import requests
from bs4 import BeautifulSoup
from pymongo import MongoClient

class QuNaEr():
  def __init__(self, keyword, page=1):
    self.keyword = keyword
    self.page = page

  def qne_spider(self):
    url = 'https://piao.qunar.com/ticket/list.htm?keyword=%s&region=&from=mpl_search_suggest&page=%s' % (self.keyword, self.page)
    response = requests.get(url)
    response.encoding = 'utf-8'
    text = response.text
    bs_obj = BeautifulSoup(text, 'html.parser')

    arr = bs_obj.find('div', {'class': 'result_list'}).contents
    for i in arr:
      info = i.attrs
      # 景區(qū)名稱
      name = info.get('data-sight-name')
      # 地址
      address = info.get('data-address')
      # 近期售票數(shù)
      count = info.get('data-sale-count')
      # 經(jīng)緯度
      point = info.get('data-point')

      # 起始價(jià)格
      price = i.find('span', {'class': 'sight_item_price'})
      price = price.find_all('em')
      price = price[0].text

      conn = MongoClient('localhost', port=27017)
      db = conn.QuNaEr # 庫
      table = db.qunaer_51 # 表

      table.insert_one({
        'name'   :  name,
        'address'  :  address,
        'count'   :  int(count),
        'point'   :  point,
        'price'   :  float(price),
        'city'   :  self.keyword
      })

if __name__ == '__main__':
  citys = ['北京', '上海', '成都', '三亞', '廣州', '重慶', '深圳', '西安', '杭州', '廈門', '武漢', '大連', '蘇州']
  for i in citys:
    for page in range(1, 5):
      qne = QuNaEr(i, page=page)
      qne.qne_spider()

from pymongo import MongoClient
# 設(shè)置字體，不然無法顯示中文
from pylab import *

mpl.rcParams['font.sans-serif'] = ['SimHei']

conn = MongoClient('localhost', port=27017)
db = conn.QuNaEr # 庫
table = db.qunaer_51 # 表

result = table.find().sort([('count', -1)]).limit(15)
# x,y軸數(shù)據(jù)
x_arr = [] # 景區(qū)名稱
y_arr = [] # 銷量
for i in result:
  x_arr.append(i['name'])
  y_arr.append(i['count'])

"""
去哪兒月銷量排行榜
"""
plt.bar(x_arr, y_arr, color='rgb') # 指定color，不然所有的柱體都會是一個(gè)顏色
plt.gcf().autofmt_xdate() # 旋轉(zhuǎn)x軸，避免重疊
plt.xlabel(u'景點(diǎn)名稱') # x軸描述信息
plt.ylabel(u'月銷量') # y軸描述信息
plt.title(u'拉鉤景點(diǎn)月銷量統(tǒng)計(jì)表') # 指定圖表描述信息
plt.ylim(0, 4000) # 指定Y軸的高度
plt.savefig('去哪兒月銷售量排行榜') # 保存為圖片
plt.show()

arr = [[0, 50], [50,100], [100, 200], [200,300], [300,500], [500,1000]]
name_arr = []
total_arr = []
for i in arr:
  result = table.count({'price': {'$gte': i[0], '$lt': i[1]}})
  name = '%s元 ~ %s元 ' % (i[0], i[1])
  name_arr.append(name)
  total_arr.append(result)

color = 'red', 'orange', 'green', 'blue', 'gray', 'goldenrod' # 各類別顏色
explode = (0.2, 0, 0, 0, 0, 0) # 各類別的偏移半徑

# 繪制餅狀圖
pie = plt.pie(total_arr, colors=color, explode=explode, labels=name_arr, shadow=True, autopct='%1.1f%%')

plt.axis('equal')
plt.title(u'熱點(diǎn)旅游景區(qū)門票價(jià)格比例', fontsize=12)

plt.legend(loc=0, bbox_to_anchor=(0.82, 1)) # 圖例
# 設(shè)置legend的字體大小
leg = plt.gca().get_legend()
ltext = leg.get_texts()
plt.setp(ltext, fontsize=6)
# 顯示圖
plt.show()