epub2md/tests/test_integration.py

import json
import re
from pathlib import Path

import pytest
import yaml

from conftest import ROOT, Book, Doc, build_epub

BEISPIELE = sorted((ROOT / "Beispiele").glob("*.epub"))


def simple_book() -> Book:
    return Book(docs=[Doc("a.xhtml", "<h1>A</h1><p>Text</p>"), Doc("b.xhtml", "<h1>B</h1><p>Mehr</p>")])


def test_output_structure_and_reports(convert):
    run = convert(simple_book(), "--copy-original")
    assert run.rc == 0
    for rel in ("README.md", "metadata.yaml", "manifest.json", "conversion-report.json",
                "diagnostics/warnings.md", "diagnostics/unsupported-elements.md"):
        assert (run.out / rel).is_file(), rel
    assert len(list((run.out / "source").glob("*.epub"))) == 1
    manifest = json.loads((run.out / "manifest.json").read_text(encoding="utf-8"))
    assert [c["file"] for c in manifest["chapters"]] == ["chapters/001-a.md", "chapters/002-b.md"]
    for c in manifest["chapters"]:
        assert (run.out / c["file"]).is_file()
        assert c["word_count"] > 0
    assert run.report["status"] == "success"
    assert run.report["input"]["sha256"]
    fm = yaml.safe_load(run.chapter("001").split("---\n")[1])
    assert fm["chapter_number"] == 1 and fm["source_href"] == "a.xhtml" and fm["book_title"] == "Testbuch"


def test_no_frontmatter_no_toc(convert):
    run = convert(simple_book(), "--no-frontmatter", "--no-toc")
    assert not run.chapter("001").startswith("---")
    assert not (run.out / "README.md").exists()


def test_dry_run_writes_nothing(convert, tmp_path):
    report = tmp_path / "r.json"
    run = convert(simple_book(), "--dry-run", "--report", str(report))
    assert run.rc == 0
    assert not run.out.exists()
    assert not report.exists()
    assert "chapters/001-a.md" in run.stdout


def test_existing_output_requires_force(convert, tmp_path):
    out = tmp_path / "ziel"
    out.mkdir()
    (out / "wichtig.txt").write_text("nicht löschen")
    run = convert(simple_book(), out=out)
    assert run.rc == 1
    assert (out / "wichtig.txt").exists()
    # --force refuses directories that are not epub2md output
    assert convert(simple_book(), "--force", out=out).rc == 1
    assert (out / "wichtig.txt").exists()


def test_force_replaces_previous_output(convert, tmp_path):
    out = tmp_path / "ziel"
    assert convert(simple_book(), out=out).rc == 0
    (out / "alt.txt").write_text("alt")
    run = convert(simple_book(), "--force", out=out)
    assert run.rc == 0
    assert not (out / "alt.txt").exists()
    assert (out / "README.md").exists()
    assert not list(tmp_path.glob(".epub2md-*"))  # temp dir cleaned up


def test_report_option_and_verbose(convert, tmp_path):
    report = tmp_path / "rep" / "bericht.json"
    run = convert(simple_book(), "--report", str(report), "--verbose")
    assert json.loads(report.read_text(encoding="utf-8"))["tool"] == "epub2md"
    assert "Rendere" in run.stderr


def test_invalid_arguments_exit_1(convert, tmp_path):
    assert convert(simple_book(), "--math", "falsch").rc == 1
    assert convert(simple_book(), "--chapter-level", "9").rc == 1
    assert convert(tmp_path / "gibt-es-nicht.epub").rc == 1


def test_deterministic_output(convert, tmp_path):
    epub = build_epub(tmp_path / "same.epub", simple_book())
    a = convert(epub, out=tmp_path / "x1")
    b = convert(epub, out=tmp_path / "x2")
    for p in a.chapters():
        assert p.read_text() == (b.out / "chapters" / p.name).read_text()


def test_encoding_edge_cases(convert):
    text_in = "Ärger über Öl – „Zitat“ — café naïve, Ωμέγα αβγ, ∀x∈ℝ: x² ≥ 0 … ‚einfach‘"
    book = Book(docs=[Doc("Kapitel-é.xhtml", f'<h1 id="über">Grüße</h1><p>{text_in}</p>')])
    run = convert(book)
    text = run.chapter("001")
    assert text_in in text
    assert "<EFBFBD>" not in text
    assert run.chapters()[0].name == "001-gruesse.md"


@pytest.mark.skipif(not BEISPIELE, reason="keine Beispiel-EPUBs vorhanden")
@pytest.mark.parametrize("epub", BEISPIELE, ids=[p.stem[:20] for p in BEISPIELE])
def test_example_books_structural_invariants(convert, epub: Path):
    """Checks structure only – no copyrighted text is asserted."""
    run = convert(epub)
    assert run.rc == 0, run.stderr
    report = run.report
    chapters = run.chapters()
    assert len(chapters) >= 3
    assert report["output_link_check"]["broken_count"] == 0, report["output_link_check"]["broken"][:5]
    assert not report["errors"]
    for path in chapters:
        text = path.read_text(encoding="utf-8")
        body = text.split("---\n", 2)[2] if text.startswith("---\n") else text
        yaml.safe_load(text.split("---\n")[1])
        assert body.strip(), f"leere Kapiteldatei {path.name}"
        fences = re.findall(r"^(`{3,})", body, re.M)
        assert len(fences) % 2 == 0, f"unbalancierte Code-Fences in {path.name}"
        for block in re.findall(r"^```[\w+-]*\n(.*?)\n```$", body, re.M | re.S):
            assert "&lt;" not in block and "&amp;" not in block
    for m in report["media"]["files"]:
        assert (run.out / m["file"]).is_file()
    assert sum(c["stats"]["word_count"] for c in report["chapters"]) > 1000