forked from Apollo1840/Happy-Crawler
-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathdouban_crawler.py
203 lines (146 loc) · 6.75 KB
/
douban_crawler.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
# -*- coding: utf-8 -*-
'''
todo: add for loop， track the change of some specific words
'''
import jieba
import jieba.posseg
# "n"是名词，“a”是形容词，“v”是动词，“d”是副词，“x”是非语素词
# https://blog.csdn.net/suibianshen2012/article/details/53487157
import requests as req
from bs4 import BeautifulSoup
from basic_crawler import basic_crawler
import numpy as np
import pandas as pd
from pyecharts import WordCloud
import time
class douban_crawler(basic_crawler):
def __init__(self):
url = 'https://www.douban.com/group/blabla//discussion?start=0'
super(douban_crawler, self).__init__(url)
self.temp_soup = None
self.post_titles = []
self.historical_post_titles = []
self.post_heat = []
self.words = [] # each item i, i[0] is word, i[1] is flag, i[2] is heat
self.words_refine = []
self.list_of_drop_words = ['人',
'人物',
'大家',
'帖子',
'时候',
'鹅',
'鹅们',
'全国',
'图',
'料',
'集',
'楼']
def get_posts(self, include_heat=False):
'''
this function will enumerate the post titles of temp_soup
and store it in self.post_titles with modification mode.
if include_heat, it will also store the number of responses in self.post_heat.
'''
self.post_titles = []
self.post_heat = []
tbody = self.temp_soup.find('table', class_='olt')
trs = tbody.find_all('tr')
for tr in trs[1:]:
title = tr.td.a['title']
self.post_titles.append(title)
if include_heat:
heat = tr.find_all('td')[2].text
self.post_heat.append(heat)
def get_words(self, include_heat=False, tags_off=False):
self.get_posts(include_heat)
if tags_off:
# tags_off can get preciser result
for p_title in self.post_titles:
words_in_p_title = jieba.cut(p_title, cut_all=False, HMM=True)
self.words_refine.extend(words_in_p_title)
else:
for i in range(len(self.post_titles)):
if self.post_titles[i] not in self.historical_post_titles:
words_in_p_title = jieba.posseg.cut(self.post_titles[i], HMM=True)
for w in words_in_p_title:
if include_heat:
self.words.append([w.word, w.flag, self.post_heat[i]])
else:
self.words.append([w.word, w.flag, 1])
self.historical_post_titles.extend(self.post_titles)
def get_words_list(self, num_pages=5, include_heat=True):
self.historical_post_titles = []
for id_page in range(num_pages):
time.sleep(1)
url = 'https://www.douban.com/group/blabla//discussion?start={}'.format(num_pages-id_page)
c = basic_crawler(url)
self.temp_soup = c.soup
del c
self.get_words(include_heat)
print(id_page)
# print(self.historical_post_titles)
def create_words_table(self, adjustment=None, get_raw_data=False):
# preprocess the raw data
df_wl = pd.DataFrame(self.words)
df_wl.columns = ['words','flag','heat']
df_wl.heat = df_wl.heat.astype('float')
if get_raw_data:
df_wl.to_csv('material/raw_data.csv')
if adjustment=='log':
df_wl.heat = np.log(df_wl.heat)
# df_wl.to_csv('raw_data2.csv')
df_wlh = df_wl.groupby(['words','flag'])['heat'].agg([np.sum])
df_wlh.reset_index(inplace=True)
df_wlh.columns = ['words','flag','heat']
# df_wlh.to_csv('raw_data3.csv')
return df_wlh
def run(self, num_pages=5, consider_flags=['n','nr','nrt'], include_heat=True, adjustment=None):
# "n"是名词，“a”是形容词，“v”是动词，“d”是副词，“x”是非语素词
# https://blog.csdn.net/suibianshen2012/article/details/53487157
self. get_words_list(num_pages, include_heat)
df_wl = self.create_words_table(adjustment)
df_part = df_wl.loc[df_wl.flag.isin(consider_flags),:]
df_part = df_part.loc[~df_part.words.isin(self.list_of_drop_words),:]
print(df_part.shape[0])
graph_title = "The hot words in douban/blabla within {} pages at {}".format(num_pages, time.asctime( time.localtime(time.time())))
wordcloud_2 = WordCloud(title=graph_title, title_pos='center', width=1600, height=800)
wordcloud_2.add("", list(df_part.words), df_part.heat, word_size_range=[10, 50],
shape='diamond')
wordcloud_2.render("material/Hot_words_cloud.html")
def test():
'''
this part of program is just playing ground of the programmer
'''
url = 'https://www.douban.com/group/blabla//discussion?start=0'
soup=BeautifulSoup(req.get(url).text, 'lxml')
post_titles = []
post_heat = []
words_h = []
tbody = soup.find('table', class_='olt')
trs = tbody.find_all('tr')
for tr in trs[1:]:
post_titles.append(tr.td.a['title'])
post_heat.append(tr.find_all('td')[2].text)
for i in range(len(post_titles)):
words_in_p_title = jieba.posseg.cut(post_titles[i], HMM=True)
for w in words_in_p_title:
words_h.append([w.word, w.flag, post_heat[i]])
df_wl = pd.DataFrame(words_h)
print(df_wl)
df_wl.columns = ['words','flag','heat']
df_wl.heat.astype('int',inplace=True)
max(df_wl.heat)
df_wl.loc[df_wl.heat==max(df_wl.heat),:]
df_wlh = df_wl.groupby(['words','flag'])['heat'].agg([np.sum])
print(df_wlh)
df_wlh.reset_index(inplace=True)
print(df_wlh)
print(df_wlh.shape)
df_wlh.columns = ['words','flag','heat']
pd.set_option('display.max_rows', None)
df=pd.read_csv('raw_data.csv')
print(df.loc[df.flag.isin(['n','nr','nrt']),:].sort_values(['heat','words']))
print(df.loc[df.flag.isin(['n','nr','nrt']),:].sort_values(['words']))
if __name__ == '__main__':
dc = douban_crawler()
dc.run(num_pages=10, adjustment='normal')