Estructura inicial funcional con documentación en el README.md de la FastAPI
This commit is contained in:
202
tests/test_api.py
Normal file
202
tests/test_api.py
Normal file
@@ -0,0 +1,202 @@
|
||||
"""Pruebas de la API con un Elasticsearch simulado.
|
||||
|
||||
Las respuestas simuladas tienen la misma forma que las de un índice de FSCrawler
|
||||
(campos file.*, path.virtual, matched_queries, highlight). Lo que NO prueban es
|
||||
cómo Elasticsearch interpreta la consulta: eso se verifica contra el servidor real
|
||||
(ver README, "Verificar contra Elasticsearch").
|
||||
"""
|
||||
import json
|
||||
import time
|
||||
|
||||
import httpx
|
||||
import jwt
|
||||
import pytest
|
||||
from fastapi.testclient import TestClient
|
||||
|
||||
SECRETO = "secreto-de-prueba-con-mas-de-32-caracteres"
|
||||
|
||||
DOCS = {
|
||||
"a1b2": {"file": {"filename": "cv_juan.docx", "extension": "docx", "filesize": 30,
|
||||
"last_modified": "2026-05-10T12:00:00.000+00:00"},
|
||||
"path": {"virtual": "/CURRICULUMS/cv_juan.docx"}},
|
||||
"c3d4": {"file": {"filename": "resolucion_45.pdf", "extension": "pdf", "filesize": 26,
|
||||
"last_modified": "2025-01-02T08:00:00.000+00:00"},
|
||||
"path": {"virtual": "/DEMANDAS - VARIOS/resolucion_45.pdf"}},
|
||||
"e5f6": {"file": {"filename": "informe.txt", "extension": "txt", "filesize": 10},
|
||||
"path": {"virtual": "/DATOS POLICIA/informe.txt"}},
|
||||
"malo": {"file": {"filename": "passwd"}, "path": {"virtual": "/../../etc/passwd"}},
|
||||
}
|
||||
|
||||
|
||||
class FakeES:
|
||||
"""Imita las dos llamadas que usa la API: _search y _doc/{id}."""
|
||||
|
||||
def __init__(self):
|
||||
self.ultima_busqueda = None
|
||||
self.caido = False
|
||||
self.hits = []
|
||||
|
||||
def __call__(self, request: httpx.Request) -> httpx.Response:
|
||||
if self.caido:
|
||||
raise httpx.ConnectError("sin conexión")
|
||||
if request.url.path.endswith("/_search"):
|
||||
self.ultima_busqueda = json.loads(request.content)
|
||||
return httpx.Response(200, json={"hits": {"total": {"value": len(self.hits), "relation": "eq"},
|
||||
"hits": self.hits}})
|
||||
if "/_doc/" in request.url.path:
|
||||
doc_id = request.url.path.rsplit("/", 1)[1]
|
||||
if doc_id in DOCS:
|
||||
return httpx.Response(200, json={"_id": doc_id, "found": True, "_source": DOCS[doc_id]})
|
||||
return httpx.Response(404, json={"found": False})
|
||||
if request.url.path.endswith("/_count"):
|
||||
return httpx.Response(200, json={"count": len(DOCS)})
|
||||
return httpx.Response(400)
|
||||
|
||||
|
||||
def _hit(doc_id, nombradas, fragmento):
|
||||
return {"_id": doc_id, "_score": 1, "_source": DOCS[doc_id], "matched_queries": nombradas,
|
||||
"highlight": {"content": [fragmento]}}
|
||||
|
||||
|
||||
@pytest.fixture()
|
||||
def api(tmp_path, monkeypatch):
|
||||
raiz = tmp_path / "archivos"
|
||||
(raiz / "CURRICULUMS").mkdir(parents=True)
|
||||
(raiz / "DEMANDAS - VARIOS").mkdir()
|
||||
(raiz / "DATOS POLICIA").mkdir()
|
||||
(raiz / "CURRICULUMS" / "cv_juan.docx").write_bytes(b"contenido del curriculum")
|
||||
(raiz / "DATOS POLICIA" / "informe.txt").write_bytes(b"reservado")
|
||||
|
||||
monkeypatch.chdir(tmp_path)
|
||||
for k, v in {"STORAGE_TYPE": "local", "STORAGE_ROOT": str(raiz), "EXCLUDE_FOLDERS": "DATOS POLICIA",
|
||||
"ES_URL": "http://es:9200", "ES_INDEX": "archivos",
|
||||
"DATABASE_URL": f"sqlite:///{tmp_path / 'logs.db'}", "JWT_SECRET": SECRETO,
|
||||
"DOWNLOAD_SECRET": "descarga-de-prueba-con-mas-de-32-caracteres",
|
||||
"LOG_FILE": str(tmp_path / "logs" / "api.log")}.items():
|
||||
monkeypatch.setenv(k, v)
|
||||
|
||||
import app.config
|
||||
import app.database as dbm
|
||||
import app.elastic
|
||||
import app.main
|
||||
import app.storage
|
||||
|
||||
app.config.get_settings.cache_clear()
|
||||
app.storage.get_storage.cache_clear()
|
||||
app.elastic.cliente.cache_clear()
|
||||
fake = FakeES()
|
||||
monkeypatch.setattr(app.elastic, "cliente",
|
||||
lambda: httpx.Client(base_url="http://es:9200", transport=httpx.MockTransport(fake)))
|
||||
dbm.engine = dbm._crear_engine()
|
||||
dbm.SessionLocal.configure(bind=dbm.engine)
|
||||
with TestClient(app.main.app) as c:
|
||||
c.es = fake
|
||||
yield c
|
||||
|
||||
|
||||
def auth(rol="admin", exp_seg=600):
|
||||
t = jwt.encode({"sub": "7", "rol": rol, "exp": int(time.time()) + exp_seg}, SECRETO, "HS256")
|
||||
return {"Authorization": f"Bearer {t}"}
|
||||
|
||||
|
||||
# ---------------------------------------------------------------- la consulta que se manda
|
||||
|
||||
def test_consulta_cedula_busca_las_tres_formas_y_la_ruta():
|
||||
from app.busqueda import variantes_cedula
|
||||
|
||||
assert variantes_cedula("1234567") == ["1234567", "1.234.567", "1,234,567"]
|
||||
assert variantes_cedula("123456") == ["123456", "123.456", "123,456"]
|
||||
|
||||
|
||||
def test_consulta_armada(api):
|
||||
api.get("/archivos", params={"cedula": "1.234.567", "nombres": "Juan Carlos",
|
||||
"apellidos": "Pérez González"}, headers=auth())
|
||||
q = api.es.ultima_busqueda
|
||||
texto = json.dumps(q, ensure_ascii=False)
|
||||
should = q["query"]["bool"]["should"]
|
||||
assert [c["constant_score"]["_name"] for c in should] == ["cedula", "nombre_completo", "nombre_parcial"]
|
||||
assert '"1.234.567"' in texto and '"1234567"' in texto and "*1234567*" in texto
|
||||
# Nombre: busca con y sin tilde, en cualquier orden
|
||||
completo = should[1]["constant_score"]["filter"]["intervals"]["content"]["all_of"]
|
||||
assert completo["ordered"] is False and len(completo["intervals"]) == 4
|
||||
assert '"pérez"' in texto and '"perez"' in texto
|
||||
parcial = should[2]["constant_score"]["filter"]["intervals"]["content"]["all_of"]
|
||||
assert len(parcial["intervals"]) == 2 and parcial["max_gaps"] == 2
|
||||
# Carpeta excluida y nunca se trae el contenido completo
|
||||
assert q["query"]["bool"]["must_not"] == [{"prefix": {"path.virtual": "/DATOS POLICIA/"}}]
|
||||
assert "content" not in q["_source"]
|
||||
assert q["highlight"]["pre_tags"] == ["«"]
|
||||
|
||||
|
||||
def test_caracteres_raros_no_llegan_a_la_consulta(api):
|
||||
api.get("/archivos", params={"nombres": 'Juan" OR *', "apellidos": "Pérez}{"}, headers=auth())
|
||||
texto = json.dumps(api.es.ultima_busqueda)
|
||||
assert '"OR"' not in texto.replace('"or"', "") and "}{" not in texto
|
||||
|
||||
|
||||
# ---------------------------------------------------------------- la respuesta de la API
|
||||
|
||||
def test_respuesta_con_coincidencias(api):
|
||||
api.es.hits = [
|
||||
_hit("a1b2", ["cedula", "nombre_completo", "nombre_parcial"], "Juan Carlos Pérez\nC.I. «1.234.567»"),
|
||||
_hit("c3d4", ["nombre_parcial"], "Demandado: «Perez», «Juan»"),
|
||||
]
|
||||
r = api.get("/archivos", params={"cedula": "1234567", "nombres": "Juan Carlos",
|
||||
"apellidos": "Pérez González"}, headers=auth())
|
||||
assert r.status_code == 200
|
||||
d = r.json()
|
||||
assert d["total"] == d["total_coincidencias"] == 2
|
||||
a, b = d["resultados"]
|
||||
assert (a["id"], a["nombre"], a["carpeta"], a["extension"]) == ("a1b2", "cv_juan.docx", "CURRICULUMS", ".docx")
|
||||
assert a["coincide_cedula"] is True and a["coincide_nombre"] == "completo"
|
||||
assert a["fragmento"] == "Juan Carlos Pérez C.I. «1.234.567»"
|
||||
assert b["coincide_cedula"] is False and b["coincide_nombre"] == "parcial"
|
||||
|
||||
|
||||
def test_validaciones_y_seguridad(api):
|
||||
assert api.get("/archivos", headers=auth()).status_code == 422
|
||||
assert api.get("/archivos", params={"nombres": "Juan"}, headers=auth()).status_code == 422
|
||||
assert api.get("/archivos", params={"cedula": "12"}, headers=auth()).status_code == 422
|
||||
assert api.get("/archivos", params={"cedula": "1234567"}).status_code == 401
|
||||
assert api.get("/archivos", params={"cedula": "1234567"}, headers=auth(rol="invitado")).status_code == 403
|
||||
assert api.get("/archivos", params={"cedula": "1234567"}, headers=auth(exp_seg=-5)).status_code == 401
|
||||
|
||||
|
||||
def test_elasticsearch_caido(api):
|
||||
api.es.caido = True
|
||||
assert api.get("/archivos", params={"cedula": "1234567"}, headers=auth()).status_code == 503
|
||||
salud = api.get("/salud").json()
|
||||
assert salud["estado"] == "degradado" and salud["elasticsearch_accesible"] is False
|
||||
|
||||
|
||||
# ---------------------------------------------------------------- descargas
|
||||
|
||||
def _enlace(api, doc_id):
|
||||
api.es.hits = [_hit(doc_id, ["cedula"], "x")]
|
||||
url = api.get("/archivos", params={"cedula": "1234567"}, headers=auth()).json()["resultados"][0]["url_descarga"]
|
||||
return url.split("/descargas/", 1)[1]
|
||||
|
||||
|
||||
def test_descarga_ok_y_enlace_manipulado(api):
|
||||
enlace = _enlace(api, "a1b2")
|
||||
r = api.get("/descargas/" + enlace)
|
||||
assert r.status_code == 200 and r.content == b"contenido del curriculum"
|
||||
assert api.get("/descargas/" + enlace.replace("a1b2", "c3d4", 1)).status_code == 403 # firma no coincide
|
||||
|
||||
|
||||
def test_descarga_de_archivo_que_ya_no_esta_en_disco(api):
|
||||
assert api.get("/descargas/" + _enlace(api, "c3d4")).status_code == 404
|
||||
|
||||
|
||||
def test_descarga_de_carpeta_excluida_bloqueada(api):
|
||||
assert api.get("/descargas/" + _enlace(api, "e5f6")).status_code == 404
|
||||
|
||||
|
||||
def test_ruta_maliciosa_en_elasticsearch_no_sale_de_la_raiz(api):
|
||||
# Aunque alguien lograra meter un documento con una ruta "../..", no se entrega
|
||||
assert api.get("/descargas/" + _enlace(api, "malo")).status_code == 404
|
||||
|
||||
|
||||
def test_salud_ok(api):
|
||||
assert api.get("/salud").json() == {"estado": "ok", "elasticsearch_accesible": True,
|
||||
"tipo_almacenamiento": "local", "almacenamiento_accesible": True}
|
||||
Reference in New Issue
Block a user