去年書いたスクリプト↓がだいたいそのまま動いたのでアドベントカレンダーのはてなブックマークの合計数のランキングと記事別のランキングを今年も出してみた
sucrose.hatenablog.com
全件記事を見たわけではないので変なゴミが混じっているかも
今年はディープラーニングの記事がたくさんはてブを集めていて人気を感じる
↓最終結果も出した
カレンダーの合計ランキング
記事別ランキング
アドベントカレンダー一覧ページも含む
ソースコード
# -*- coding: utf-8 -*- import pyquery import time import requests banned = set(['http://www.4gamer.net/games/074/G007497/20120713064/', 'http://horiemon.com/talk/11346/']) def getQiitaCalendarList(year, page): calendar_list = pyquery.PyQuery(url='http://qiita.com/advent-calendar/{}/calendars?page={}'.format(year, page)) calendars = set() for elm in calendar_list.find('.adventCalendarList_calendarTitle > a'): a = pyquery.PyQuery(elm) href = a.attr('href') calendars.add((href[22:], a.text())) return calendars def getQiitaArticles(name): calendar = pyquery.PyQuery(url='http://qiita.com/advent-calendar/2016/{}'.format(name)) article = set() for elm in calendar.find('.adventCalendarItem_entry > a'): a = pyquery.PyQuery(elm) url = a.attr('href') if 'http' in url and url not in banned: article.add(url) return article def getAdventarCalendarList(year): calendar_list = pyquery.PyQuery(url='http://www.adventar.org/calendars?year={}'.format(year)) urls = set() for elm in calendar_list.find('.mod-calendarList-title > a'): a = pyquery.PyQuery(elm) href = 'http://www.adventar.org' + a.attr('href') urls.add((href, a.text())) return urls def getAdventarArticles(url): calendar = pyquery.PyQuery(url=url) article = set() for elm in calendar.find('.mod-entryList-url > a'): a = pyquery.PyQuery(elm) if 'http' in a.text() and a.attr('href') not in banned: url = a.attr('href') article.add(url) return article def getHatenaBookmarkCount(urls): assert len(urls) <= 50 try: return requests.get('http://api.b.st-hatena.com/entry.counts', params={'url': urls}).json() except: time.sleep(2) return requests.get('http://api.b.st-hatena.com/entry.counts', params={'url': urls}).json() if __name__ == '__main__': result = [] result_article = [] calendars = set() for i in range(1, 27): calendars |= getQiitaCalendarList(2016, i) time.sleep(1) for name, title in calendars: articles = getQiitaArticles(name) calendar_url = 'http://qiita.com/advent-calendar/2016/' + name urls = [calendar_url] + list(articles) hatebu_count = getHatenaBookmarkCount(urls) result.append((sum(hatebu_count.values()), calendar_url, title)) result_article.extend([(url, count, calendar_url, title) for url, count in hatebu_count.items()]) time.sleep(1) calendars = getAdventarCalendarList(2016) time.sleep(1) for calendar_url, title in calendars: articles = getAdventarArticles(calendar_url) urls = [calendar_url] + list(articles) hatebu_count = getHatenaBookmarkCount(urls) result.append((sum(hatebu_count.values()), calendar_url, title)) result_article.extend([(url, count, calendar_url, title) for url, count in hatebu_count.items()]) time.sleep(1) result.sort(reverse=True) print u'|*順位|*カレンダー名|*はてなブックマーク数|'.encode('utf-8') for i, (count, name, title) in enumerate(result[:100], 1): print u'|{0}|<a href="{1}">{2}</a>|{3}|'.format(i, name, title, count).encode('utf-8') result_article.sort(key=lambda x: -x[1]) print print u'|*順位|*記事名|*はてなブックマーク数|*カレンダー名|'.encode('utf-8') for i, (url, count, calendar_url, title) in enumerate(result_article[:100], 1): print u'|{0}|<a href="{1}">{2}</a>|{3}|<a href="{4}">{5}</a>|'.format(i, url, pyquery.PyQuery(url=url).find('title').text().replace('|', '|'), count, calendar_url, title).encode('utf-8') time.sleep(1)
