sufw-anweisenheiten/tests/test_validators.py

"""Tests for src/validators.py — pre-generation validation logic."""

import json
import re
from pathlib import Path
from unittest.mock import MagicMock, patch

import pytest

from src.validators import validate


# ---------------------------------------------------------------------------
# Helpers
# ---------------------------------------------------------------------------

def _write_config(tmp_path: Path, overrides: dict = None) -> Path:
    """Write a valid config.json into tmp_path, optionally applying overrides."""
    cfg = {
        "antragsnummer": "12345",
        "ort": "Teststadt",
        "template_pdf": "template/sab.pdf",
        "output_dir": "output",
        "excel_year_pattern": "(19\\d{2}|20\\d{2})",
        "excel_name_header": "Name, Vorname",
        "excel_name_column": 1,
        "excel_seq_column": 0,
        "excel_day_start_col": 9,
        "excel_day_end_col": 39,
        "month_mapping": {"Feb": "02", "Jan": "01"},
        "valid_status_codes": ["K", "KK", "E", "F", "P", "U", "A", "V"],
        "plan_hours_value": "6",
        "pdf_field_ort": "unterschrift.ort",
        "pdf_field_erstellungsdatum": "unterschrift.datum",
        "pdf_pauschale_prefix": "pauschale_5",
        "pauschale_present_value": "1",
    }
    if overrides:
        cfg.update(overrides)
    cfg_path = tmp_path / "config.json"
    cfg_path.write_text(json.dumps(cfg), encoding="utf-8")
    return cfg_path


def _write_dummy_xlsx(tmp_path: Path, name: str = "Anwesenheit_2026.xlsx") -> Path:
    """Create a minimal valid xlsx file using openpyxl."""
    import openpyxl
    wb = openpyxl.Workbook()
    ws = wb.active
    ws.title = "Feb"
    # Header row
    ws.cell(row=1, column=2, value="Name, Vorname")
    # Day headers (columns J–AN = col 10–40 in 1-based)
    for col in range(10, 41):
        ws.cell(row=1, column=col, value=col - 9)  # day number as header
    # Person row
    ws.cell(row=2, column=1, value=1)
    ws.cell(row=2, column=2, value="Mustermann, Max")
    ws.cell(row=2, column=10, value=6.0)  # day 1 → present 6h
    xlsx_path = tmp_path / name
    wb.save(str(xlsx_path))
    return xlsx_path


def _write_dummy_pdf(tmp_path: Path) -> Path:
    """Create a 1-byte placeholder file that passes existence checks."""
    pdf_path = tmp_path / "template" / "sab.pdf"
    pdf_path.parent.mkdir(parents=True, exist_ok=True)
    pdf_path.write_bytes(b"%PDF-1.4 placeholder")
    return pdf_path


# ---------------------------------------------------------------------------
# Tests
# ---------------------------------------------------------------------------

class TestValidate:
    def test_all_checks_pass(self, tmp_path):
        cfg_path = _write_config(tmp_path)
        xlsx_path = _write_dummy_xlsx(tmp_path)
        _write_dummy_pdf(tmp_path)

        # Reset config cache so the test config is used
        from src.config_loader import reset_config_cache
        reset_config_cache()

        errors = validate(xlsx_path, "Feb", cfg_path)
        assert errors == []

    def test_excel_file_missing(self, tmp_path):
        cfg_path = _write_config(tmp_path)
        _write_dummy_pdf(tmp_path)

        from src.config_loader import reset_config_cache
        reset_config_cache()

        errors = validate(tmp_path / "nonexistent.xlsx", "Feb", cfg_path)
        assert any("nicht gefunden" in e for e in errors)

    def test_year_not_in_filename(self, tmp_path):
        cfg_path = _write_config(tmp_path)
        xlsx_path = _write_dummy_xlsx(tmp_path, name="Anwesenheit_kein_jahr.xlsx")
        _write_dummy_pdf(tmp_path)

        from src.config_loader import reset_config_cache
        reset_config_cache()

        errors = validate(xlsx_path, "Feb", cfg_path)
        assert any("Jahr" in e for e in errors)

    def test_unknown_sheet_name(self, tmp_path):
        cfg_path = _write_config(tmp_path)
        xlsx_path = _write_dummy_xlsx(tmp_path)
        _write_dummy_pdf(tmp_path)

        from src.config_loader import reset_config_cache
        reset_config_cache()

        errors = validate(xlsx_path, "Xyz", cfg_path)
        assert any("month_mapping" in e or "Sheet" in e for e in errors)

    def test_pdf_template_missing(self, tmp_path):
        # No _write_dummy_pdf call → template missing
        cfg_path = _write_config(tmp_path)
        xlsx_path = _write_dummy_xlsx(tmp_path)

        from src.config_loader import reset_config_cache
        reset_config_cache()

        errors = validate(xlsx_path, "Feb", cfg_path)
        assert any("Template" in e or "template" in e.lower() for e in errors)

    def test_config_missing_key(self, tmp_path):
        cfg = {
            "antragsnummer": "123",
            # omit 'ort' and other required keys
        }
        cfg_path = tmp_path / "config.json"
        cfg_path.write_text(json.dumps(cfg), encoding="utf-8")

        from src.config_loader import reset_config_cache
        reset_config_cache()

        errors = validate(tmp_path / "file.xlsx", "Feb", cfg_path)
        assert any("Konfigurationsfehler" in e or "fehlen" in e.lower() for e in errors)

    def test_non_xlsx_extension_flagged(self, tmp_path):
        cfg_path = _write_config(tmp_path)
        _write_dummy_pdf(tmp_path)

        csv_path = tmp_path / "Anwesenheit_2026.csv"
        csv_path.write_text("dummy")

        from src.config_loader import reset_config_cache
        reset_config_cache()

        errors = validate(csv_path, "Feb", cfg_path)
        assert any(".xlsx" in e or "Excel" in e for e in errors)