Files
mostovik-backend/tests/apps/parsers/test_exchange_spb.py
2026-10-05 17:04:11 +02:00

181 lines
6.0 KiB
Python

"""Public CSV contract derived from the current SPB listing export."""
import csv
import importlib
from io import BytesIO, StringIO
from pathlib import Path
import pytest
from apps.parsers.registry_snapshots import SnapshotValidationError
from django.test import override_settings
FIXTURES = Path(__file__).parents[2] / "fixtures" / "exchanges"
def adapter():
return importlib.import_module("apps.parsers.exchange_spb")
def data(**changes):
rows = list(
csv.reader(
StringIO((FIXTURES / "listing.csv").read_text(encoding="utf-8-sig")),
delimiter=";",
)
)
for key, value in changes.items():
rows[1][rows[0].index(key)] = value
stream = StringIO()
csv.writer(stream, delimiter=";").writerows(rows)
return BytesIO(stream.getvalue().encode())
def test_preserves_inn_vendor_identity_and_decimal():
row = next(adapter().iter_spb_records(data()))
assert row.inn == "0123456789"
assert row.external_id == "spb:fixture-security-1"
assert row.payload["exchange"] == "spb"
assert row.amount is None
assert row.payload["nominal_value"] == "0.1"
assert row.payload["listing_date"] == "2024-11-06"
assert row.payload["duration"] is None
assert row.payload["yield"] is None
@pytest.mark.parametrize("name", ["unknown_headers.csv"])
def test_unknown_headers_rejected(name):
with (FIXTURES / name).open("rb") as f, pytest.raises(SnapshotValidationError):
list(adapter().iter_spb_records(f))
@pytest.mark.parametrize("value", [b"not csv", b"PK\x03\x04invalid", b""])
def test_invalid_input_rejected(value):
with pytest.raises(SnapshotValidationError):
list(adapter().iter_spb_records(BytesIO(value)))
def test_genuine_empty_with_valid_headers():
with (FIXTURES / "empty.csv").open("rb") as f:
assert list(adapter().iter_spb_records(f)) == []
def test_missing_inn_stays_missing_and_name_is_not_identity():
row = next(adapter().iter_spb_records(data(e_INN_code="")))
assert row.inn == ""
assert row.organization_uid is None
@pytest.mark.parametrize(
"field,value",
[("s_RTS_code", ""), ("s_quot_list_in_date", "tomorrow"), ("s_face_value", "NaN")],
)
def test_malformed_required_or_numeric_fields_rejected(field, value):
with pytest.raises(SnapshotValidationError):
list(adapter().iter_spb_records(data(**{field: value})))
def test_bounded_rows():
with override_settings(EXCHANGE_SPB_MAX_ROWS=0), pytest.raises(
SnapshotValidationError
):
list(adapter().iter_spb_records(data()))
def test_latest_date_not_link_order():
html = '<a href="/listing/securities/list">Список ценных бумаг на 05.10.2026</a><a href="/listing/securities/list/">Список ценных бумаг на 02.10.2026</a>'
assert adapter().discover_spb_date(html).isoformat() == "2026-10-05"
def test_no_dated_current_link_rejected():
with pytest.raises(SnapshotValidationError):
adapter().discover_spb_date(
'<a href="https://evil.test/list">Список ценных бумаг на 05.10.2026</a>'
)
def test_handle_left_open_and_text_inn_not_coerced():
handle = data(e_INN_code="0010000025")
assert next(adapter().iter_spb_records(handle)).inn == "0010000025"
assert not handle.closed
def test_tls_trust_stays_default():
session = adapter().registry_session()
assert session.verify is True
session.close()
@pytest.mark.django_db
def test_download_checksum_date_private_file_and_three_bounded_reads(tmp_path):
import hashlib
from contextlib import contextmanager
from unittest.mock import patch
module = adapter()
body = (FIXTURES / "listing.csv").read_bytes()
html = b'<a href="/listing/securities/list">\xd0\xa1\xd0\xbf\xd0\xb8\xd1\x81\xd0\xbe\xd0\xba \xd1\x86\xd0\xb5\xd0\xbd\xd0\xbd\xd1\x8b\xd1\x85 \xd0\xb1\xd1\x83\xd0\xbc\xd0\xb0\xd0\xb3 \xd0\xbd\xd0\xb0 05.10.2026</a>'
urls = []
@contextmanager
def response(session, url, **kwargs):
urls.append(url)
class Response:
def iter_content(self, chunk_size):
yield body if url == module.SPB_EXPORT else html
yield Response()
with override_settings(PARSER_PRIVATE_ARTIFACT_ROOT=str(tmp_path)), patch.object(
module, "registry_response", response
):
artifact = module.download_spb_artifact(load_batch=1)
assert artifact.sha256 == hashlib.sha256(body).hexdigest()
assert artifact.source_published_at.isoformat() == "2026-10-05"
assert artifact.metadata["input_only"] is True
assert len(urls) == 3
with pytest.raises(ValueError):
_ = artifact.file.url
@pytest.mark.django_db
def test_download_size_limit_rejects_artifact(tmp_path):
from contextlib import contextmanager
from unittest.mock import patch
from apps.parsers.models import ParserSourceArtifact
module = adapter()
html = '<a href="/listing/securities/list">Список ценных бумаг на 05.10.2026</a>'.encode()
@contextmanager
def response(session, url, **kwargs):
class Response:
def iter_content(self, chunk_size):
yield b"oversized" if url == module.SPB_EXPORT else html
yield Response()
with override_settings(
PARSER_PRIVATE_ARTIFACT_ROOT=str(tmp_path), EXCHANGE_SPB_MAX_BYTES=1
), patch.object(module, "registry_response", response), pytest.raises(
SnapshotValidationError
):
module.download_spb_artifact(load_batch=1)
assert ParserSourceArtifact.objects.get().status == "rejected"
def test_redirect_host_rejected_before_second_request():
from unittest.mock import Mock
module = adapter()
session = Mock()
response = Mock(
is_redirect=True, headers={"Location": "https://evil.test/export.csv"}
)
session.get.return_value = response
with pytest.raises(SnapshotValidationError), module.registry_response(
session, module.SPB_EXPORT, hosts=module.SPB_HOSTS
):
pytest.fail("unsafe redirect yielded")
assert session.get.call_count == 1