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
|
#!/usr/bin/env python3
import argparse
import hashlib
import imghdr
import os
import pickle
import posixpath
import re
import signal
import socket
import threading
import time
import urllib.parse
import urllib.request
# config
output_dir = './bing' # default output dir
socket.setdefaulttimeout(2)
tried_urls = []
image_md5s = {}
in_progress = 0
urlopenheader = {'User-Agent': 'Mozilla/5.0 (X11; Fedora; Linux x86_64; rv:94.0) Gecko/20100101 Firefox/94.0'}
def download(pool_sema: threading.Semaphore, img_sema: threading.Semaphore, url: str, output_dir: str, limit: int):
global in_progress
if url in tried_urls:
print('SKIP: Already checked url, skipping')
return
pool_sema.acquire()
in_progress += 1
acquired_img_sema = False
path = urllib.parse.urlsplit(url).path
filename = posixpath.basename(path).split('?')[0] # Strip GET parameters from filename
name, ext = os.path.splitext(filename)
name = name[:36].strip()
filename = name + ext
try:
request = urllib.request.Request(url, None, urlopenheader)
image = urllib.request.urlopen(request).read()
if not imghdr.what(None, image):
print('SKIP: Invalid image, not saving ' + filename)
return
md5_key = hashlib.md5(image).hexdigest()
if md5_key in image_md5s:
print('SKIP: Image is a duplicate of ' + image_md5s[md5_key] + ', not saving ' + filename)
return
i = 0
while os.path.exists(os.path.join(output_dir, filename)):
if hashlib.md5(open(os.path.join(output_dir, filename), 'rb').read()).hexdigest() == md5_key:
print('SKIP: Already downloaded ' + filename + ', not saving')
return
i += 1
filename = "%s-%d%s" % (name, i, ext)
image_md5s[md5_key] = filename
img_sema.acquire()
acquired_img_sema = True
if limit is not None and len(tried_urls) >= limit:
return
imagefile = open(os.path.join(output_dir, filename), 'wb')
imagefile.write(image)
imagefile.close()
print(" OK : " + filename)
tried_urls.append(url)
except Exception as e:
print("FAIL: " + filename)
finally:
pool_sema.release()
if acquired_img_sema:
img_sema.release()
in_progress -= 1
def fetch_images_from_keyword(pool_sema: threading.Semaphore, img_sema: threading.Semaphore, keyword: str,
output_dir: str, filters: str, limit: int):
current = 0
last = ''
while True:
time.sleep(0.5)
if in_progress > 10:
continue
request_url = 'https://www.bing.com/images/async?q=' + urllib.parse.quote_plus(keyword) + '&first=' + str(
current) + '&count=35&qft=' + ('' if filters is None else filters)
request = urllib.request.Request(request_url, None, headers=urlopenheader)
response = urllib.request.urlopen(request)
html = response.read().decode('utf8')
links = re.findall('murl":"(.*?)"', html)
try:
if links[-1] == last:
return
for index, link in enumerate(links):
if limit is not None and len(tried_urls) >= limit:
exit(0)
t = threading.Thread(target=download, args=(pool_sema, img_sema, link, output_dir, limit))
t.start()
current += 1
last = links[-1]
except IndexError:
print('FAIL: No search results for "{0}"'.format(keyword))
return
def backup_history(*args):
download_history = open(os.path.join(output_dir, 'download_history.pickle'), 'wb')
pickle.dump(tried_urls, download_history)
copied_image_md5s = dict(
image_md5s) # We are working with the copy, because length of input variable for pickle must not be changed during dumping
pickle.dump(copied_image_md5s, download_history)
download_history.close()
print('history_dumped')
if args:
exit(0)
if __name__ == "__main__":
parser = argparse.ArgumentParser(description='Bing image bulk downloader')
parser.add_argument('-s', '--search-string', help='Keyword to search', required=False)
parser.add_argument('-f', '--search-file', help='Path to a file containing search strings line by line',
required=False)
parser.add_argument('-o', '--output', help='Output directory', required=False)
parser.add_argument('--adult-filter-off', help='Disable adult filter', action='store_true', required=False)
parser.add_argument('--filters',
help='Any query based filters you want to append when searching for images, e.g. +filterui:license-L1',
required=False)
parser.add_argument('--limit', help='Make sure not to search for more than specified amount of images.',
required=False, type=int)
parser.add_argument('--threads', help='Number of threads', type=int, default=20)
args = parser.parse_args()
if (not args.search_string) and (not args.search_file):
parser.error('Provide Either search string or path to file containing search strings')
if args.output:
output_dir = args.output
if not os.path.exists(output_dir):
os.makedirs(output_dir)
output_dir_origin = output_dir
signal.signal(signal.SIGINT, backup_history)
try:
download_history = open(os.path.join(output_dir, 'download_history.pickle'), 'rb')
tried_urls = pickle.load(download_history)
image_md5s = pickle.load(download_history)
download_history.close()
except (OSError, IOError):
tried_urls = []
if args.adult_filter_off:
urlopenheader['Cookie'] = 'SRCHHPGUSR=ADLT=OFF'
pool_sema = threading.BoundedSemaphore(args.threads)
img_sema = threading.Semaphore()
if args.search_string:
fetch_images_from_keyword(pool_sema, img_sema, args.search_string, output_dir, args.filters, args.limit)
elif args.search_file:
try:
inputFile = open(args.search_file)
except (OSError, IOError):
print("FAIL: Couldn't open file {}".format(args.search_file))
exit(1)
for keyword in inputFile.readlines():
output_sub_dir = os.path.join(output_dir_origin, keyword.strip().replace(' ', '_'))
if not os.path.exists(output_sub_dir):
os.makedirs(output_sub_dir)
fetch_images_from_keyword(pool_sema, img_sema, keyword, output_sub_dir, args.filters, args.limit)
backup_history()
time.sleep(10)
inputFile.close()
|
