app/services/catalog_hub.py

"""API-first catalog hub — resolve book identity and platform listings.

Uses only public JSON APIs (Google Books, Open Library, Goodreads autocomplete,
Apple iTunes) plus Goodreads buy links. No retailer HTML scraping.
"""

from __future__ import annotations

import asyncio
import urllib.parse
from dataclasses import dataclass, field

import httpx
import structlog

from app.services.book_url_scraper import (
    _google_by_isbn,
    _google_search_best,
    _normalize_isbn,
    _score_itunes_item,
)
from app.services.isbn_catalog import (
    classify_retailer_url,
    enrich_retailer_links_from_isbn,
    extract_amazon_asin,
    is_confirmed_isbn13,
    lookup_open_library,
    resolve_google_books_html,
    sanitize_platform_url,
)

logger = structlog.get_logger(__name__)

_TIMEOUT = 8.0
VERIFIED_THRESHOLD = 0.85
LIKELY_THRESHOLD = 0.70

# Import platform IDs that map to scored platform IDs
SOURCE_PLATFORM_ALIASES: dict[str, str] = {
    "google_play": "google_books",
}


@dataclass
class CatalogSnapshot:
    """Canonical book identity from public catalog APIs."""

    title: str = ""
    author: str = ""
    isbn13: str | None = None
    isbn10: str | None = None
    asin: str | None = None
    gb_volume_id: str | None = None
    gb_confidence: float = 0.0
    ol_key: str | None = None
    ol_confidence: float = 0.0
    goodreads_id: str | None = None
    goodreads_url: str | None = None
    goodreads_confidence: float = 0.0
    apple_url: str | None = None
    apple_confidence: float = 0.0
    retailer_links: dict[str, str] = field(default_factory=dict)

    @property
    def confirmed_isbn13(self) -> str | None:
        """ISBN-13 confirmed by at least one catalog API or valid import."""
        if self.isbn13 and (self.gb_confidence >= LIKELY_THRESHOLD or self.ol_confidence >= LIKELY_THRESHOLD):
            return self.isbn13
        if self.isbn13 and self.gb_volume_id and self.ol_key:
            return self.isbn13
        return None

    def google_books_url(self) -> str | None:
        if self.gb_volume_id:
            return f"https://books.google.com/books?id={self.gb_volume_id}"
        return None

    def open_library_url(self) -> str | None:
        if self.isbn13:
            return f"https://openlibrary.org/isbn/{self.isbn13}"
        if self.ol_key:
            return f"https://openlibrary.org{self.ol_key}"
        return None

    def amazon_product_url(self) -> str | None:
        if self.asin:
            return f"https://www.amazon.com/dp/{self.asin}"
        if self.isbn10 and self.isbn10.isdigit():
            return f"https://www.amazon.com/dp/{self.isbn10}"
        return None


@dataclass
class PlatformHit:
    """Discovery result for one retailer."""

    confidence: float
    listing_url: str | None
    isbn_match: bool = False


def normalize_source_platform(platform_id: str | None) -> str | None:
    """Map import platform IDs to scored platform IDs."""
    if not platform_id:
        return None
    return SOURCE_PLATFORM_ALIASES.get(platform_id, platform_id)


def isbn_retailer_urls(isbn13: str) -> dict[str, str]:
    """ISBN-targeted lookup URLs — honest search, not fake product pages."""
    q = urllib.parse.quote(isbn13)
    return {
        "barnes_noble": f"https://www.barnesandnoble.com/s/{q}",
        "kobo": f"https://www.kobo.com/us/en/search?query={q}",
        "abebooks": f"https://www.abebooks.com/servlet/SearchResults?isbn={q}",
        "thriftbooks": f"https://www.thriftbooks.com/browse/?b.search={q}",
        "bookshop": f"https://bookshop.org/search?keywords={q}",
    }


async def _resolve_google(
    client: httpx.AsyncClient,
    title: str,
    author: str,
    isbn: str | None,
) -> dict:
    """Google Books API lookup with HTML fallback when quota is exceeded."""
    from app.services.platform_presence import title_match_score

    gb: dict = {}
    if isbn:
        gb = await _google_by_isbn(isbn, client, author_hint=author)
    if not gb.get("title"):
        gb = await _google_search_best(title, author, client)
    if not gb.get("title") and is_confirmed_isbn13(isbn):
        gb = await resolve_google_books_html(client, isbn, title, author)
    if not gb.get("title"):
        return {}
    conf = title_match_score(title, gb["title"], author, gb.get("author", ""))
    isbn_match = bool(
        isbn and gb.get("isbn") and _normalize_isbn(isbn) == _normalize_isbn(gb["isbn"])
    )
    if isbn_match or gb.get("isbn_match"):
        conf = max(conf, 0.95)
    return {
        "confidence": gb.get("confidence", conf),
        "volume_id": gb.get("volume_id") or "",
        "isbn13": _normalize_isbn(gb.get("isbn13") or gb.get("isbn") or ""),
        "isbn10": _normalize_isbn(gb.get("isbn10") or ""),
        "title": gb.get("title", ""),
        "author": gb.get("author", ""),
        "isbn_match": isbn_match or bool(gb.get("isbn_match")),
    }


async def _resolve_apple(
    client: httpx.AsyncClient,
    title: str,
    author: str,
) -> dict:
    """Apple iTunes Search API lookup."""
    from app.services.platform_presence import search_title, title_match_score

    core = search_title(title) or title
    if not core:
        return {}
    term = f"{core} {author}".strip() if author else core
    try:
        resp = await client.get(
            "https://itunes.apple.com/search",
            params={"term": term, "entity": "ebook", "limit": "8", "country": "US"},
            timeout=_TIMEOUT,
        )
        if resp.status_code != 200:
            return {}
        results = resp.json().get("results", [])
        if not results:
            return {}
        best = max(results, key=lambda item: _score_itunes_item(item, core, author))
        if _score_itunes_item(best, core, author) < 4.0:
            return {}
        found_title = best.get("trackName") or ""
        conf = title_match_score(core, found_title, author, best.get("artistName") or "")
        if conf < LIKELY_THRESHOLD:
            return {}
        return {"confidence": conf, "url": best.get("trackViewUrl")}
    except Exception as exc:
        logger.debug("Apple catalog lookup failed", error=str(exc))
        return {}


async def build_catalog_snapshot(
    client: httpx.AsyncClient,
    *,
    title: str,
    author: str,
    isbn: str | None = None,
    asin: str | None = None,
    goodreads_links: dict[str, str] | None = None,
    goodreads_url: str | None = None,
    goodreads_confidence: float = 0.0,
    resolved_author: str = "",
    skip_enrich: bool = False,
) -> CatalogSnapshot:
    """Resolve book identity from public APIs in parallel."""
    from app.services.platform_presence import search_title

    core_title = search_title(title) or title.strip()
    norm_isbn = _normalize_isbn(isbn) if isbn else None
    if is_confirmed_isbn13(norm_isbn):
        catalog_isbn = norm_isbn
    else:
        catalog_isbn = None
    norm_asin = (asin or "").upper() or None
    if not norm_asin and isbn and len(_normalize_isbn(isbn) or "") == 10:
        candidate = (isbn or "").upper()
        if candidate.startswith("B0"):
            norm_asin = candidate

    gb_task = asyncio.create_task(_resolve_google(client, core_title, author, catalog_isbn))
    ol_task = asyncio.create_task(
        lookup_open_library(client, core_title, author, catalog_isbn)
    )
    apple_task = asyncio.create_task(_resolve_apple(client, core_title, author))

    gb, ol, apple = await asyncio.gather(gb_task, ol_task, apple_task)

    snap = CatalogSnapshot(title=core_title, author=resolved_author or author)
    snap.retailer_links = dict(goodreads_links or {})

    if gb:
        snap.gb_volume_id = gb.get("volume_id") or None
        snap.gb_confidence = gb.get("confidence", 0.0)
        if gb.get("isbn13"):
            snap.isbn13 = gb["isbn13"]
        if gb.get("isbn10"):
            snap.isbn10 = gb["isbn10"]
        if gb.get("author") and not snap.author:
            snap.author = gb["author"]

    if ol:
        snap.ol_key = ol.get("work_key")
        snap.ol_confidence = ol.get("confidence", 0.0)
        if ol.get("isbn13") and not snap.isbn13:
            snap.isbn13 = ol["isbn13"]
        if ol.get("author") and not snap.author:
            snap.author = ol["author"]

    if is_confirmed_isbn13(norm_isbn):
        snap.isbn13 = norm_isbn
    elif catalog_isbn and not snap.isbn13:
        snap.isbn13 = catalog_isbn

    if norm_asin:
        snap.asin = norm_asin
    elif isbn:
        extracted = extract_amazon_asin(isbn) or (
            isbn.upper() if isbn.upper().startswith("B0") and len(isbn) == 10 else None
        )
        if extracted:
            snap.asin = extracted

    if apple:
        snap.apple_url = apple.get("url")
        snap.apple_confidence = apple.get("confidence", 0.0)

    if goodreads_url:
        snap.goodreads_url = goodreads_url
        snap.goodreads_confidence = max(goodreads_confidence, 0.95)

    if not skip_enrich and (
        is_confirmed_isbn13(snap.isbn13) or (core_title and (snap.author or author))
    ):
        snap.retailer_links = await enrich_retailer_links_from_isbn(
            client,
            snap.retailer_links,
            isbn13=snap.isbn13,
            title=core_title,
            author=snap.author or author,
        )

    return snap


def discover_platform_hit(
    platform_id: str,
    snap: CatalogSnapshot,
    *,
    source_platform: str | None = None,
    source_url: str | None = None,
    buy_url: str | None = None,
) -> PlatformHit:
    """Return a hit only when the book is definitively listed (product page or API)."""
    src = normalize_source_platform(source_platform)

    if src == platform_id and (source_url or buy_url):
        url = sanitize_platform_url(buy_url or source_url or "", platform_id)
        if url and classify_retailer_url(url, platform_id) == "product":
            return PlatformHit(1.0, url, isbn_match=True)

    if platform_id == "google_books":
        if snap.gb_volume_id and snap.gb_confidence >= VERIFIED_THRESHOLD:
            return PlatformHit(
                snap.gb_confidence,
                snap.google_books_url(),
                snap.gb_confidence >= 0.95,
            )

    if platform_id == "open_library" and snap.ol_confidence >= VERIFIED_THRESHOLD:
        url = snap.open_library_url()
        if url:
            return PlatformHit(snap.ol_confidence, url)

    if (
        platform_id == "goodreads"
        and snap.goodreads_url
        and snap.goodreads_confidence >= VERIFIED_THRESHOLD
    ):
        return PlatformHit(snap.goodreads_confidence, snap.goodreads_url)

    if (
        platform_id == "apple_books"
        and snap.apple_url
        and snap.apple_confidence >= VERIFIED_THRESHOLD
    ):
        return PlatformHit(snap.apple_confidence, snap.apple_url)

    gr_url = snap.retailer_links.get(platform_id)
    if gr_url:
        clean = sanitize_platform_url(gr_url, platform_id)
        if clean and classify_retailer_url(clean, platform_id) == "product":
            return PlatformHit(0.95, clean)

    if platform_id == "amazon" and snap.asin:
        amazon_url = snap.amazon_product_url()
        if amazon_url:
            return PlatformHit(0.98, amazon_url, isbn_match=True)

    return PlatformHit(0.0, None)