personal-data/personal_data/fetchers/psnprofiles.py

import dataclasses
import logging
import re
from collections.abc import Iterator

import bs4
import requests_util

import personal_data.html_util
from personal_data import secrets
from personal_data.data import DeduplicateMode, Scraper

from .. import parse_util

logger = logging.getLogger(__name__)

URL_API_ROOT = 'https://psnprofiles.com/'
URL_PROFILE = URL_API_ROOT + '{psn_id}'
URL_USER_GAME_TROPHIES = URL_API_ROOT + 'trophies/{game_id}/{psn_id}'


def game_psnprofiles_id_from_url(relative_url: str) -> int:
    m = re.match(r'/(?:trophy|trophies)/(\d+)\-(?:[\w-]+)(/[\w-]*)?', relative_url)
    result = m.group(1)
    return int(result)


MAX_NUMBER_GAMES_TO_PARSE = 1000


@dataclasses.dataclass(frozen=True)
class PsnProfilesScraper(Scraper):
    dataset_name = 'games_played_playstation'
    deduplicate_mode = DeduplicateMode.BY_ALL_COLUMNS

    @staticmethod
    def requires_cfscrape() -> bool:
        return True

    def scrape(self):
        self._setup_cache()
        games_rows = list(self._scrape_games_overview())
        games_ids = {row['psnprofiles.game_id']: row['game.name'] for row in games_rows}

        SCRAPE_FROM_OVERVIEW = False
        if SCRAPE_FROM_OVERVIEW:
            yield from games_rows

        idx = 0
        for game_id, game_name in games_ids.items():
            yield from self._scrape_game_trophies(game_id, game_name)
            del game_id
            idx += 1
            if idx >= MAX_NUMBER_GAMES_TO_PARSE:
                break

    def _setup_cache(self):
        requests_util.setup_limiter(
            self.session,
            URL_API_ROOT,
            per_minute=5,
        )

    def _scrape_games_overview(self) -> Iterator[dict]:
        # Request to get overview
        logger.info('Getting Overview')
        url = URL_PROFILE.format(psn_id=secrets.PLAYSTATION_PSN_ID)
        response = self.session.get(url)
        response.raise_for_status()

        NOW = parse_util.parse_response_datetime(response)

        # Parse data
        soup = bs4.BeautifulSoup(response.content, 'lxml')
        soup = personal_data.html_util.normalize_soup_slightly(soup, classes=False)

        # Recent trophies.
        soup_recent_tropies = soup.select('ul#recent-trophies > li')
        assert len(soup_recent_tropies) > 0, url
        for row in soup_recent_tropies:
            cells = row.select_one('.info .box td').find_all('div')

            trophy_name = cells[0].get_text().strip()
            trophy_desc = cells[1].get_text().strip()
            game_name = cells[2].a.extract().get_text().strip()

            psnprofiles_id = game_psnprofiles_id_from_url(cells[0].find('a')['href'])
            trophy_icon = row.find(class_='icon').find('img')['src']

            gotten_at = (
                cells[2].get_text().strip().removesuffix(' in').removesuffix(' ago')
            )
            gotten_at = parse_util.parse_duration(gotten_at)
            time_acquired = NOW - gotten_at

            yield {
                'game.name': game_name,
                'me.last_played_time': time_acquired.date(),
                # Trophy Data
                'trophy.name': trophy_name,
                'trophy.desc': trophy_desc,
                'trophy.icon': trophy_icon,
                'psnprofiles.game_id': psnprofiles_id,
            }

            del row, cells, time_acquired

        # Games table
        table_rows = soup.find(id='gamesTable').find_all('tr')
        assert len(table_rows) > 0, url

        for row in table_rows:
            cells = row.find_all('td')

            # Check for pagination
            if re.match(
                r'show \d+ more games',
                cells[0].get_text().strip(),
                re.IGNORECASE,
            ):
                break

            game_name = cells[1].find(class_='title').get_text()
            psnprofiles_id = game_psnprofiles_id_from_url(cells[0].find('a')['href'])
            game_icon = cells[0].find('img')['src']

            game_name = row.select_one('.title').get_text()
            game_platform = row.select_one('.platform').get_text()

            small_infos = cells[1].find_all('div')
            if len(small_infos) > 2:
                time_played_div = small_infos[2]
                time_played_div.sup.extract()
                time_played = parse_util.parse_date(
                    time_played_div.get_text(),
                )
            else:
                time_played = None

            d = {
                # Important fields
                'game.name': game_name,
                # Secondary fields
                'game.platform': game_platform,
                'game.icon': game_icon,
                'psnprofiles.game_id': psnprofiles_id,
            }
            if time_played:
                d['me.last_played_time'] = time_played
            yield d

    def _scrape_game_trophies(
        self,
        psnprofiles_id: int,
        game_name: str,
    ) -> Iterator[dict]:
        assert isinstance(psnprofiles_id, int), psnprofiles_id
        assert isinstance(game_name, str), game_name

        logger.info('Getting Game Trophies %s', psnprofiles_id)

        url = URL_USER_GAME_TROPHIES.format(
            psn_id=secrets.PLAYSTATION_PSN_ID,
            game_id=psnprofiles_id,
        )
        response = self.session.get(url)
        response.raise_for_status()

        # Parse data
        soup = bs4.BeautifulSoup(response.content, 'lxml')
        soup = personal_data.html_util.normalize_soup_slightly(soup, classes=False)

        # Remove redundant
        for redundant in soup.select('.wide-ad'):
            redundant.extract()
        for redundant in soup.select('div.col-xs-4'):
            redundant.extract()

        # Recent trophies.
        soup_tropies = soup.select(
            '#content.page > .row > div.col-xs div.box table.zebra tr.completed',
        )
        for row in soup_tropies:
            cells = row.find_all('td')

            trophy_name_a = cells[1].a
            if trophy_name_a is None:
                continue
            trophy_name = trophy_name_a.get_text().strip()
            trophy_name_a.extract()
            trophy_desc = cells[1].get_text().strip()

            trophy_icon = cells[0].img['src']

            cells[2].span.span.nobr.sup.extract()
            gotten_at = parse_util.parse_time(cells[2].get_text())

            yield {
                'game.name': game_name,
                'me.last_played_time': gotten_at,
                # Trophy Data
                'trophy.name': trophy_name,
                'trophy.desc': trophy_desc,
                'trophy.icon': trophy_icon,
                'psnprofiles.game_id': psnprofiles_id,
            }

            del row, cells