5 files changed, 244 insertions, 0 deletions
diff --git a/tests/__init__.py b/tests/__init__.py
new file mode 100644
index 0000000..e69de29
--- /dev/null
+++ b/tests/__init__.py
diff --git a/tests/test_audio.py b/tests/test_audio.py
new file mode 100644
index 0000000..a629224
--- /dev/null
+++ b/tests/test_audio.py
@@ -0,0 +1,57 @@
+"""Tests for audio helpers: speed parameters and chunk cleanup."""
+
+import tempfile
+import unittest
+from pathlib import Path
+
+from converter import config
+from converter.audio import cleanup_chunks, speed_export_params
+
+
+class SpeedExportParamsTests(unittest.TestCase):
+    def test_normal_speed_no_filter(self):
+        self.assertEqual(speed_export_params(1.0), [])
+
+    def test_simple_speedup(self):
+        self.assertEqual(speed_export_params(1.5), ["-filter:a", "atempo=1.5"])
+
+    def test_simple_slowdown(self):
+        self.assertEqual(speed_export_params(0.75), ["-filter:a", "atempo=0.75"])
+
+    def test_chained_speedup_beyond_2x(self):
+        self.assertEqual(speed_export_params(3.0), ["-filter:a", "atempo=2.0,atempo=1.5"])
+
+    def test_chained_slowdown_below_half(self):
+        self.assertEqual(speed_export_params(0.25), ["-filter:a", "atempo=0.5,atempo=0.5"])
+
+    def test_zero_speed_rejected(self):
+        with self.assertRaises(ValueError):
+            speed_export_params(0)
+
+    def test_negative_speed_rejected(self):
+        with self.assertRaises(ValueError):
+            speed_export_params(-1.5)
+
+
+class CleanupChunksTests(unittest.TestCase):
+    def test_removes_only_chunk_files(self):
+        with tempfile.TemporaryDirectory() as tmp:
+            chunks_dir = Path(tmp)
+            (chunks_dir / "chunk_0001.wav").write_bytes(b"stale")
+            (chunks_dir / "chunk_0002.wav").write_bytes(b"stale")
+            (chunks_dir / "keep.txt").write_bytes(b"keep")
+
+            original = config.CHUNKS_FOLDER
+            config.CHUNKS_FOLDER = chunks_dir
+            try:
+                cleanup_chunks()
+            finally:
+                config.CHUNKS_FOLDER = original
+
+            self.assertFalse((chunks_dir / "chunk_0001.wav").exists())
+            self.assertFalse((chunks_dir / "chunk_0002.wav").exists())
+            self.assertTrue((chunks_dir / "keep.txt").exists())
+
+
+if __name__ == "__main__":
+    unittest.main()
diff --git a/tests/test_chunking.py b/tests/test_chunking.py
new file mode 100644
index 0000000..da45f65
--- /dev/null
+++ b/tests/test_chunking.py
@@ -0,0 +1,39 @@
+"""Tests for text chunking."""
+
+import unittest
+
+from converter.chunking import split_into_chunks
+
+
+class SplitIntoChunksTests(unittest.TestCase):
+    def test_empty_input(self):
+        self.assertEqual(split_into_chunks(""), [])
+        self.assertEqual(split_into_chunks("   \n  "), [])
+
+    def test_short_text_single_chunk(self):
+        self.assertEqual(split_into_chunks("One short sentence."), ["One short sentence."])
+
+    def test_respects_word_limit_across_sentences(self):
+        # 10 sentences of 9 words each = 90 words total
+        sentences = [f"S{i} " + " ".join(["word"] * 8) + "." for i in range(10)]
+        chunks = split_into_chunks(" ".join(sentences), max_words=25)
+        self.assertGreater(len(chunks), 1)
+        self.assertTrue(all(len(c.split()) <= 25 for c in chunks))
+        self.assertEqual(sum(len(c.split()) for c in chunks), 90)
+
+    def test_long_sentence_split_keeps_punctuation(self):
+        # 10 clauses of 5 words each, joined by commas
+        sentence = ",".join([" ".join(["w"] * 5) for _ in range(10)]) + "."
+        chunks = split_into_chunks(sentence, max_words=12)
+        self.assertGreater(len(chunks), 1)
+        self.assertTrue(all(len(c.split()) <= 12 for c in chunks))
+        self.assertIn(",", chunks[0])  # commas retained for TTS prosody
+
+    def test_single_oversized_sentence_stays_intact(self):
+        sentence = " ".join(["word"] * 30) + "."
+        chunks = split_into_chunks(sentence, max_words=10)
+        self.assertEqual(chunks, [sentence])
+
+
+if __name__ == "__main__":
+    unittest.main()
diff --git a/tests/test_cleaning.py b/tests/test_cleaning.py
new file mode 100644
index 0000000..41f4ed7
--- /dev/null
+++ b/tests/test_cleaning.py
@@ -0,0 +1,54 @@
+"""Tests for text and HTML cleaning."""
+
+import unittest
+
+from converter.extractors import clean_html, clean_text
+
+
+class CleanTextTests(unittest.TestCase):
+    def test_empty_input(self):
+        self.assertEqual(clean_text(""), "")
+        self.assertEqual(clean_text(None), "")
+
+    def test_collapses_whitespace(self):
+        self.assertEqual(clean_text("a\n\n   b \t c"), "a b c")
+
+    def test_preserves_inline_numbers(self):
+        self.assertEqual(clean_text("He was 42 years old."), "He was 42 years old.")
+
+    def test_preserves_grouped_and_decimal_numbers(self):
+        self.assertEqual(
+            clean_text("Over 1,000 pages and 3.5 stars."),
+            "Over 1,000 pages and 3.5 stars.",
+        )
+
+    def test_removes_standalone_page_numbers(self):
+        self.assertEqual(
+            clean_text("End of page.\n7\nNext page text."),
+            "End of page. Next page text.",
+        )
+
+    def test_page_number_removal_leaves_single_spacing(self):
+        result = clean_text("Chapter one\n\n12\n\nChapter two")
+        self.assertEqual(result, "Chapter one Chapter two")
+        self.assertNotIn("  ", result)
+
+
+class CleanHtmlTests(unittest.TestCase):
+    def test_strips_tags(self):
+        self.assertEqual(clean_html("<p>Hello <b>world</b></p>"), "Hello world")
+
+    def test_removes_script_and_style(self):
+        html = "<style>.x{color:red}</style><p>Text</p><script>var a=1;</script>"
+        self.assertEqual(clean_html(html), "Text")
+
+    def test_unescapes_entities(self):
+        self.assertEqual(clean_html("Tom &amp; Jerry"), "Tom & Jerry")
+
+    def test_empty(self):
+        self.assertEqual(clean_html(""), "")
+        self.assertEqual(clean_html(None), "")
+
+
+if __name__ == "__main__":
+    unittest.main()
diff --git a/tests/test_extractors.py b/tests/test_extractors.py
new file mode 100644
index 0000000..c497267
--- /dev/null
+++ b/tests/test_extractors.py
@@ -0,0 +1,94 @@
+"""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) -> None:
+    from ebooklib import epub
+
+    book = epub.EpubBook()
+    book.set_identifier("test-id")
+    book.set_title("Test Book")
+    book.set_language("en")
+
+    chapter1 = epub.EpubHtml(title="One", file_name="chap1.xhtml", lang="en")
+    chapter1.content = "<html><body><p>First chapter text.</p></body></html>"
+    chapter2 = epub.EpubHtml(title="Two", file_name="chap2.xhtml", lang="en")
+    chapter2.content = "<html><body><p>Second chapter text.</p></body></html>"
+
+    book.add_item(chapter1)
+    book.add_item(chapter2)
+    book.toc = (chapter1, chapter2)
+    book.spine = ["nav", chapter1, chapter2]
+    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 _extract_epub_ebooklib
+
+        with tempfile.TemporaryDirectory() as tmp:
+            path = Path(tmp) / "book.epub"
+            _build_test_epub(path)
+            text = _extract_epub_ebooklib(path)
+
+        self.assertIn("First chapter text.", text)
+        self.assertIn("Second chapter text.", text)
+
+    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."))
+
+
+if __name__ == "__main__":
+    unittest.main()