/
web_scrape.py
198 lines (162 loc) · 6.35 KB
/
web_scrape.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
198
import requests
import bs4
import pandas as pd
from pandas import Series, DataFrame
import numpy as np
from datetime import datetime
import sys
# FUNCTIONS
def download_website(website, file_name):
page = requests.get(website)
try:
page.raise_for_status()
except Exception as exc:
print('There was a problem: %s' % (exc))
file_w = open(file_name, 'wb')
for chunk in page.iter_content(100000):
file_w.write(chunk)
file_w.close()
def naming_links(prefix, num_files):
num = range(1, num_files + 1)
file_names = [prefix + '_' + str(c) + '.txt' for c in num]
print file_names
return file_names
def site_iterator(websites, file_names):
for i, j in zip(websites, file_names):
download_website(i, j)
print 'Created ', j
def unique_links(link):
output = []
for x in link:
if x not in output:
output.append(x)
return output
def download_iterator(websites, prefix, num_files):
file_names = naming_links(prefix, num_files)
site_iterator(websites, file_names)
print 'Downloaded websites'
# Scraping web links
def scrape_links(link):
response = requests.get(link)
soup = bs4.BeautifulSoup(response.text)
links = [a.attrs.get('href') for a in soup.select('#Body a[href^=http://www.spine-health.com/forum/discussion]')]
print links
print len(links)
return links
# Find unique links
def scrape_unique_links(links):
links_unq = unique_links(links)
print len(links_unq)
matching = [s for s in links_unq if "#latest" in s]
print len(matching)
# download_iterator(matching, 'lbp', len(links_unq))
return matching
# Scraping page
def scrape_element(soup, prefix, element):
msgs = soup.select(element)
length = len(msgs)
print length
n = range(0, length)
print n
element_dict = {}
for i in n:
element_dict[prefix + str(i)] = soup.select(element)[i].get_text()
return element_dict
# Metadata
# forum = scrape_element('forum', '.MItem.Category')
def title_data(soup):
title = str(soup.find('div', class_='PageTitle').h1)
title_text = title.replace('<h1>', '').replace('</h1>', '')
df_title1 = DataFrame(pd.Series(title_text))
df_title2 = DataFrame(pd.Series(len(title_text)))
df_title = pd.concat([df_title1,df_title2],axis=1)
df_title.columns = ['title','title_length']
return df_title
# Data
# Distributions
# Length per post
def messages_data(soup,message_csv):
messages = scrape_element(soup, 'messages', '.Message')
msg_lengths = []
pd.set_option('display.max_colwidth', -1)
for k, v in messages.items():
msg_lengths.append(len(v))
text = Series(str(np.array(v.encode('utf-8'))))
print text
text.to_csv(message_csv, sep=',', header=False, index=False, mode='a')
df_msg_lgth = DataFrame(msg_lengths)
df_msg_describe = DataFrame(df_msg_lgth.describe()).T
cols = df_msg_describe.columns
df_msg_describe.columns = ['msg_' + c for c in cols]
return df_msg_describe
# Author post history
def posts_cnt_data(soup):
posts_cnt = scrape_element(soup, 'posts_cnt', '.MItem.PostCount')
posts_cnts = []
for k, v in posts_cnt.items():
posts_cnts.append(v)
print posts_cnts
df_posts_cnts = pd.DataFrame(posts_cnts)
df_strip = df_posts_cnts[0].apply(lambda x: int(x.strip('Posts: ').replace(',', '')))
df_strip_describe = DataFrame(df_strip.describe()).T
cols = df_strip_describe.columns
df_strip_describe.columns = ['postshistory_' + c for c in cols]
return df_strip_describe
# Post created dates
def create_date_diff(df_create, var, prefix):
df_create_describe = DataFrame(df_create[[var]].describe()).T
cols = df_create_describe.columns
df_create_describe.columns = [prefix + c for c in cols]
return df_create_describe
def posts_create_data(soup):
created = scrape_element(soup, 'created', '.MItem.DateCreated')
create_dates = []
for k, v in created.items():
create_dates.append(v)
print create_dates
df_create_dates = pd.DataFrame(create_dates)
df_create_dates[1] = df_create_dates[0].apply(lambda x: x.replace('\n', '').replace('Today',datetime.today().strftime("%m/%d/%Y")))
df_create_dates['date'] = df_create_dates[1].apply(lambda x: pd.to_datetime(x))
date_size = len(df_create_dates)
df_next = df_create_dates['date'].ix[1:date_size - 1].reset_index()
df_next.columns = ['index', 'next_date']
df_now_next = DataFrame(df_create_dates['date']).join(DataFrame(df_next['next_date']))
df_now_next['diff'] = (df_now_next['next_date'] - df_now_next['date']) / np.timedelta64(1, 'D')
df_date = create_date_diff(df_now_next, 'date', 'create_').reset_index().drop('index', 1)
df_diff = create_date_diff(df_now_next, 'diff', 'datediff_').reset_index().drop('index', 1).astype(float)
dfs = pd.concat([df_date, df_diff], axis=1)
return dfs
def combine_data(links, title, author, posts, create):
all_data = pd.concat([links, title, author, posts, create], axis=1)
print all_data.shape
return all_data
# updated = scrape_element('updated', '.DateUpdated')
# author = scrape_element('author', '.Author')
def main(f_link, message_csv, stats_csv):
forum_link = f_link
web_links = scrape_links(forum_link)
matching = scrape_unique_links(web_links)
i = 1
for m in matching:
print m, '\n', i, '\n', datetime.now()
response = requests.get(m)
soup = bs4.BeautifulSoup(response.text)
df_links = DataFrame(Series(m))
df_links.columns = ['links']
df_title = title_data(soup)
df_msg = messages_data(soup,message_csv)
df_posts = posts_cnt_data(soup)
df_date = posts_create_data(soup)
all_data = combine_data(df_links, df_title, df_msg, df_posts, df_date)
if i == 1:
all_data.to_csv(stats_csv, delimiter=',', header=True, index=True, mode='w')
else:
all_data.to_csv(stats_csv, delimiter=',', header=False, index=True, mode='a')
i += 1
print 'Done'
# MAIN CODE
if __name__=="__main__":
forumlink = sys.argv[1]
messagecsv = sys.argv[2]
statscsv = sys.argv[3]
main(forumlink, messagecsv, statscsv)