VanguardReMaster/tools/import_card_art.py

#!/usr/bin/env python3
"""Скачать арты триальной деки с Cardfight!! Vanguard Wiki.

Пример:
    python tools/import_card_art.py \
        --url "https://cardfight.fandom.com/wiki/Trial_Deck_1:_Blaster_Blade" \
        --output client/Assets/Vanguard/Resources/CardArt

Fandom иногда отвечает Cloudflare 403 на запросы не из браузера. В этом
случае сохраните HTML/API-ответ страницы из браузера и передайте:
    python tools/import_card_art.py --html saved.html --output ...
"""

from __future__ import annotations

import argparse
import html
import json
import re
import sys
import urllib.parse
import urllib.request
from pathlib import Path


def fetch(url: str) -> str:
    request = urllib.request.Request(
        url,
        headers={
            "User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) "
                          "AppleWebKit/537.36 Chrome/126 Safari/537.36",
            "Accept": "text/html,application/json;q=0.9,*/*;q=0.8",
        },
    )
    with urllib.request.urlopen(request, timeout=60) as response:
        return response.read().decode("utf-8", errors="replace")


def page_html(url: str, local_html: str | None) -> str:
    if local_html:
        raw = Path(local_html).read_text(encoding="utf-8", errors="replace")
        if raw.lstrip().startswith("{"):
            payload = json.loads(raw)
            return payload["parse"]["text"]
        return raw

    # Пытаемся через API: он отдаёт HTML статьи без рекламной обвязки.
    page = url.split("/wiki/", 1)[-1]
    api = (
        "https://cardfight.fandom.com/api.php?action=parse&format=json"
        "&formatversion=2&prop=text&page=" + urllib.parse.quote(page)
    )
    raw = fetch(api)
    payload = json.loads(raw)
    if "error" in payload:
        raise RuntimeError(payload["error"].get("info", "Wiki API error"))
    return payload["parse"]["text"]


def set_gallery_links(page: str) -> list[str]:
    links = re.findall(r'href="(/wiki/Set_Gallery:[^"]+)"', page)
    return list(dict.fromkeys("https://cardfight.fandom.com" + link for link in links))


def image_candidates(html_text: str, code: str) -> list[str]:
    """Ищем картинку, у которой alt начинается с кода принта."""
    pattern = re.compile(r"<img\b[^>]*>", re.I)
    result: list[tuple[int, str]] = []
    for tag in pattern.findall(html_text):
        alt_match = re.search(r'\balt="([^"]*)"', tag, re.I)
        src_match = re.search(r'\bsrc="([^"]*)"', tag, re.I)
        if not alt_match or not src_match:
            continue

        alt = html.unescape(alt_match.group(1))
        src = html.unescape(src_match.group(1))
        if not alt.startswith(code) or not src.startswith("https://static.wikia.nocookie.net/"):
            continue

        score = 0
        if "Sample" in alt:
            score += 10
        if "Holo" in alt or "Parallel" in alt:
            score += 5
        result.append((score, src))

    result.sort(key=lambda item: item[0])
    urls = []
    for _, src in result:
        # Берём оригинал, а не превью 126px.
        src = re.sub(r"/scale-to-width-down/\d+", "", src)
        src = src.split("?")[0] + "?format=original"
        if src not in urls:
            urls.append(src)
    return urls


def save_image(url: str, output: Path, image_key: str) -> None:
    extension = ".png" if ".png" in url.lower() else ".jpg"
    target = output / (image_key + extension)
    if target.exists():
        return

    request = urllib.request.Request(url, headers={"User-Agent": "Mozilla/5.0"})
    with urllib.request.urlopen(request, timeout=60) as response:
        data = response.read()
    target.write_bytes(data)
    print(f"  {target.name} ({len(data) // 1024} KiB)")


def main() -> int:
    parser = argparse.ArgumentParser()
    parser.add_argument("--url", required=False, help="Страница триал-деки")
    parser.add_argument("--html", help="Сохранённый HTML/API-ответ страницы")
    parser.add_argument("--output", required=True, help="Папка CardArt")
    args = parser.parse_args()

    if not args.url and not args.html:
        parser.error("нужен --url или --html")

    output = Path(args.output)
    output.mkdir(parents=True, exist_ok=True)

    source_url = args.url or "https://cardfight.fandom.com/wiki/"
    page = page_html(source_url, args.html)

    galleries = set_gallery_links(page)
    if not galleries:
        print("Не нашёл ссылку Set_Gallery на странице.", file=sys.stderr)
        return 1

    # Английские VGE-галереи приоритетнее, чтобы не перезаписать арт
    # японской/тайский версией.
    galleries.sort(key=lambda value: 0 if "/VGE-" in value else 1 if "/VG-" in value else 2)

    downloaded = 0
    for gallery in galleries:
        try:
            gallery_html = page_html(gallery, None)
        except Exception as exc:  # noqa: BLE001
            print(f"Пропуск {gallery}: {exc}", file=sys.stderr)
            continue

        # Коды вида TD01/001EN, alt картинки: "TD01/001EN Name".
        codes = sorted(set(re.findall(r"\b(TD\d{2}/\d{3})(?:EN)?\b", gallery_html)))
        for code in codes:
            urls = image_candidates(gallery_html, code)
            if not urls:
                continue
            image_key = code.lower().replace("/", "_")
            try:
                save_image(urls[0], output, image_key)
                downloaded += 1
            except Exception as exc:  # noqa: BLE001
                print(f"Ошибка {code}: {exc}", file=sys.stderr)

    print(f"Готово: обработано карт {downloaded}.")
    return 0


if __name__ == "__main__":
    raise SystemExit(main())