CIRCL
/
AIL-framework
mirror of https://github.com/CIRCL/AIL-framework
2
1
Fork 0
Code Issues Releases Wiki Activity

Fix crawler rotation 

Before this, crawler processed prioritized onions and after all starts prioritized regular.
pull/350/head
kovacsbalu 2019-05-15 09:57:18 +02:00
parent 0389b9c23b
commit 6092f482e6
1 changed files with 3 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
bin/Crawler.py
View File

@ -12,6 +12,7 @@ import time
import subprocess import subprocess
import requests import requests
from collections import deque
from pyfaup.faup import Faup from pyfaup.faup import Faup
sys.path.append(os.environ['AIL_BIN']) sys.path.append(os.environ['AIL_BIN'])
@ -303,7 +304,7 @@ if __name__ == '__main__':
#mode = sys.argv[1] #mode = sys.argv[1]
splash_port = sys.argv[1] splash_port = sys.argv[1]
rotation_mode = ['onion', 'regular'] rotation_mode = deque(["onion", "regular"])
default_proto_map = {'http': 80, 'https': 443} default_proto_map = {'http': 80, 'https': 443}
######################################################## add ftp ??? ######################################################## add ftp ???
@ -361,6 +362,7 @@ if __name__ == '__main__':
update_auto_crawler() update_auto_crawler()
rotation_mode.rotate()
to_crawl = get_elem_to_crawl(rotation_mode) to_crawl = get_elem_to_crawl(rotation_mode)
if to_crawl: if to_crawl:
url_data = unpack_url(to_crawl['url']) url_data = unpack_url(to_crawl['url'])