Tests de POST /signup (main.py l.28)"""
def test_email_invalide_422(self):
skip_no_app()
r = client.post("/signup", json={"email": "pasunemail", "password": "test", "full_name": "T"})
assert r.status_code == 422
def test_champs_manquants_422(self):
skip_no_app()
r = client.post("/signup", json={"email": "test@test.com"})
assert r.status_code == 422
def test_email_deja_existant_400(self):
"""Email déjà pris → 400 'Email already registered' (l.33-37)"""
skip_no_app()
email = f"doublon_{int(time.time())}@regwatch.fr"
client.post("/signup", json={"email": email, "password": "test123", "full_name": "T"})
r = client.post("/signup", json={"email": email, "password": "autre", "full_name": "T"})
assert r.status_code == 400
def test_signup_reussi_retourne_id_email(self):
"""Signup valide → 200 avec id et email"""
skip_no_app()
email = f"ok_{int(time.time())}@regwatch.fr"
r = client.post("/signup", json={"email": email, "password": "test123", "full_name": "OK"})
assert r.status_code == 200
data = r.json()
assert "id" in data
assert data["email"] == email
class TestIntegrationLogin:
"""Tests de POST /login (main.py l.55)"""
def test_email_inexistant_401(self):
"""Email inconnu → 401 'User not found' (l.65-68)"""
skip_no_app()
r = client.post("/login", json={"email": "xyz_inconnu@regwatch.fr", "password": "test"})
assert r.status_code == 401
assert "not found" in r.json()["detail"].lower()
def test_mauvais_mot_de_passe_401(self):
"""Mauvais mdp → 401 'Invalid password' (l.70-74)"""
skip_no_app()
email = f"mdp_{int(time.time())}@regwatch.fr"
client.post("/signup", json={"email": email, "password": "bonmdp", "full_name": "T"})
r = client.post("/login", json={"email": email, "password": "mauvaismdp"})
assert r.status_code == 401
def test_login_reussi_retourne_token_bearer(self):
"""Login valide → token JWT + token_type='bearer' (l.76-82)"""
skip_no_app()
email = f"jwt_{int(time.time())}@regwatch.fr"
client.post("/signup", json={"email": email, "password": "test123", "full_name": "T"})
r = client.post("/login", json={"email": email, "password": "test123"})
assert r.status_code == 200
data = r.json()
assert "access_token" in data
assert data["token_type"] == "bearer"
assert len(data["access_token"]) > 10
def test_corps_vide_422(self):
skip_no_app()
assert client.post("/login", json={}).status_code == 422
class TestIntegrationMe:
"""Tests de GET /me?token=... (main.py l.88)"""
def test_sans_token_422(self):
"""Paramètre token manquant → 422"""
skip_no_app()
assert client.get("/me").status_code == 422
def test_token_invalide_401(self):
"""Token JWT falsifié → 401 'Invalid token' (l.90-93)"""
skip_no_app()
r = client.get("/me?token=token_completement_faux")
assert r.status_code == 401
assert "invalid" in r.json()["detail"].lower() or "token" in r.json()["detail"].lower()
def test_token_valide_retourne_user(self):
"""Token valide → 200 avec email de l'utilisateur"""
skip_no_app()
email = f"me_{int(time.time())}@regwatch.fr"
client.post("/signup", json={"email": email, "password": "test123", "full_name": "T"})
token = client.post("/login", json={"email": email, "password": "test123"}).json()["access_token"]
r = client.get(f"/me?token={token}")
assert r.status_code == 200
assert r.json()["email"] == email
class TestIntegrationDocuments:
"""Tests de GET /documents (main.py l.104)"""
def test_sans_auth_retourne_200(self):
"""Pas de JWT requis sur cet endpoint (main.py l.104)"""
skip_no_app()
assert client.get("/documents").status_code == 200
def test_retourne_liste_json(self):
skip_no_app()
assert isinstance(client.get("/documents").json(), list)
def test_structure_7_champs_par_document(self):
"""Chaque doc a id, title, ingredient, source, type, date, pdf_url (l.115-123)"""
skip_no_app()
docs = client.get("/documents").json()
if docs:
for champ in ["id", "title", "ingredient", "source", "type", "date", "pdf_url"]:
assert champ in docs[0], f"Champ manquant : {champ}"
def test_source_cir_ou_sccs(self):
"""La source est 'CIR' ou 'SCCS' pour chaque document"""
skip_no_app()
docs = client.get("/documents").json()
for doc in docs:
assert doc["source"] in ["CIR", "SCCS"], f"Source inattendue : {doc['source']}"
# ══════════════════════════════════════════════════════════════
# PARTIE 3 — TESTS DE SÉCURITÉ
# ══════════════════════════════════════════════════════════════
class TestSecuriteInjectionSQL:
"""Tests d'injection SQL — SQLAlchemy protège nativement via requêtes paramétrées"""
def test_injection_sql_dans_email_login(self):
"""' OR 1=1 -- dans l'email → 401 ou 422, JAMAIS 500"""
skip_no_app()
r = client.post("/login", json={"email": "' OR 1=1 --", "password": "test"})
assert r.status_code in [401, 422], f"Possible injection SQL : {r.status_code}"
def test_injection_sql_dans_email_signup(self):
"""'; DROP TABLE users; -- → 400 ou 422, JAMAIS 500"""
skip_no_app()
r = client.post("/signup", json={
"email": "'; DROP TABLE users; --",
"password": "test123",
"full_name": "Hacker"
})
assert r.status_code in [400, 422], f"Possible injection SQL : {r.status_code}"
def test_injection_sql_guillemets_doubles(self):
"""\" OR \"1\"=\"1 dans l'email → pas de 500"""
skip_no_app()
r = client.post("/login", json={"email": "\" OR \"1\"=\"1", "password": "test"})
assert r.status_code != 500
class TestSecuriteJWT:
"""Tests de sécurité JWT"""
def test_token_falsifie_retourne_401(self):
"""JWT avec signature incorrecte → 401 'Invalid token'"""
skip_no_app()
r = client.get("/me?token=eyJhbGciOiJIUzI1NiJ9.faux.mauvaise_signature")
assert r.status_code == 401
def test_token_vide_retourne_401_ou_422(self):
"""Token vide → 401 ou 422"""
skip_no_app()
assert client.get("/me?token=").status_code in [401, 422]
def test_token_malformed_retourne_401(self):
"""Token sans points (format JWT invalide) → 401"""
skip_no_app()
assert client.get("/me?token=cecinestunepasjwt").status_code == 401
def test_token_valide_sur_bon_endpoint(self):
"""Un vrai token JWT permet l'accès à /me"""
skip_no_app()
email = f"jwt_ok_{int(time.time())}@regwatch.fr"
client.post("/signup", json={"email": email, "password": "test123", "full_name": "T"})
token = client.post("/login", json={"email": email, "password": "test123"}).json()["access_token"]
assert client.get(f"/me?token={token}").status_code == 200
class TestSecuriteXSS:
"""Tests XSS — l'API REST retourne du JSON, React échappe côté frontend"""
def test_xss_dans_full_name_stocke_sans_crash(self):
"""Payload XSS dans full_name → accepté sans crash (200 ou 422)"""
skip_no_app()
email = f"xss_{int(time.time())}@regwatch.fr"
r = client.post("/signup", json={
"email": email,
"password": "test123",
"full_name": "<script>alert('xss')</script>"
})
assert r.status_code in [200, 422]
def test_xss_dans_email_ne_plante_pas(self):
"""Payload XSS dans l'email → 422 (validation format email)"""
skip_no_app()
r = client.post("/login", json={
"email": "<script>alert(1)</script>@test.com",
"password": "test"
})
assert r.status_code in [401, 422]
# ══════════════════════════════════════════════════════════════
# PARTIE 4 — CONFORMITÉ RGPD
# ══════════════════════════════════════════════════════════════
class TestRGPD:
"""
Tests de conformité RGPD — exigés par le guide Nexa
Vérifie que les données personnelles sont bien protégées
"""
def test_mot_de_passe_absent_de_la_reponse_signup(self):
"""
Le mot de passe ne doit JAMAIS apparaître dans la réponse API
UserResponse = id + email + full_name SEULEMENT (pas de password)
"""
skip_no_app()
email = f"rgpd_{int(time.time())}@regwatch.fr"
r = client.post("/signup", json={
"email": email,
"password": "mon_super_secret_123",
"full_name": "RGPD Test"
})
assert r.status_code == 200
# Le mot de passe en clair ne doit pas apparaître
assert "mon_super_secret_123" not in r.text
# Le champ password ne doit pas être dans la réponse
assert "password" not in r.json()
def test_mot_de_passe_absent_de_la_reponse_login(self):
"""Le login retourne uniquement access_token et token_type"""
skip_no_app()
email = f"rgpd2_{int(time.time())}@regwatch.fr"
client.post("/signup", json={"email": email, "password": "secret456", "full_name": "T"})
r = client.post("/login", json={"email": email, "password": "secret456"})
assert r.status_code == 200
assert "secret456" not in r.text
data = r.json()
assert "password" not in data
assert set(data.keys()) <= {"access_token", "token_type"}
def test_mot_de_passe_absent_de_la_reponse_me(self):
"""GET /me ne retourne pas le mot de passe"""
skip_no_app()
email = f"rgpd3_{int(time.time())}@regwatch.fr"
client.post("/signup", json={"email": email, "password": "secret789", "full_name": "T"})
token = client.post("/login", json={"email": email, "password": "secret789"}).json()["access_token"]
r = client.get(f"/me?token={token}")
assert r.status_code == 200
assert "secret789" not in r.text
assert "password" not in r.json()
def test_hachage_argon2_dans_requirements(self):
"""
La dépendance passlib[argon2] est bien dans requirements.txt
→ preuve que les mots de passe sont hachés avec Argon2
"""
req_path = os.path.join(
os.path.dirname(__file__), "..", "requirements.txt"
)
if os.path.exists(req_path):
with open(req_path) as f:
content = f.read()
assert "argon2" in content.lower(), \
"passlib[argon2] absent de requirements.txt"
else:
pytest.skip("requirements.txt non trouvé")
def test_token_jwt_expire_apres_acces(self):
"""
Un token JWT est bien structuré (3 parties séparées par des points)
→ preuve que l'expiration est encodée dans le payload
"""
skip_no_app()
email = f"expire_{int(time.time())}@regwatch.fr"
client.post("/signup", json={"email": email, "password": "test123", "full_name": "T"})
token = client.post("/login", json={"email": email, "password": "test123"}).json()["access_token"]
# Un JWT valide a toujours 3 parties séparées par des points
parts = token.split(".")
assert len(parts) == 3, "Le token JWT n'a pas la structure header.payload.signature"
def test_cors_autorise_frontend(self):
"""
Le middleware CORS permet au frontend React de communiquer avec l'API
allow_origins=["*"] configuré dans main.py l.13-18
"""
skip_no_app()
r = client.get("/", headers={"Origin": "http://localhost:5173"})
assert r.status_code == 200