test_modules/test_integration.py

"""
Comprehensive integration tests for the ReadBookMom app.

Tests cover:
  - Module imports and structure
  - Story loading and text processing
  - HTML rendering functions (dashboard, library, voices, story text)
  - Gradio component wiring (inputs/outputs match handler signatures)
  - State machine transitions (book select → play → pause → ask → resume)
  - XSS escaping in all HTML renderers
  - Voice profile integration in player
  - TTS module chunk splitting
  - Q&A answer rendering safety
"""
import os
import sys
import html as html_module
import inspect

sys.path.insert(0, os.path.dirname(os.path.dirname(__file__)))

PASS = 0
FAIL = 0


def check(label, condition, detail=""):
    global PASS, FAIL
    if condition:
        PASS += 1
        print(f"  [OK] {label}")
    else:
        FAIL += 1
        print(f"  [FAIL] {label} -- {detail}")


# ── Module imports ──────────────────────────────────────────────

def test_module_imports():
    print("\n=== Module imports ===")
    try:
        from tts import split_into_chunks, generate_audio_stream
        check("tts module imports", True)
    except Exception as e:
        check("tts module imports", False, str(e))

    try:
        from voice_clone import (
            create_voice_profile, synthesize_cloned, synthesize_cloned_preview,
            synthesize_custom_voice, has_profile, get_model_mode,
            _trim_reference_audio, _select_dtype, _select_attn_impl,
            GENERATION_PARAMS, BASE_MODEL_ID, CUSTOM_VOICE_MODEL_ID,
        )
        check("voice_clone module imports", True)
    except Exception as e:
        check("voice_clone module imports", False, str(e))

    try:
        from inference import transcribe_audio, answer_story_question
        check("inference module imports", True)
    except Exception as e:
        check("inference module imports", False, str(e))


# ── Story loading ───────────────────────────────────────────────

def test_story_loading():
    print("\n=== Story loading ===")
    from app import load_books_from_stories, load_paragraphs

    books = load_books_from_stories()
    check("Stories found", len(books) > 0, f"found {len(books)}")

    for b in books:
        required = ["id", "title", "story_path", "cover_url", "voice_name", "percentage"]
        missing = [k for k in required if k not in b]
        check(f"Book '{b['title']}' has all keys", len(missing) == 0, f"missing: {missing}")

    # Test paragraph loading
    if books:
        paras = load_paragraphs(books[0]["story_path"])
        check(f"Paragraphs loaded from '{books[0]['title']}'", len(paras) > 0, f"got {len(paras)}")

    # Test non-existent file
    paras = load_paragraphs("nonexistent_file.txt")
    check("Non-existent file returns empty list", paras == [])


# ── TTS chunk splitting ────────────────────────────────────────

def test_chunk_splitting():
    print("\n=== TTS chunk splitting ===")
    from tts import split_into_chunks

    chunks = split_into_chunks("Hello world. How are you? I am fine!")
    check("Splits on sentence boundaries", len(chunks) == 3, f"got {len(chunks)}: {chunks}")

    chunks = split_into_chunks("Single sentence without ending period")
    check("Single sentence stays intact", len(chunks) == 1, f"got {len(chunks)}")

    chunks = split_into_chunks("")
    check("Empty text returns empty list", len(chunks) == 0)

    chunks = split_into_chunks("   ")
    check("Whitespace-only returns empty list", len(chunks) == 0)

    long_text = "First sentence. Second sentence! Third? Fourth; Fifth."
    chunks = split_into_chunks(long_text)
    check("Splits on .!?; delimiters", len(chunks) == 5, f"got {len(chunks)}: {chunks}")


# ── HTML rendering ──────────────────────────────────────────────

def test_html_rendering():
    print("\n=== HTML rendering ===")
    from app import (
        generate_dashboard_html, generate_library_html,
        render_cloned_voices_html, render_story_text, mock_books, mock_voices,
    )

    # Dashboard
    dash_html = generate_dashboard_html(mock_books)
    check("Dashboard renders without error", isinstance(dash_html, str) and len(dash_html) > 100)
    check("Dashboard contains book titles", any(b["title"] in dash_html for b in mock_books))

    # Library
    lib_html = generate_library_html(mock_books)
    check("Library renders without error", isinstance(lib_html, str) and len(lib_html) > 100)
    check("Library contains Tap to Play", "Tap to Play" in lib_html)

    # Library with search filter
    if mock_books:
        first_title = mock_books[0]["title"]
        filtered = generate_library_html(mock_books, query=first_title)
        check(f"Library filters to '{first_title}'", first_title in filtered)

    # Library empty result
    empty = generate_library_html(mock_books, query="xyznonexistent999")
    check("Library shows no-match message", "No audiobooks match" in empty)

    # Cloned voices
    voices_html = render_cloned_voices_html(mock_voices)
    check("Voices grid renders", isinstance(voices_html, str) and len(voices_html) > 50)

    # Story text
    paras = ["Paragraph one.", "Paragraph two.", "Paragraph three."]
    story_html = render_story_text(paras, 1)
    check("Story text highlights current paragraph", "border-left: 3px solid #f5841f" in story_html)
    check("Story text renders all paragraphs", "Paragraph one" in story_html and "Paragraph three" in story_html)

    # Empty paragraphs
    check("Empty paragraphs returns empty string", render_story_text([], 0) == "")


# ── XSS escaping ───────────────────────────────────────────────

def test_xss_escaping():
    print("\n=== XSS escaping ===")
    from app import render_story_text, render_cloned_voices_html

    # Story text with HTML injection
    xss_paras = ['<script>alert("xss")</script>', 'Normal text']
    result = render_story_text(xss_paras, 0)
    check("Story text escapes <script>", "<script>" not in result)
    check("Story text contains escaped version", "&lt;script&gt;" in result)

    # Voices with HTML injection
    xss_voices = [{
        "id": "v1", "name": '<img onerror="alert(1)" src=x>',
        "avatar_url": "👩", "description": "test", "gender": "Female",
        "created_date": "2026-01-01", "status": "ready"
    }]
    result = render_cloned_voices_html(xss_voices)
    check("Voice name escapes HTML", "<img onerror" not in result)
    check("Voice name contains escaped version", "&lt;img" in result)


# ── Gradio component wiring ────────────────────────────────────

def test_gradio_wiring():
    print("\n=== Gradio component wiring ===")
    from app import demo

    # Verify the demo has the expected blocks
    blocks = demo.blocks
    check("Demo has blocks", len(blocks) > 0, f"found {len(blocks)}")

    # Check that no component references a deleted 'loading_spinner'
    source = open("app.py", encoding="utf-8").read()
    check("No loading_spinner references", "loading_spinner" not in source)
    check("No empty_state_sim references", "empty_state_sim" not in source)
    check("No offline_toggle references", "offline_toggle" not in source)
    check("No system_error_modal references", "system_error_modal" not in source)
    check("No narrative_subtitles references", "narrative_subtitles" not in source)


# ── Handler signatures vs outputs ──────────────────────────────

def test_handler_signatures():
    print("\n=== Handler signatures ===")
    source = open("app.py", encoding="utf-8").read()

    # Verify key handler functions exist
    handlers = [
        "def filter_books_shelf",
        "def handle_book_select",
        "def animate_cloning_pipeline",
        "def save_new_cloned_voice",
        "def stream_tts",
        "def enter_paused_state",
        "def update_story_highlight",
        "def enter_asking_state",
        "def handle_question_submit",
        "def enter_resume_state",
    ]
    for h in handlers:
        check(f"Handler {h.split('def ')[1]} exists", h in source)

    # Verify removed handlers are gone
    check("check_story_finished removed (inline in stream_tts)", "def check_story_finished" not in source)
    check("toggle_outage_sandbox removed", "def toggle_outage_sandbox" not in source)


# ── Voice profile integration ──────────────────────────────────

def test_voice_profile_integration():
    print("\n=== Voice profile integration ===")
    from voice_clone import has_profile

    check("has_profile(None) is False", has_profile(None) is False)
    check("has_profile('') is False", has_profile("") is False)
    check("has_profile('nonexistent') is False", has_profile("nonexistent") is False)

    # Verify handle_book_select accepts profile_id
    source = open("app.py", encoding="utf-8").read()
    check("handle_book_select takes profile_id",
          "def handle_book_select(title_chosen, current_inventory, profile_id)" in source)
    check("book_selector inputs include voice_profile_state",
          "inputs=[book_selector, books_state, voice_profile_state]" in source)


# ── State machine consistency ──────────────────────────────────

def test_state_machine():
    print("\n=== State machine consistency ===")
    source = open("app.py", encoding="utf-8").read()

    # Play → shows pause_btn, hides play_btn
    check("stream_tts hides play_btn during playback",
          "gr.Button(visible=False), gr.Button(visible=True), gr.Button(visible=True)" in source)

    # Pause → shows play_btn, hides pause_btn
    check("enter_paused_state shows play_btn",
          "gr.Button(visible=True),\n            gr.Button(visible=False)," in source or
          ("enter_paused_state" in source and "gr.Button(visible=True)" in source))

    # Resume → shows play_btn (ready state), hides resume_btn
    check("enter_resume_state shows ready state",
          "READY" in source and "Tap Play to continue" in source)

    # Ask → shows resume_btn, hides play/pause/ask
    check("enter_asking_state shows resume_btn",
          "PAUSED" in source and "ASKING" in source)

    # stream_tts updates timeline_slider
    check("stream_tts outputs include timeline_slider",
          "timeline_slider, story_text_display, story_finished_panel" in source)


# ── App launch readiness ───────────────────────────────────────

def test_app_launch():
    print("\n=== App launch readiness ===")
    from app import demo

    check("Demo object exists", demo is not None)
    check("Demo is a gr.Blocks instance", "Blocks" in type(demo).__name__)

    # Check allowed_paths in launch call
    source = open("app.py", encoding="utf-8").read()
    check("Launch has allowed_paths for assets", 'allowed_paths' in source)
    check("Launch uses .queue()", '.queue()' in source)


# ── File structure ──────────────────────────────────────────────

def test_file_structure():
    print("\n=== File structure ===")
    required_files = [
        "app.py", "tts.py", "voice_clone.py", "inference.py",
        "requirements.txt", "static/style.css",
    ]
    for f in required_files:
        path = os.path.join(os.path.dirname(os.path.dirname(__file__)), f)
        check(f"File exists: {f}", os.path.exists(path))

    # Check stories directory has content
    stories_dir = os.path.join(os.path.dirname(os.path.dirname(__file__)), "stories")
    if os.path.isdir(stories_dir):
        txt_files = [f for f in os.listdir(stories_dir) if f.endswith(".txt")]
        check("Stories directory has .txt files", len(txt_files) > 0, f"found {len(txt_files)}")
    else:
        check("Stories directory exists", False, "stories/ not found")


if __name__ == "__main__":
    print("=" * 60)
    print("Comprehensive Integration Tests — ReadBookMom")
    print("=" * 60)

    test_module_imports()
    test_story_loading()
    test_chunk_splitting()
    test_html_rendering()
    test_xss_escaping()
    test_gradio_wiring()
    test_handler_signatures()
    test_voice_profile_integration()
    test_state_machine()
    test_app_launch()
    test_file_structure()

    print("\n" + "=" * 60)
    print(f"Results: {PASS} passed, {FAIL} failed")
    print("=" * 60)
    sys.exit(1 if FAIL > 0 else 0)