personal-data/personal_data/fetchers/playstation.py

import secrets
import logging

from personal_data.data import Scraper, DeduplicateMode

logger = logging.getLogger(__name__)

URL_RECENTLY_PLAYED_HTML = 'https://library.playstation.com/recently-played'
URL_RECENTLY_PLAYED_API = "https://web.np.playstation.com/api/graphql/v1/op?operationName=getUserGameList&variables=%7B%22limit%22%3A50%2C%22categories%22%3A%22ps4_game%2Cps5_native_game%22%7D&extensions=%7B%22persistedQuery%22%3A%7B%22version%22%3A1%2C%22sha256Hash%22%3A%22e0136f81d7d1fb6be58238c574e9a46e1c0cc2f7f6977a08a5a46f224523a004%22%7D%7D"

def scrape_played_last(session):
    # Initial request to trigger cookie.
    logger.warning('Trying to trigger initial cookie usage')

    response = session.get(URL_RECENTLY_PLAYED_HTML, cookies = session.cookies)
    response.raise_for_status()

    print('From herp')
    for cookie in session.cookies:
        print('    ', cookie.domain, cookie)
    exit(1)

    # Now trigger API call.
    logger.warning('Trying to fetch data from API')
    headers = {
        "User-Agent": "Mozilla/5.0 (X11; Linux x86_64; rv:109.0) Gecko/20100101 Firefox/119.0",
        "Accept": "application/json",
        "Accept-Language": "en-US,en;q=0.5",
        "content-type": "application/json",
        "X-PSN-App-Ver": "my-playstation/0.1.0-20230720235210-hotfix-1-g1e9f07ff-1e9f07ff247eafea4d9d9236f73863cb9cd5d3e8",
        "X-PSN-Correlation-Id": "bc7a39b1-a99c-478b-9494-d2fddf189875",
        "apollographql-client-name": "my-playstation",
        "apollographql-client-version": "0.1.0-20230720235210-hotfix-1-g1e9f07ff",
        "X-PSN-Request-Id": "8ad64653-d8b5-4941-b565-b5536c9853df",
        'Referer': 'https://library.playstation.com/',
        'Origin': 'https://library.playstation.com',
    }
    result = session.get(URL_RECENTLY_PLAYED_API, headers = headers)
    result.raise_for_status()

    print(result.json())
    games_data = result.json()['data']['gameLibraryTitlesRetrieve']['games']
    print(games_data)
    for game_data in games_data:
        yield {
                # Important fields
                'game.name': game_data['name'],
                'me.last_played_time': game_data['lastPlayedDateTime'],
                'playstation.product_id': game_data['productId'],

                # Secondary fields
                'playstation.concept_id': game_data['conceptId'],
                'playstation.title_id': game_data['titleId'],
                'playstation.entitlement_id': game_data['entitlementId'],
                'me.acquired_by_playstation_membership': game_data['membership'],
                'game.platform': game_data['platform'],
                'game.icon': game_data['image']['url'],
        }

SCRAPERS = [
        Scraper(scrape_played_last, 'games_played_playstation',
                deduplicate_mode = DeduplicateMode.BY_ALL_COLUMNS)
]