"""Tests for file text extraction.""" import tempfile import unittest from pathlib import Path from converter.extractors import extract_text class TxtExtractionTests(unittest.TestCase): def _extract(self, data: bytes) -> str: with tempfile.TemporaryDirectory() as tmp: path = Path(tmp) / "book.txt" path.write_bytes(data) return extract_text(path) def test_utf8(self): self.assertEqual(self._extract("héllo wörld".encode("utf-8")), "héllo wörld") def test_utf16_with_bom(self): self.assertEqual(self._extract("héllo".encode("utf-16")), "héllo") def test_cp1252(self): self.assertEqual(self._extract("“quotes”".encode("cp1252")), "“quotes”") def test_latin1_fallback(self): # 0x81 is undefined in cp1252, forcing the latin-1 catch-all self.assertEqual(self._extract(b"caf\x81"), "caf\x81") def test_unsupported_format(self): with tempfile.TemporaryDirectory() as tmp: path = Path(tmp) / "book.xyz" path.write_bytes(b"data") with self.assertRaises(ValueError): extract_text(path) def _build_test_epub(path: Path, chapters=(("One", "First chapter text."), ("Two", "Second chapter text."))) -> None: from ebooklib import epub book = epub.EpubBook() book.set_identifier("test-id") book.set_title("Test Book") book.set_language("en") book.add_author("Test Author") items = [] for index, (title, text) in enumerate(chapters, 1): chapter = epub.EpubHtml(title=title, file_name=f"chap{index}.xhtml", lang="en") chapter.content = f"

{text}

" book.add_item(chapter) items.append(chapter) book.toc = tuple(items) book.spine = ["nav", *items] book.add_item(epub.EpubNcx()) book.add_item(epub.EpubNav()) epub.write_epub(str(path), book) class EpubExtractionTests(unittest.TestCase): def setUp(self): try: import ebooklib # noqa: F401 except ImportError: self.skipTest("ebooklib not installed") def test_ebooklib_extraction(self): # Regression test: the ebooklib path used to silently return "" due to # isinstance(item, ebooklib.ITEM_DOCUMENT) (an int, not a class). from converter.extractors import _read_epub_ebooklib with tempfile.TemporaryDirectory() as tmp: path = Path(tmp) / "book.epub" _build_test_epub(path) items = _read_epub_ebooklib(path) html = "\n".join(content for _, content in items) self.assertIn("First chapter text.", html) self.assertIn("Second chapter text.", html) def test_epub_extraction_follows_spine_order(self): with tempfile.TemporaryDirectory() as tmp: path = Path(tmp) / "book.epub" _build_test_epub(path) text = extract_text(path) self.assertIn("First chapter text.", text) self.assertIn("Second chapter text.", text) self.assertLess(text.index("First chapter text."), text.index("Second chapter text.")) class ExtractSectionsTests(unittest.TestCase): def setUp(self): try: import ebooklib # noqa: F401 except ImportError: self.skipTest("ebooklib not installed") def test_epub_sections_split_on_chapters(self): from converter.extractors import extract_sections with tempfile.TemporaryDirectory() as tmp: path = Path(tmp) / "book.epub" _build_test_epub(path) sections = extract_sections(path) self.assertEqual(len(sections), 2) self.assertEqual(sections[0].title, "One") self.assertEqual(sections[1].title, "Two") self.assertIn("First chapter text.", sections[0].text) self.assertIn("Second chapter text.", sections[1].text) def test_txt_is_single_section(self): from converter.extractors import extract_sections with tempfile.TemporaryDirectory() as tmp: path = Path(tmp) / "book.txt" path.write_text("Hello world.", encoding="utf-8") sections = extract_sections(path) self.assertEqual(len(sections), 1) self.assertEqual(sections[0].title, "book") self.assertEqual(sections[0].text, "Hello world.") def test_single_chapter_epub_keeps_chapter_title(self): from converter.extractors import extract_sections with tempfile.TemporaryDirectory() as tmp: path = Path(tmp) / "book.epub" _build_test_epub(path, chapters=(("Only", "Just one chapter."),)) sections = extract_sections(path) self.assertEqual(len(sections), 1) self.assertEqual(sections[0].title, "Only") self.assertIn("Just one chapter.", sections[0].text) class ExtractBookTests(unittest.TestCase): def test_txt_falls_back_to_stem_and_blank_author(self): from converter.extractors import extract_book with tempfile.TemporaryDirectory() as tmp: path = Path(tmp) / "mybook.txt" path.write_text("Hello world.", encoding="utf-8") book = extract_book(path) self.assertEqual(book.title, "mybook") self.assertEqual(book.author, "") self.assertEqual(len(book.sections), 1) def test_epub_metadata_harvested(self): try: import ebooklib # noqa: F401 except ImportError: self.skipTest("ebooklib not installed") from converter.extractors import extract_book with tempfile.TemporaryDirectory() as tmp: path = Path(tmp) / "book.epub" _build_test_epub(path) book = extract_book(path) self.assertEqual(book.title, "Test Book") self.assertEqual(book.author, "Test Author") self.assertEqual([s.title for s in book.sections], ["One", "Two"]) def test_pdf_metadata_harvested(self): from converter.extractors import extract_book try: from pypdf import PdfWriter except ImportError: self.skipTest("pypdf not installed") with tempfile.TemporaryDirectory() as tmp: path = Path(tmp) / "book.pdf" writer = PdfWriter() writer.add_metadata({"/Title": "PDF Title", "/Author": "PDF Author"}) writer.add_blank_page(width=612, height=792) with open(path, "wb") as handle: writer.write(handle) book = extract_book(path) self.assertEqual(book.title, "PDF Title") self.assertEqual(book.author, "PDF Author") self.assertEqual(len(book.sections), 1) def test_pdf_without_metadata_falls_back(self): from converter.extractors import extract_book try: from pypdf import PdfWriter except ImportError: self.skipTest("pypdf not installed") with tempfile.TemporaryDirectory() as tmp: path = Path(tmp) / "plain.pdf" writer = PdfWriter() writer.add_blank_page(width=612, height=792) with open(path, "wb") as handle: writer.write(handle) book = extract_book(path) self.assertEqual(book.title, "plain") self.assertEqual(book.author, "") if __name__ == "__main__": unittest.main()