import threading import time import zipfile from pathlib import Path import pytest from PIL import Image from PySide6.QtCore import QSettings from mdhelper.batch_documents import BatchDocument, combine_documents, fingerprint, publish_outputs from mdhelper.batch_export_dialog import BatchExportDialog from mdhelper.docx_exporter import export_markdown_to_docx from mdhelper.export_jobs import ExportCancelled from mdhelper.export_settings import PageSettings from mdhelper.markdown_renderer import MarkdownRenderer from mdhelper import main_window, batch_documents def documents(tmp_path): first, second = tmp_path / 'first.md', tmp_path / 'sub' / 'second.md' second.parent.mkdir() Image.new('RGB', (16, 16), 'red').save(second.parent / '图.png') first.write_text('# Shared\n\n[chapter](sub/second.md)\n\n[other](sub/second.md#shared)\n\n[ref][same]\n\n[same]: sub/second.md#shared', encoding='utf-8') second.write_text('# Shared\n\n![图](图.png)\n\n[ref][same]\n\n[same]: ../first.md#shared\n\n```\n[code](../first.md)\n```', encoding='utf-8') return [BatchDocument(path, path.read_text(encoding='utf-8')) for path in (first, second)] def test_combined_resources_duplicate_headings_and_real_docx(tmp_path): docs = documents(tmp_path) combined = combine_documents(docs, MarkdownRenderer(tmp_path)) assert [h.slug for h in combined.headings] == ['mdh1', 'mdh1h1', 'mdh2', 'mdh2h1'] assert 'href="#mdh2h1"' in combined.html_body and 'href="#mdh1h1"' in combined.html_body assert 'id="mdh2h1"' in combined.html_body and '(#mdh2)' in combined.word_source assert '[code](../first.md)' in combined.word_source output = tmp_path / 'combined.docx' export_markdown_to_docx(combined.word_source, output, engine='pandoc', source_path=docs[0].path, page_settings=PageSettings(header='Batch header', footer='Batch footer')) with zipfile.ZipFile(output) as archive: xml = archive.read('word/document.xml').decode() assert 'w:anchor="mdh2h1"' in xml and 'w:name="mdh2h1"' in xml assert any(name.startswith('word/media/') for name in archive.namelist()) assert 'TOC ' in xml assert 'Batch header' in archive.read('word/header1.xml').decode() def test_publication_cancellation_race_and_rollback(tmp_path, monkeypatch): destinations = [tmp_path / f'{i}.out' for i in range(2)] staged = [tmp_path / f'{i}.stage' for i in range(2)] for path in destinations: path.write_bytes(b'old') for path in staged: path.write_bytes(b'new') outputs = dict(zip(destinations, staged)); expected = {p: fingerprint(p) for p in destinations} cancel = threading.Event(); cancel.set() with pytest.raises(ExportCancelled): publish_outputs(outputs, expected, tmp_path, cancel) destinations[0].write_bytes(b'external') with pytest.raises(OSError, match='发生变化'): publish_outputs(outputs, expected, tmp_path) destinations[0].write_bytes(b'old') real_write = batch_documents.atomic_write def fail(path, data): if path == destinations[1] and data == b'new': raise OSError('disk full') real_write(path, data) monkeypatch.setattr(batch_documents, 'atomic_write', fail) with pytest.raises(OSError, match='disk full'): publish_outputs(outputs, expected, tmp_path) assert all(path.read_bytes() == b'old' for path in destinations) @pytest.mark.parametrize('kind,combined', [('docx', False), ('pdf', True), ('pdf', False)]) def test_queue_generates_complete_outputs(qt_application, tmp_path, monkeypatch, kind, combined): docs = documents(tmp_path) settings = QSettings(str(tmp_path / 'settings.ini'), QSettings.Format.IniFormat) monkeypatch.setattr(main_window, 'QSettings', lambda *_: settings) errors = [] monkeypatch.setattr(main_window.QMessageBox, 'warning', lambda *args: errors.append(str(args[2]))) window = main_window.MainWindow(tmp_path) panel = BatchExportDialog(window) try: window.open_document(docs[0].path) output = [tmp_path / f'export-{i}.{kind}' for i in range(1 if combined else 2)] panel.begin(docs, output, kind, combined, window._export_snapshot(), PageSettings(header='Batch', include_toc=True)) deadline = time.monotonic() + 60 while panel.running and time.monotonic() < deadline: qt_application.processEvents(); time.sleep(.02) assert not panel.running, panel.status.text() assert not errors and all(path.exists() for path in output), (panel.status.text(), errors) if kind == 'pdf': from pypdf import PdfReader text = '\n'.join(page.extract_text() for path in output for page in PdfReader(path).pages) assert 'Shared' in text and 'Batch' in text assert sum(len(page.images) for path in output for page in PdfReader(path).pages) >= 1 finally: panel.cancel(); window.close()