feat(reader): add PDF reading support with outline chapters, CLI blocks, and noise filtering (task 7)

tests/test_api.py

@@ -891,3 +891,123 @@ def test_duplicate_prevention_and_in_library_flag():
     data_new = resp_lookup_new.json()
     assert data_new["in_library"] is False
     assert data_new["library_card"] is None
+
+
+def make_mock_pdf(title="CCNA 200-301 Official Guide", author="Wendell Odom", with_outline=True):
+    import io
+    import pypdf
+    from pypdf.generic import DecodedStreamObject, NameObject, DictionaryObject
+
+    writer = pypdf.PdfWriter()
+
+    def add_text_page(text_lines):
+        page = writer.add_blank_page(width=400, height=400)
+        stream_content = [b"BT /F1 12 Tf 50 350 Td"]
+        for idx, l in enumerate(text_lines):
+            if idx > 0:
+                stream_content.append(b"0 -15 Td")
+            escaped = l.replace("\\", "\\\\").replace("(", "\\(").replace(")", "\\)")
+            stream_content.append(f"({escaped}) Tj".encode("latin1"))
+        stream_content.append(b"ET")
+
+        content = DecodedStreamObject()
+        content.set_data(b" ".join(stream_content))
+        page[NameObject("/Contents")] = content
+        page[NameObject("/Resources")] = DictionaryObject({
+            NameObject("/Font"): DictionaryObject({
+                NameObject("/F1"): DictionaryObject({
+                    NameObject("/Type"): NameObject("/Font"),
+                    NameObject("/Subtype"): NameObject("/Type1"),
+                    NameObject("/BaseFont"): NameObject("/Helvetica"),
+                })
+            })
+        })
+
+    p0_lines = ["Introduction to Networking", "CCNA Official Guide Header"] + [f"Introductory networking line {i} with several words for threshold." for i in range(12)]
+    p1_lines = ["Chapter 1: Ethernet Switching", "Router# show ip interface brief", "GigabitEthernet0/0 10.0.0.1 up up", "Router# configure terminal", "Switch(config)# vlan 10"] + [f"Ethernet switching explanation line {i} with concepts." for i in range(12)]
+    p2_lines = ["Chapter 2: IPv4 Routing", "Router# show ip route", "C 10.0.0.0/24 is directly connected"] + [f"IPv4 routing explanation line {i} with concepts." for i in range(12)]
+
+    add_text_page(p0_lines)
+    add_text_page(p1_lines)
+    add_text_page(p2_lines)
+
+    if with_outline:
+        writer.add_outline_item("Introduction", 0)
+        writer.add_outline_item("Chapter 1: Ethernet Switching", 1)
+        writer.add_outline_item("Chapter 2: IPv4 Routing", 2)
+
+    buf = io.BytesIO()
+    writer.write(buf)
+    return buf.getvalue()
+
+
+def test_pdf_reader_unit():
+    import io
+    import pypdf
+    from app.reader import parse_pdf, format_cli_blocks, clean_page_lines
+
+    # 1. Test CLI block detection
+    sample_cli = "Check interfaces:\nRouter# show ip interface brief\nGigabitEthernet0/0 10.0.0.1 up up\nRouter# exit\nAll done."
+    formatted = format_cli_blocks(sample_cli)
+    assert "```text" in formatted
+    assert "Router# show ip interface brief" in formatted
+    assert "```" in formatted
+
+    # 2. Test outline-based PDF parsing
+    pdf_bytes = make_mock_pdf("CCNA Guide", "Cisco Press", with_outline=True)
+    title, author, chapters = parse_pdf(pdf_bytes, filename="ccna_guide.pdf")
+    assert title == "ccna_guide"
+    assert len(chapters) == 3
+    assert chapters[0]["title"] == "Introduction"
+    assert chapters[1]["title"] == "Chapter 1: Ethernet Switching"
+    assert chapters[2]["title"] == "Chapter 2: IPv4 Routing"
+    assert "```text" in chapters[1]["content"]
+
+    # 3. Test fallback when no outline (heading regex detection)
+    pdf_no_outline = make_mock_pdf("Manual", "Admin", with_outline=False)
+    t2, a2, chs2 = parse_pdf(pdf_no_outline, filename="network_manual.pdf")
+    assert len(chs2) >= 2
+    assert any("Chapter 1" in ch["title"] for ch in chs2)
+
+    # 4. Test Scanned PDF Guard (< 100 words raises ValueError)
+    empty_writer = pypdf.PdfWriter()
+    empty_writer.add_blank_page(width=100, height=100)
+    buf_scanned = io.BytesIO()
+    empty_writer.write(buf_scanned)
+    import pytest
+    with pytest.raises(ValueError, match="текстового слоя"):
+        parse_pdf(buf_scanned.getvalue(), filename="scan.pdf")
+
+
+def test_books_pdf_multipart_upload():
+    # 1. Successful PDF upload
+    pdf_data = make_mock_pdf()
+    files = {"file": ("ccna_cert_guide.pdf", pdf_data, "application/pdf")}
+    resp = client.post("/api/books", files=files)
+    assert resp.status_code == 200
+    book = resp.json()
+    assert book["title"] == "ccna_cert_guide"
+    assert book["source_type"] == "pdf"
+    assert len(book["chapters"]) == 3
+    assert book["total_words"] > 100
+    book_id = book["id"]
+
+    # Verify retrieval
+    get_resp = client.get(f"/api/books/{book_id}")
+    assert get_resp.status_code == 200
+    assert get_resp.json()["source_type"] == "pdf"
+
+    # 2. Upload scanned PDF without OCR text returns 400 with informative detail
+    import io, pypdf
+    empty_writer = pypdf.PdfWriter()
+    empty_writer.add_blank_page(width=100, height=100)
+    buf = io.BytesIO()
+    empty_writer.write(buf)
+
+    scan_files = {"file": ("scanned_cisco.pdf", buf.getvalue(), "application/pdf")}
+    err_resp = client.post("/api/books", files=scan_files)
+    assert err_resp.status_code == 400
+    assert "текстового слоя" in err_resp.json()["detail"]
+
+    # Cleanup
+    client.delete(f"/api/books/{book_id}")