arm-tabular-parser/api/tests/unit/test_fixtures_helpers.py
Andrey Epifancev 9b8b8ea727 fix: финальное ревью — лимиты чтения xlsx (I-1), лог traceback (I-2), ACCOUNT_RE (M-1), ИНН/КПП (M-3)
- I-1 (R-23): readers/_limits.py — сумма ZipInfo.file_size ≤ 10 × PARSE_MAX_BYTES (параметр read_book от
  OsvParser), потоковый скан <mergeCell ref> до openpyxl и площадь merged после чтения любого формата
  (MERGED_AREA_LIMIT = 200 000; реальный максимум 10 090) → UNSUPPORTED_FORMAT; дедлайн в циклах detect
  (поиск строки групп, дополнительные колонки); README «Лимиты» исправлен, «Известные ограничения» (defusedxml).
- I-2: OsvParser логирует стек непредвиденной ошибки (loguru ERROR), UNREADABLE — WARNING без стека.
- M-1: ACCOUNT_RE с re.ASCII | re.DOTALL (многострочное название после запятой, только цифры 0–9).
- M-3 (R-24): правило детектора ИНН не дополняет 9/11 цифр нулём; дополнение — только при извлечении
  (numbers.inn_digits) из колонки, выбранной по заголовку или по строгому правилу.
- T9.1: тесты R-21 (свёрнутый родитель на противоположной стороне; (0,0) при детях (X, X)).
- T1.1: хелперы фикстур пропускают скрытые файлы/каталоги и lock-файлы Excel (~$…).
- T12.3: golden-update не удаляет снимки части без исходного каталога (osv/synthetic или реальные файлы).

Golden-снимки не изменились (82/82).

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_012vJkmBGKKqMoEMzvsXG3y2
2026-10-05 16:29:53 +04:00

104 lines
3.9 KiB
Python
Raw Blame History

This file contains ambiguous Unicode characters

This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.

"""Хелперы доступа к ../fixtures: каталог из ARM_KA_FIXTURES, список файлов ОСВ без синтетики, причина skip."""
from pathlib import Path
import pytest
from fixtures_helpers import (
FIXTURES_ENV,
fixtures_skip_reason,
osv_fixture_files,
resolve_fixtures_dir,
synthetic_files,
synthetic_skip_reason,
)
def test_env_overrides_fixtures_dir(tmp_path: Path, monkeypatch: pytest.MonkeyPatch) -> None:
monkeypatch.setenv(FIXTURES_ENV, str(tmp_path))
assert resolve_fixtures_dir() == tmp_path
def test_default_fixtures_dir_is_sibling_of_repository(monkeypatch: pytest.MonkeyPatch) -> None:
monkeypatch.delenv(FIXTURES_ENV, raising=False)
assert resolve_fixtures_dir() == Path(__file__).resolve().parents[4] / "fixtures"
def test_no_osv_directory_gives_no_files(tmp_path: Path) -> None:
assert osv_fixture_files(tmp_path) == []
assert synthetic_files(tmp_path) == []
def test_synthetic_excluded_and_foreign_files_ignored(tmp_path: Path) -> None:
osv = tmp_path / "osv"
(osv / "rb").mkdir(parents=True)
(osv / "synthetic").mkdir()
(osv / "rb" / "b.xlsx").write_bytes(b"")
(osv / "a.xls").write_bytes(b"")
(osv / "c.xlsb").write_bytes(b"")
(osv / ".DS_Store").write_bytes(b"")
(osv / "notes.txt").write_bytes(b"")
(osv / "synthetic" / "s.xlsx").write_bytes(b"")
(osv / "synthetic" / ".gitkeep").write_bytes(b"")
(osv / "synthetic" / "README.md").write_bytes(b"")
assert osv_fixture_files(tmp_path) == [osv / "a.xls", osv / "c.xlsb", osv / "rb" / "b.xlsx"]
assert synthetic_files(tmp_path) == [osv / "synthetic" / "s.xlsx"]
def test_hidden_and_excel_lock_files_are_ignored(tmp_path: Path) -> None:
"""T1.1: скрытые файлы и каталоги и lock-файлы Excel (`~$…`, пока книга открыта) — не фикстуры."""
osv = tmp_path / "osv"
(osv / "rb" / ".cache").mkdir(parents=True)
(osv / "synthetic").mkdir()
for path in (
osv / "rb" / "b.xlsx",
osv / "rb" / "~$b.xlsx",
osv / "rb" / ".b.xlsx",
osv / "rb" / ".cache" / "c.xlsx",
osv / "synthetic" / "s.xlsx",
osv / "synthetic" / "~$s.xlsx",
osv / "synthetic" / ".s.xls",
):
path.write_bytes(b"")
assert osv_fixture_files(tmp_path) == [osv / "rb" / "b.xlsx"]
assert synthetic_files(tmp_path) == [osv / "synthetic" / "s.xlsx"]
def test_skip_reason_mentions_env_and_path(tmp_path: Path, monkeypatch: pytest.MonkeyPatch) -> None:
missing = tmp_path / "нет-такого"
monkeypatch.setenv(FIXTURES_ENV, str(missing))
reason = fixtures_skip_reason()
assert reason is not None
assert FIXTURES_ENV in reason
assert str(missing) in reason
def test_skip_reason_none_when_available(tmp_path: Path, monkeypatch: pytest.MonkeyPatch) -> None:
(tmp_path / "osv").mkdir()
monkeypatch.setenv(FIXTURES_ENV, str(tmp_path))
assert fixtures_skip_reason() is None
def test_synthetic_skip_reason_when_fixtures_exist_but_synthetic_is_missing_or_empty(
tmp_path: Path, monkeypatch: pytest.MonkeyPatch
) -> None:
(tmp_path / "osv").mkdir()
monkeypatch.setenv(FIXTURES_ENV, str(tmp_path))
reason = synthetic_skip_reason()
assert reason is not None
assert "osv" in reason
assert "synthetic" in reason
(tmp_path / "osv" / "synthetic").mkdir()
(tmp_path / "osv" / "synthetic" / "README.md").write_bytes(b"")
assert synthetic_skip_reason() == reason
(tmp_path / "osv" / "synthetic" / "s01.xlsx").write_bytes(b"")
assert synthetic_skip_reason() is None
def test_synthetic_skip_reason_is_none_without_fixtures(tmp_path: Path, monkeypatch: pytest.MonkeyPatch) -> None:
monkeypatch.setenv(FIXTURES_ENV, str(tmp_path / "нет"))
assert synthetic_skip_reason() is None # причину тогда даёт fixtures_skip_reason