Python获取暗黑破坏神3战网前1000命位玩家的英雄技能统计

yipeiwu_com5年前Python基础

说实在的个人对游戏并没有多大的兴趣,但唯独对暴雪的Diablo系列很有感情,去年年初开始玩Diablo3,断断续续,感觉最麻烦的是选择技能,每次版本更新可能都有更优的build,这对于我这样的业余玩家来说可不是件好事,好在宏伟秘境后有了天梯,借鉴排名在前的高级玩家们build总没错,于是花了点时间写了这个脚本。

脚本只是统计了主动技能、被动技能和传奇宝石的使用情况,理论上统计其它如装备等信息也是一样简单可行的,但Diablo装备的生成机制使得统计这个没有多大意义,相同的装备属性可能各有优劣,难以比较,而且某些装备坑爹的掉率也不是你想要就能有的。

题外话,不得不说Python太适合写这类功能相对简单的脚本了,一个字:快。

# -*- coding: utf-8 -*-
"""
Diablo3 排名前1000玩家英雄使用技能统计

python diablo.py help
python diablo.py [barbarian|crusader|demon-hunter|monk'|witch-doctor|wizard]

默认使用的是亚服的数据,如果需要美服或欧服,更改`_rank_page`和`_api`变量地址即可

Copyright (c) 2015 JinnLynn <eatfishlin@gmail.com>
Released under the terms of the MIT license.
"""
from __future__ import unicode_literals, print_function, absolute_import
import os
import sys
import urllib2
import json
import re

__version__ = '1.0.0'
__author__ = 'JinnLynn <eatfishlin@gmail.com>'
__license__ = 'The MIT License'
__copyright__ = 'Copyright 2015 JinnLynn'

# 排名页面
_rank_page = 'http://tw.battle.net/d3/zh/rankings/'
# api
_api = 'http://tw.battle.net/api/d3/'
_api_profile = os.path.join(_api, 'profile')
_api_data = os.path.join(_api, 'data')

_hero_classes = {
  'barbarian': '野蠻人', 'crusader': '聖教軍', 'demon-hunter': '狩魔獵人',
  'monk': '武僧', 'witch-doctor': '巫醫', 'wizard': '秘術師'}

_retry = 5

_hero_class = ''
_active_skills = {}
_passive_skills = {}
_unique_gems = {}


def _clear_output(msg=''):
  sys.stdout.write('\r{:30}'.format(' '))
  sys.stdout.write('\r{}'.format(msg))
  sys.stdout.flush()


def _process(stated, total):
  msg = '英雄数据分析中... {}/{}'.format(stated, total)
  _clear_output(msg)


def _get(url, is_json=True):
  # print('GET: ', url)
  retry = 5 if _retry < 1 else _retry
  while retry > 0:
    try:
      req = urllib2.urlopen(url.encode('utf8'), timeout=10)
      return json.load(req) if is_json else req.read()
    except KeyboardInterrupt, e:
      raise e
    except Exception, e:
      retry -= 1
      # print('retry', retry, e)
      # raise e


def _api_url(*args, **kwargs):
  slash = kwargs.get('slash', False)
  args = [unicode(arg) for arg in args]
  url = os.path.join(*args).rstrip('/')
  return url + '/' if slash else url


def get_era():
  req = urllib2.urlopen(_rank_page)
  return req.geturl().split('/')[-2]


def get_rank_page_url(era):
  url_part = 'rift-'
  if _hero_class == 'demon-hunter':
    url_part += 'dh'
  elif _hero_class == 'witch-doctor':
    url_part += 'wd'
  else:
    url_part += _hero_class
  return os.path.join(_rank_page, 'era', era, url_part)


def fetch_rank_list():
  tags = []
  try:
    _clear_output('获取当前游戏纪元...')
    era = get_era()
    _clear_output('获取当前排名前1000的玩家...')
    url = get_rank_page_url(era)
    html = _get(url, is_json=False)
    # re parse
    lst = re.findall(
      r"a href=\"(.*)\" title=.*class=\"icon-profile link-first\">",
      html.decode('utf8'),
      re.UNICODE)
    # BeautifulSoup parse
    # import bs4
    # soup = bs4.BeautifulSoup(html)
    # lst = soup.select('#ladders-table tbody tr .battletag a')['href']
    for item in lst:
      try:
        tags.append(item.split('/')[-2])
      except:
        pass
  except Exception, e:
    print('fetch rank list fail. {}'.format(_rank_page))
    raise e
  return tags


def get_hero(player_tag):
  url = _api_url(_api_profile, player_tag, slash=True)
  data = _get(url)
  hero_selected = None
  for hero in data.get('heroes', []):
    if hero['class'] != _hero_class:
      continue
    last_updated = hero_selected['last-updated']
    # 最近使用的英雄
    if hero_selected is None or last_updated < hero['last-updated']:
      hero_selected = hero
  if not hero_selected:
    raise Exception('{} hero missing.'.format(player_tag))
  url = _api_url(_api_profile, player_tag, 'hero', hero_selected['id'])
  return _get(url)


# 主动技能符文
def stat_active_skill_rune(skill_slug, rune):
  global _active_skills
  if not rune:
    return
  slug = rune.get('slug')
  if slug in _active_skills[skill_slug]['rune']:
    _active_skills[skill_slug]['rune'][slug]['count'] += 1
  else:
    _active_skills[skill_slug]['rune'][slug] = {
      'count': 1,
      'name': rune.get('name')
    }


# 主动技能
def stat_active_skill(active):
  global _active_skills
  slug = active.get('skill', {}).get('slug')
  # d3 API 返回的数据中可能存在空的数据
  if not slug:
    return
  if slug in _active_skills:
    _active_skills[slug]['count'] += 1
  else:
    _active_skills[slug] = {
      'count': 1,
      'name': active.get('skill').get('name'),
      'rune': {}
    }
  stat_active_skill_rune(slug, active.get('rune'))


# 被动技能
def stat_passive_skill(passive):
  global _passive_skills
  slug = passive.get('skill', {}).get('slug')
  # d3 API 返回的数据中可能存在空的数据
  if not slug:
    return
  if slug in _passive_skills:
    _passive_skills[slug]['count'] += 1
  else:
    _passive_skills[slug] = {
      'count': 1,
      'name': passive.get('skill').get('name')
    }


def stat_unique_gem(items):
  global _unique_gems

  def get_gem(tooltip):
    if not tooltip:
      return None, None
    url = _api_url(_api_data, tooltip)
    data = _get(url)
    gems = data.get('gems')
    if not gems:
      return None, None
    gem = gems[0].get('item', {})
    return gem.get('id'), gem.get('name')

  if not items:
    return

  lst = [items.get(s, {}) for s in ['leftFinger', 'rightFinger', 'neck']]
  for tooltip in [d.get('tooltipParams', None) for d in lst]:
    id_, name = get_gem(tooltip)
    if not id_:
      continue
    if id_ in _unique_gems:
      _unique_gems[id_]['count'] += 1
    else:
      _unique_gems[id_] = {
        'count': 1,
        'name': name
      }


def stat(hero):
  global _active_skills, _passive_skills

  map(stat_active_skill, hero.get('skills', {}).get('active', []))
  map(stat_passive_skill, hero.get('skills', {}).get('passive', []))

  items = hero.get('items', {})
  stat_unique_gem(items)


def output(hero_stated, hero_stat_failed):
  def sort(data, count=10):
    d = sorted(data.items(), key=lambda d: d[1]['count'], reverse=True)
    return d if count <= 0 else d[0:count]

  _clear_output()

  # print('======')
  # print(hero_stated, hero_stat_failed)
  # print('======')
  # pprint(_active_skills)
  # print('======')
  # pprint(_passive_skills)
  # print('======')
  # pprint(_unique_gems)
  # pprint(_active_skills.items())
  # print('======')

  print('\n=== RESULT ===\n')
  print('统计英雄数\n')
  print(' 成功: {} 失败: {}\n'.format(hero_stated, hero_stat_failed))

  print('主动技能使用排名: ')
  for _, d in sort(_active_skills):
    runes = []
    for _, r in sort(d.get('rune', {})):
      runes.append('{name}[{count}]'.format(**r))
    d.update({'rune_rank': ', '.join(runes)})
    print(' {name}[{count}]: {rune_rank}'.format(**d))
  print()

  print('被动技能使用排名: ')
  for _, d in sort(_passive_skills):
    print(' {name}[{count}]'.format(**d))
  print()

  print('传奇宝石使用排名: ')
  for _, d in sort(_unique_gems):
    print(' {name}[{count}]'.format(**d))
  print()


def prepare():
  global _hero_class

  def print_hc():
    print('仅支持以下英雄类型, 默认 demon-hunter:\n')
    for c, n in _hero_classes.items():
      print(c, ':', n)

  if len(sys.argv) == 1:
    _hero_class = 'demon-hunter'
  elif len(sys.argv) > 2:
    sys.exit('参数错误')
  else:
    arg = sys.argv[1]
    if arg == 'help':
      print_hc()
      print('\nTips: 运行中可随时Ctrl+C终止以获得已统计的数据结果')
      sys.exit()
    elif arg not in _hero_classes:
      print_hc()
      sys.exit()
    else:
      _hero_class = arg


def main():
  prepare()
  print('待分析的英雄类型:', _hero_classes[_hero_class])

  hero_stated = 0
  hero_stat_failed = 0
  try:
    tags = fetch_rank_list()
    if not tags:
      raise Exception('parse battle.net rank page fail.')
  except Exception, e:
    print('error,', e)
    sys.exit()

  total = len(tags)

  for tag in tags:
    try:
      hero = get_hero(tag)
      if not hero:
        raise Exception('no hero data')
      stat(hero)
      hero_stated += 1
      _process(hero_stated, total)
    except KeyboardInterrupt:
      break
    except Exception, e:
      # print('Fail: ', tag, e, hero)
      hero_stat_failed += 1

  output(hero_stated, hero_stat_failed)


if __name__ == '__main__':
  main()

相关文章

遗传算法之Python实现代码

写在前面 之前的文章中已经讲过了遗传算法的基本流程,并且用MATLAB实现过一遍了。这一篇文章主要面对的人群是看过了我之前的文章,因此我就不再赘述遗传算法是什么以及基本的内容了,假设大家...

python生成不重复随机数和对list乱序的解决方法

andom.sample(list, n)即是从list中随机选取n个不同的元素 # -*- coding: utf-8 -*- import random # 从一个list中...

Python求解任意闭区间的所有素数

题目:请求出任意区间[a,b]的所有素数,简单考虑实用性 这道题看起来应该很easy是吧,但任意区间(这个问题有没get 到) Afanty的分析: 1、首先明白什么叫素数,注意用求余法...

Python切片用法实例教程

本文以实例形式讲述了Python中切片操作的用法,分享给大家供大家参考借鉴,具体如下: 取一个list或tuple的部分元素是非常常见的操作。比如,一个list如下: >>...

python以环状形式组合排列图片并输出的方法

本文实例讲述了python以环状形式组合排列图片并输出的方法。分享给大家供大家参考。具体分析如下: 这段代码可以自定义一个空白画板,然后将指定的图片以圆环状的方式排列起来,用到了pil库...