youtube_abschrift_korrigieren/tests/test_vtt2txt.py

"""Tests for vtt2txt.py."""
import shutil
import sys
from pathlib import Path

import pytest

import vtt2txt

REPO = Path(__file__).resolve().parent.parent
SAMPLE_VTT = REPO / "testdaten" / "ramstein.de.vtt"
SAMPLE_TXT = REPO / "testdaten" / "ramstein.de.txt"


@pytest.fixture
def sample_copy(tmp_path):
    target = tmp_path / SAMPLE_VTT.name
    shutil.copy(SAMPLE_VTT, target)
    return target


def cue(start, *lines):
    return f"00:00:{start:02d}.000 --> 00:00:{start + 1:02d}.000\n" + "\n".join(lines) + "\n\n"


def run_main(monkeypatch, *args):
    monkeypatch.setattr(sys, "argv", ["vtt2txt.py", *map(str, args)])
    vtt2txt.main()


def test_vtt_to_lines_strips_tags_and_rolling_duplicates():
    vtt = (
        "WEBVTT\nKind: captions\n\n"
        "00:00:01.000 --> 00:00:02.000 align:start\n"
        "Erste<00:00:01.500><c> Zeile</c>\n\n"
        "00:00:02.000 --> 00:00:03.000\nErste Zeile\n \n\n"
        "cue-id\n00:00:03.000 --> 00:00:04.000\nErste Zeile\nA &amp; B\n"
    )
    assert vtt2txt.vtt_to_lines(vtt) == ["Erste Zeile", "A & B"]


def test_vtt_to_lines_keeps_text_after_space_only_line():
    # YouTube cues start with a line holding a single space; the text after it
    # must survive even if no later cue repeats it (here: end of file)
    vtt = "WEBVTT\n\n00:00:01.000 --> 00:00:02.000\n \n[applaus]\n"
    assert vtt2txt.vtt_to_lines(vtt) == ["[applaus]"]


def test_vtt_to_lines_skips_note_and_normalizes_whitespace():
    vtt = (
        "WEBVTT\n\nNOTE Kommentar\nnicht übernehmen\n\n"
        "00:00:01.000 --> 00:00:02.000\nA&nbsp; B\n\n"
        "00:00:02.000 --> 00:00:03.000\nA B\nC\n"
    )
    assert vtt2txt.vtt_to_lines(vtt) == ["A B", "C"]


def test_vtt_to_lines_keeps_real_repetition():
    # Rolling window: each cue repeats the previous line, then adds new text.
    # The speaker really says "Nein." twice; both must survive.
    vtt = "WEBVTT\n\n" + "".join([
        cue(1, " ", "Er sagt:"),
        cue(2, "Er sagt:", " "),
        cue(3, "Er sagt:", "Nein."),
        cue(4, "Nein.", " "),
        cue(5, "Nein.", "Nein."),
        cue(6, " ", " "),
        cue(7, "Nein.", "Ende."),
    ])
    assert vtt2txt.vtt_to_lines(vtt) == ["Er sagt:", "Nein.", "Nein.", "Ende."]


def test_vtt_to_lines_without_rolling_window():
    vtt = "WEBVTT\n\n1\n" + cue(1, "A", "B") + "2\n" + cue(2, "C")
    assert vtt2txt.vtt_to_lines(vtt) == ["A", "B", "C"]


def test_main_default_output_matches_sample(sample_copy, monkeypatch, capsys):
    run_main(monkeypatch, "-i", sample_copy)
    target = sample_copy.with_name("ramstein.de.txt")
    assert target.read_text(encoding="utf-8") == SAMPLE_TXT.read_text(encoding="utf-8")
    assert f"Erstellt: {target}" in capsys.readouterr().out


def test_main_multiple_inputs(sample_copy, monkeypatch):
    second = sample_copy.with_name("rede.en.vtt")
    shutil.copy(sample_copy, second)
    run_main(monkeypatch, "-i", sample_copy, second)
    assert sample_copy.with_suffix(".txt").is_file()
    assert second.with_suffix(".txt").is_file()


def test_main_expands_glob_without_duplicates(sample_copy, monkeypatch, capsys):
    shutil.copy(sample_copy, sample_copy.with_name("ramstein.en.vtt"))
    shutil.copy(sample_copy, sample_copy.with_name("anders.de.vtt"))
    run_main(monkeypatch, "-i", sample_copy.parent / "ramstein.*.vtt", sample_copy)
    out = capsys.readouterr().out
    assert out.count("Erstellt:") == 2
    assert not sample_copy.with_name("anders.de.txt").exists()


@pytest.mark.parametrize("make_dir, suffix", [(True, ""), (False, "/")])
def test_main_output_directory(sample_copy, monkeypatch, make_dir, suffix):
    second = sample_copy.with_name("rede.en.vtt")
    shutil.copy(sample_copy, second)
    out_dir = sample_copy.parent / "out"
    if make_dir:
        out_dir.mkdir()
    run_main(monkeypatch, "-i", sample_copy, second, "-o", f"{out_dir}{suffix}")
    assert (out_dir / "ramstein.de.txt").is_file()
    assert (out_dir / "rede.en.txt").is_file()


def test_main_output_option_creates_directories(sample_copy, monkeypatch):
    target = sample_copy.parent / "neu" / "unter" / "rede.txt"
    run_main(monkeypatch, "-i", sample_copy, "-o", target)
    assert target.is_file()


@pytest.mark.parametrize("content, message", [
    ("WEBVTT\n".encode("utf-8"), "Keine Untertitel"),
    ("WEBVTT\n\n00:00:01.000 --> 00:00:02.000\nGrüße\n".encode("latin-1"),
     "Keine UTF-8-Datei"),
])
def test_main_bad_file_continues_batch(sample_copy, monkeypatch, capsys, content, message):
    bad = sample_copy.with_name("aaa.vtt")
    bad.write_bytes(content)
    with pytest.raises(SystemExit) as exc:
        run_main(monkeypatch, "-i", bad, sample_copy)
    assert exc.value.code == 1
    assert message in capsys.readouterr().err
    assert not bad.with_suffix(".txt").exists()
    assert sample_copy.with_suffix(".txt").is_file()


@pytest.mark.parametrize("args, message", [
    ([], "required"),
    (["-i", "gibtsnicht.vtt"], "Eingabedatei nicht gefunden"),
    (["-i", "{txt}"], "Keine VTT-Datei"),
    (["-i", "{vtt}", "{vtt2}", "-o", "{dir}/x.txt"], "muss -o/--output ein Verzeichnis sein"),
    (["-i", "{dir}/*.gibtsnicht"], "Keine Datei passt zu"),
    (["-i", "{vtt}", "{dir}/sub/ramstein.de.vtt", "-o", "{dir}/"], "dieselbe Ausgabedatei"),
    (["-i", "{vtt}", "-o", "{vtt}"], "nicht die Eingabedatei"),
])
def test_cli_errors(sample_copy, monkeypatch, capsys, args, message):
    txt = sample_copy.with_suffix(".txt")
    txt.write_text("x\n", encoding="utf-8")
    vtt2 = sample_copy.with_name("rede.en.vtt")
    shutil.copy(sample_copy, vtt2)
    (sample_copy.parent / "sub").mkdir()
    shutil.copy(sample_copy, sample_copy.parent / "sub" / sample_copy.name)
    args = [a.format(vtt=sample_copy, vtt2=vtt2, txt=txt, dir=sample_copy.parent)
            for a in args]
    with pytest.raises(SystemExit) as exc:
        run_main(monkeypatch, *args)
    assert exc.value.code == 2
    assert message in capsys.readouterr().err