练习： bs4 简单爬取 + matplotlib 折线图显示 （关键词，职位数量、起薪）

# -*- coding: utf-8 -*-
# coding=utf-8


import requests
import bs4
import urllib

def getPageTxt(url):
    headers = {
        'User-Agent': 'Mozilla/5.0 (Macintosh; Intel Mac OS X 10_12_3) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/56.0.2924.87 Safari/537.36'}
    r = requests.get(url, headers=headers)
    soup = bs4.BeautifulSoup(r.content.decode("utf-8"), "lxml")
    mylist = soup.select('.search_yx_tj em')
    counts = mylist[0].text
    return counts


def getUrl(word, moneymin=0, moneymax=99999):
    url = "http://sou.zhaopin.com/jobs/searchresult.ashx?jl=北京&kw="
    url += word + "&p=1&isadv=0&isfilter=1&sf=" + str(moneymin) + "&st=" + str(moneymax)
    return url


# 一、查询关键词
word000 = "前端,python" #限最多6个值，中间半角逗号分隔
#word000 = "前端,python,php"
words = word000.split(",")


# 二、工资区间（最小值为0，最大值为99999）
moneys = [0, 6000, 12000, 20000, 30000, 99999]
money_min = list(map(lambda x: x+1, moneys[:-1]))    #[1, 6001, 12001, 20001, 30001]
money_max = list(map(lambda x: x, moneys[1:]))        #[6000, 12000, 20000, 30000, 99999]
print(money_min)
print(money_max)


# 三、url拼串，扒取页面，生成字典文件
'''
字典文件的结构为：（counts是职位数量，startmoney是起薪。这两个list，后面将用于生成折线图）
[
    {'key': '前端', 'counts': [1105, 4485, 4394, 1549, 393], 'startmoney': [1, 6001, 12001, 20001, 30001]},
    {'key': 'python', 'counts': [700, 2945, 4467, 2454, 984], 'startmoney': [1, 6001, 12001, 20001, 30001]}
]

'''
ulist = []
for word in words:
    li = {}
    counts = []
    for k in money_min:
        idx = money_min.index(k)  # 索引
        url = getUrl(word, money_min[idx], money_max[idx])    # 拼串
        counts.append(int(getPageTxt(url)))  # 扒取    ，并转换为数字，存入counts列表
    li["key"] = word
    li["counts"] = counts
    li["startmoney"] = money_min
    ulist.append(li)
print(ulist)
f = open("counts000.txt", "w")
f.write(str(ulist))
f.close()

# -*- coding: utf-8 -*-
# coding=utf-8


import json
f = open('counts000.txt', 'r+')
ulliststring = f.read()
f.close()

ulliststring = ulliststring.replace("'", '"')  # 注意这里有个坑！！  json的标准格式：要求必须 只能使用双引号作为键 或者 值的边界符号，不能使用单引号，而且“键”必须使用边界符（双引号）
ullist = json.loads(ulliststring)



import matplotlib
import matplotlib.pyplot as plt
matplotlib.rcParams['font.sans-serif'] = ['SimHei']# 为了能显示中文（而不是显示一个框）
matplotlib.rcParams['font.family']='sans-serif'
matplotlib.rcParams['axes.unicode_minus'] = False   # 为了能显示负号（而不是显示一个框）


colornames = ['#ff0000', '#ccee00', '#00aa00','#0000ff', '#9900ff', '#ff0099']

plt.title('bj薪资、职位数量折线表', color='#ff33a0')
plt.xlabel('薪资区间')  # 设置X坐标轴标题
plt.ylabel('职位数量')  # 设置Y坐标轴表
plt.xlim(0, 35000)  # 设置坐标轴的范围
plt.ylim(0, 5000)
# plt.xticks([2,4])#设置x轴的标签间隔
# plt.yticks([4,16])#设置y轴的标签间隔

i = 0
for li in ullist:
    x = li["startmoney"]
    y = li["counts"]
    plt.plot(x, y, colornames[i])
    i += 1
plt.show()