PixivSalvage.py

import itertools
from datetime import datetime, timedelta
import time
from typing import List, Optional, Tuple

import requests

PIXIV_ARTWORKS_URL = "https://www.pixiv.net/artworks/"
ARTWORKS_DETAIL_ENDPOINT = "https://www.pixiv.net/ajax/illust/"
ARTWORKS_EXT_LIST = ("jpg", "png", "gif") #複数の拡張子を混ぜるとjpgに変換されるのでjpg優先
ARTWORKS_TEMPLATE = (
    "https://i.pximg.net/img-original/img/{timestamp}/{id}_p{number}.{ext}"
)

session = requests.Session()
session.headers.update(
    {
        "Referer": "https://pixiv.net/",
        "User-Agent": "Mozilla/5.0 (iPhone; CPU iPhone OS 18_3_2 like Mac OS X) AppleWebKit/605.1.15 (KHTML, like Gecko) CriOS/135.0.7049.53 Mobile/15E148 Safari/604.1",
    }
)


def get_artwork_timestamp(artwork_id: int) -> Optional[str]:
    try:
        response = session.get(f"{ARTWORKS_DETAIL_ENDPOINT}{artwork_id}", timeout=10)
        response.raise_for_status()
    except requests.RequestException as e:
        print(f"IDの取得に失敗しました {artwork_id}: {e}")
        return None

    try:
        artwork_data = response.json().get("body", {}).get("userIllusts", {})
        timestamp = artwork_data.get(str(artwork_id), {}).get("updateDate")
        if not timestamp:
            print(f"タイムスタンプが見つかりませんでした {artwork_id}")
        return timestamp
    except ValueError as e:
        print(f"Jsonのパースに失敗しました {artwork_id}: {e}")
        return None


def find_valid_id(
    start_id: int, step: int, max_attempts: int = 30, interval: float = 0.5
) -> Optional[Tuple[int, str]]:
    current_id = start_id
    for attempt in range(max_attempts):
        timestamp = get_artwork_timestamp(current_id)
        if timestamp is not None:
            print(f"有効なIDを見つけました: {current_id} タイムスタンプ: {timestamp}")
            return current_id, timestamp
        else:
            print(f"タイムスタンプの取得に失敗しました: {current_id}, 次のIDを取得します...")
            current_id += step
            time.sleep(interval)

    print(
        f"ID {start_id} から {max_attempts} 回試行しましたが、有効なIDを見つけられませんでした"
    )
    return None


def generate_timestamps(start: datetime, end: datetime):
    current = start
    while current <= end:
        yield current.strftime("%Y/%m/%d/%H/%M/%S")
        current += timedelta(seconds=1)


def check_all_pattern_artworks(
    start: datetime, end: datetime, artwork_id: int, interval: float = 0.5
) -> Optional[Tuple[str, str, str]]:
    for ext in ARTWORKS_EXT_LIST:
        for ts in generate_timestamps(start, end):
            url = ARTWORKS_TEMPLATE.format(
                timestamp=ts, id=artwork_id, number=0, ext=ext
            )
            response = session.head(url, timeout=10)

            if response.status_code == 200:
                print("ベースリンクの取得に成功しました:" + url)
                return url, ts, ext
            else:
                print("URLが見つかりませんでした:" + url)

            time.sleep(interval)

    return None


def get_all_artworks_images(artwork_id: int, timestamp: str, ext: str, interval: float = 0.5) -> List[str]:
    urls = []
    for i in itertools.count():
        url = ARTWORKS_TEMPLATE.format(
            timestamp=timestamp, id=artwork_id, number=i, ext=ext
        )
        response = session.head(url, timeout=10)
        if response.status_code == 200:
            urls.append(url)
            print("取得に成功しました:" + url)
            time.sleep(interval)
        else:
            break

    return urls


def salvage_pixiv(artwork_id: int, interval: float = 0.5) -> Optional[List[str]]:
    start_result = find_valid_id(start_id=artwork_id - 1, step=-1, interval=interval)
    if start_result is None:
        return None
    start_id, start_timestamp = start_result

    end_result = find_valid_id(start_id=artwork_id + 1, step=1)
    if end_result is None:
        return None
    end_id, end_timestamp = end_result

    start_datetime = datetime.fromisoformat(start_timestamp)
    end_datetime = datetime.fromisoformat(end_timestamp)

    result = check_all_pattern_artworks(start_datetime, end_datetime, artwork_id, interval)
    if result is None:
        return None
    url, ts, ext = result
    return get_all_artworks_images(artwork_id, ts, ext, interval)


if "__main__" == __name__:
    artwork_id = int(input("IDを入力してください:").strip())
    interval = 0.75 #HEADがメインなので0.5秒くらいでも多分大丈夫。ただ長ければ長いほど安全
    result = salvage_pixiv(artwork_id, interval=interval)

    if result is not None:
        BAR_LENGTH = 80
        print("=" * BAR_LENGTH)
        print(f"{len(result)}件の画像をサルベージしました")
        print("-" * BAR_LENGTH)

        for i, url in enumerate(result, start=1):
            print(f"[{i:02d}] {url}")

        print("=" * BAR_LENGTH)
    else:
        print("サルベージに失敗しました")