familienarchiv/tools/import-normalizer/tests/test_persons_tree.py

import sys
from pathlib import Path
sys.path.insert(0, str(Path(__file__).parent.parent))

import persons_tree


def test_parse_year_iso_string():
    assert persons_tree._parse_year("1920-09-20") == 1920


def test_parse_year_excel_serial_birth():
    # 7568 days from 1899-12-30 = 1920-09-19 or -20 depending on leap counting
    assert persons_tree._parse_year("7568") == 1920


def test_parse_year_excel_serial_death():
    # 36222 days from 1899-12-30 ≈ 1999
    assert persons_tree._parse_year("36222") == 1999


def test_parse_year_excel_serial_small():
    # 177 days from 1899-12-30 = 1900-06-25
    assert persons_tree._parse_year("177") == 1900


def test_parse_year_german_date_string():
    assert persons_tree._parse_year("30.8.1862") == 1862


def test_parse_year_year_only():
    assert persons_tree._parse_year("1930") == 1930


def test_parse_year_free_text():
    assert persons_tree._parse_year("August 1941") == 1941


def test_parse_year_none():
    assert persons_tree._parse_year(None) is None


def test_parse_year_empty():
    assert persons_tree._parse_year("") is None


def test_parse_year_unresolvable_truncated():
    # "2.9.196" has no valid 4-digit year — returns None
    assert persons_tree._parse_year("2.9.196") is None


def test_parse_year_typo_year():
    # "4.3.1023" — year 1023 outside 1700-2100 guard — returns None
    assert persons_tree._parse_year("4.3.1023") is None


def test_parse_year_bare_out_of_range_year_is_none():
    # "1023" is a plausible typo for "1923" but is NOT an Excel serial.
    # parse_date("1023") parses it as year 1023 (out of 1700-2100 guard).
    # The serial branch must NOT re-interpret it as a serial.
    assert persons_tree._parse_year("1023") is None


def test_parse_generation_space():
    assert persons_tree._parse_generation("G 3") == 3


def test_parse_generation_no_space():
    assert persons_tree._parse_generation("G3") == 3


def test_parse_generation_extra_spaces():
    assert persons_tree._parse_generation("G  0") == 0


def test_parse_generation_trailing_garbage():
    assert persons_tree._parse_generation("G 2         de Gruyter") == 2


def test_parse_generation_empty():
    assert persons_tree._parse_generation("") is None


def test_parse_generation_none():
    assert persons_tree._parse_generation(None) is None


def test_norm_tree_basic():
    assert persons_tree._norm_tree("Werner Allemeyer") == "werner allemeyer"


def test_norm_tree_diacritics():
    assert persons_tree._norm_tree("Wöhler") == "woehler"


def test_norm_tree_strips_parens():
    assert persons_tree._norm_tree("Otto (Herbert)") == "otto"


def test_norm_tree_strips_quotes():
    assert persons_tree._norm_tree('"Tante Lolly"') == "tante lolly"


def test_norm_tree_strips_geographic_suffix():
    assert persons_tree._norm_tree("Walter Cram Aachen") == "walter cram"


def test_norm_tree_strips_mexiko():
    assert persons_tree._norm_tree("Hans Cram Mexiko") == "hans cram"


def test_norm_tree_collapses_whitespace():
    assert persons_tree._norm_tree("  Clara   de Gruyter  ") == "clara de gruyter"


def test_build_index_forward_lookup():
    persons = [{"rowId": "row_002", "firstName": "Werner", "lastName": "Allemeyer", "maidenName": None}]
    idx = persons_tree._build_index(persons)
    assert "werner allemeyer" in idx
    assert idx["werner allemeyer"] == ["row_002"]


def test_build_index_reversed_lookup():
    persons = [{"rowId": "row_002", "firstName": "Werner", "lastName": "Allemeyer", "maidenName": None}]
    idx = persons_tree._build_index(persons)
    assert idx.get("allemeyer werner") == ["row_002"]


def test_build_index_maiden_name_lookup():
    persons = [{"rowId": "row_002", "firstName": "Elsgard", "lastName": "Allemeyer", "maidenName": "Wöhler"}]
    idx = persons_tree._build_index(persons)
    assert idx.get("elsgard woehler") == ["row_002"]


def test_build_index_single_token_fallback():
    persons = [{"rowId": "row_028", "firstName": "Herbert", "lastName": "Cram", "maidenName": None}]
    idx = persons_tree._build_index(persons)
    assert idx.get("cram") == ["row_028"]


def test_build_index_ambiguous_single_token():
    persons = [
        {"rowId": "row_028", "firstName": "Herbert", "lastName": "Cram", "maidenName": None},
        {"rowId": "row_019", "firstName": "Clara", "lastName": "Cram", "maidenName": None},
    ]
    idx = persons_tree._build_index(persons)
    assert set(idx["cram"]) == {"row_028", "row_019"}


def test_resolve_one_found():
    persons = [{"rowId": "row_003", "firstName": "Werner", "lastName": "Allemeyer", "maidenName": None}]
    idx = persons_tree._build_index(persons)
    row_id, reason = persons_tree._resolve_one("Allemeyer Werner", idx)
    assert row_id == "row_003"
    assert reason is None


def test_resolve_one_not_found():
    idx = {}
    row_id, reason = persons_tree._resolve_one("Nobody Unknown", idx)
    assert row_id is None
    assert reason == "not_found"


def test_resolve_one_ambiguous():
    persons = [
        {"rowId": "row_028", "firstName": "Herbert", "lastName": "Cram", "maidenName": None},
        {"rowId": "row_019", "firstName": "Clara", "lastName": "Cram", "maidenName": None},
    ]
    idx = persons_tree._build_index(persons)
    row_id, reason = persons_tree._resolve_one("Cram", idx)
    assert row_id is None
    assert reason == "ambiguous"


def test_parse_row_serial_dates():
    fields = {
        "generation": "G 3", "last_name": "Allemeyer", "first_name": "Elsgard",
        "maiden_name": "Wöhler", "birth_date": "7568", "birth_place": "Garz",
        "death_date": "36222", "death_place": "Espelkamp",
        "spouse": "Allemeyer Werner", "notes": "Nichte von Herbert",
    }
    p = persons_tree._parse_row(2, fields)
    assert p["rowId"] == "row_002"
    assert p["firstName"] == "Elsgard"
    assert p["lastName"] == "Allemeyer"
    assert p["maidenName"] == "Wöhler"
    assert p["birthYear"] == 1920
    assert p["deathYear"] == 1999
    assert p["birthPlace"] == "Garz"
    assert p["deathPlace"] == "Espelkamp"
    assert p["generation"] == 3
    assert p["familyMember"] is True
    assert p["_spouse_raw"] == "Allemeyer Werner"
    assert p["_bemerkung_raw"] == "Nichte von Herbert"
    assert "[Geburtsdatum" not in (p["notes"] or "")


def test_parse_row_string_birth_date():
    fields = {
        "generation": "G 2", "last_name": "Cram", "first_name": "Herbert",
        "maiden_name": "", "birth_date": "25.6.1890", "birth_place": "Texas",
        "death_date": "", "death_place": "", "spouse": "", "notes": "",
    }
    p = persons_tree._parse_row(28, fields)
    assert p["birthYear"] == 1890
    assert p["deathYear"] is None
    assert p["notes"] is None or p["notes"] == ""


def test_parse_row_unresolvable_date_goes_to_notes():
    fields = {
        "generation": "G 3", "last_name": "Heydrich", "first_name": "Dieter",
        "maiden_name": "", "birth_date": "28.9.", "birth_place": "",
        "death_date": "", "death_place": "", "spouse": "", "notes": "Bruder v Ingrid",
    }
    p = persons_tree._parse_row(96, fields)
    assert p["birthYear"] is None
    assert "[Geburtsdatum: 28.9.]" in p["notes"]
    assert "Bruder v Ingrid" in p["notes"]


def test_parse_row_empty_spouse_and_notes():
    fields = {
        "generation": "G 4", "last_name": "Allemeyer", "first_name": "Jürgen",
        "maiden_name": "", "birth_date": "", "birth_place": "",
        "death_date": "", "death_place": "", "spouse": "", "notes": "",
    }
    p = persons_tree._parse_row(4, fields)
    assert p["_spouse_raw"] is None
    assert p["_bemerkung_raw"] is None


def test_deduplicate_no_duplicates():
    persons = [
        {"rowId": "row_002", "firstName": "Elsgard", "lastName": "Allemeyer", "birthYear": 1920},
        {"rowId": "row_003", "firstName": "Werner", "lastName": "Allemeyer", "birthYear": 1923},
    ]
    result, skipped = persons_tree._deduplicate(persons)
    assert len(result) == 2
    assert skipped == []


def test_deduplicate_exact_match():
    # rows 127/138: same firstName, lastName, birthYear
    persons = [
        {"rowId": "row_127", "firstName": "Christa", "lastName": "Schütz", "birthYear": 1951},
        {"rowId": "row_138", "firstName": "Christa", "lastName": "Schütz", "birthYear": 1951},
    ]
    result, skipped = persons_tree._deduplicate(persons)
    assert [p["rowId"] for p in result] == ["row_127"]
    assert len(skipped) == 1
    assert "row_138" in skipped[0]


def test_deduplicate_none_birth_year_after_known():
    # rows 129/139: row 129 has birthYear=1964, row 139 has birthYear=None
    persons = [
        {"rowId": "row_129", "firstName": "Christoph", "lastName": "Seils", "birthYear": 1964},
        {"rowId": "row_139", "firstName": "Christoph", "lastName": "Seils", "birthYear": None},
    ]
    result, skipped = persons_tree._deduplicate(persons)
    assert [p["rowId"] for p in result] == ["row_129"]
    assert len(skipped) == 1


def test_deduplicate_both_none_birth_year_kept():
    # Two people with no birth year but same name: keep first only
    persons = [
        {"rowId": "row_A", "firstName": "Hans", "lastName": "Cram", "birthYear": None},
        {"rowId": "row_B", "firstName": "Hans", "lastName": "Cram", "birthYear": None},
    ]
    result, skipped = persons_tree._deduplicate(persons)
    assert [p["rowId"] for p in result] == ["row_A"]
    assert len(skipped) == 1


def _make_persons(*args):
    """Helper: args are (rowId, firstName, lastName, maidenName, spouse_raw) tuples."""
    return [
        {"rowId": a[0], "firstName": a[1], "lastName": a[2], "maidenName": a[3],
         "_spouse_raw": a[4], "_bemerkung_raw": None,
         "birthYear": None, "deathYear": None, "birthPlace": None, "deathPlace": None,
         "generation": None, "familyMember": True, "alias": None, "notes": None}
        for a in args
    ]


def test_resolve_spouses_success():
    persons = _make_persons(
        ("row_002", "Elsgard", "Allemeyer", "Wöhler", "Allemeyer Werner"),
        ("row_003", "Werner", "Allemeyer", None, "Elsgard Wöhler"),
    )
    idx = persons_tree._build_index(persons)
    rels, unres = persons_tree._resolve_spouses(persons, idx)
    assert len(rels) == 1
    assert rels[0]["type"] == "SPOUSE_OF"
    assert set([rels[0]["personId"], rels[0]["relatedPersonId"]]) == {"row_002", "row_003"}
    assert unres == []


def test_resolve_spouses_not_found():
    persons = _make_persons(
        ("row_007", "Charlotte", "Blomquist", "Ruge", '"Tante Lolly"'),
    )
    idx = persons_tree._build_index(persons)
    rels, unres = persons_tree._resolve_spouses(persons, idx)
    assert rels == []
    assert len(unres) == 1
    assert unres[0]["rowId"] == "row_007"
    assert unres[0]["reason"] == "not_found"


def test_resolve_spouses_empty_spouse_field():
    persons = _make_persons(
        ("row_004", "Jürgen", "Allemeyer", None, None),
    )
    idx = persons_tree._build_index(persons)
    rels, unres = persons_tree._resolve_spouses(persons, idx)
    assert rels == [] and unres == []


def _register(*args):
    """Build index from (rowId, first, last, maiden) tuples."""
    persons = [
        {"rowId": a[0], "firstName": a[1], "lastName": a[2], "maidenName": a[3]}
        for a in args
    ]
    return persons, persons_tree._build_index(persons)


def test_parse_bemerkung_sohn_two_parents():
    _, idx = _register(
        ("row_019", "Clara", "Cram", "de Gruyter"),
        ("row_028", "Herbert", "Cram", None),
    )
    rels, unres, notes = persons_tree._parse_bemerkung(
        "row_021", "Sohn v Clara Cram u Herbert Cram", idx
    )
    assert len(rels) == 2
    assert all(r["type"] == "PARENT_OF" for r in rels)
    child_ids = {r["relatedPersonId"] for r in rels}
    parent_ids = {r["personId"] for r in rels}
    assert child_ids == {"row_021"}
    assert "row_019" in parent_ids and "row_028" in parent_ids
    assert unres == []
    assert notes == ""


def test_parse_bemerkung_tochter_von():
    _, idx = _register(("row_019", "Clara", "Cram", None))
    rels, unres, notes = persons_tree._parse_bemerkung(
        "row_036", "Tochter von Clara Cram", idx
    )
    assert len(rels) == 1
    assert rels[0] == {
        "personId": "row_019",
        "relatedPersonId": "row_036",
        "type": "PARENT_OF",
        "source": "bemerkung",
        "rawBemerkung": "Tochter von Clara Cram",
    }
    assert notes == ""


def test_parse_bemerkung_vater():
    _, idx = _register(("row_028", "Herbert", "Cram", None))
    rels, unres, notes = persons_tree._parse_bemerkung(
        "row_031", "Vater v Herbert Cram", idx
    )
    assert len(rels) == 1
    assert rels[0]["personId"] == "row_031"
    assert rels[0]["relatedPersonId"] == "row_028"
    assert rels[0]["type"] == "PARENT_OF"


def test_parse_bemerkung_unmatched_parent_name():
    _, idx = _register()   # empty index
    rels, unres, notes = persons_tree._parse_bemerkung(
        "row_004", "Sohn v Elsgard A.", idx
    )
    assert rels == []
    assert len(unres) == 1
    assert unres[0]["reason"] == "not_found"
    assert notes == ""


def test_parse_bemerkung_skip_nichte():
    _, idx = _register(("row_028", "Herbert", "Cram", None))
    rels, unres, notes = persons_tree._parse_bemerkung(
        "row_002", "Nichte von Herbert", idx
    )
    assert rels == []
    assert unres == []
    assert notes == "Nichte von Herbert"


def test_parse_bemerkung_skip_bruder():
    _, idx = _register(("row_028", "Herbert", "Cram", None))
    rels, unres, notes = persons_tree._parse_bemerkung(
        "row_033", "Bruder v Herbert", idx
    )
    assert rels == []
    assert unres == []
    assert notes == "Bruder v Herbert"


def test_parse_bemerkung_empty():
    _, idx = _register()
    rels, unres, notes = persons_tree._parse_bemerkung("row_004", "", idx)
    assert rels == [] and unres == [] and notes == ""


def test_parse_bemerkung_plain_remark():
    _, idx = _register()
    rels, unres, notes = persons_tree._parse_bemerkung(
        "row_029", "Verfasserin der Cram-Chronik !!", idx
    )
    assert rels == [] and unres == []
    assert notes == "Verfasserin der Cram-Chronik !!"


def test_parse_bemerkung_sohn_with_trailing_remark():
    _, idx = _register(
        ("row_019", "Clara", "Cram", "de Gruyter"),
        ("row_028", "Herbert", "Cram", None),
    )
    rels, unres, notes = persons_tree._parse_bemerkung(
        "row_021", "Sohn v Clara Cram u Herbert Cram, nach Mexiko emigriert", idx
    )
    assert len(rels) == 2
    assert unres == []
    assert notes == "nach Mexiko emigriert"


import subprocess


def test_dry_run_exits_zero(tmp_path):
    """dry-run should complete without writing any file and exit 0."""
    input_path = Path(__file__).parent.parent.parent.parent / "import" / "Personendatei 2.xlsx"
    if not input_path.exists():
        import pytest
        pytest.skip("source Excel file not present")

    result = subprocess.run(
        [
            sys.executable, str(Path(__file__).parent.parent / "persons_tree.py"),
            "--input", str(input_path),
            "--output", str(tmp_path / "out.json"),
            "--dry-run",
        ],
        capture_output=True, text=True,
    )
    assert result.returncode == 0, result.stderr
    assert not (tmp_path / "out.json").exists()
    assert "persons parsed" in result.stdout