AIL/ail-framework
7
0
Fork 0
mirror of https://github.com/ail-project/ail-framework.git synced 2024-11-13 01:58:22 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

Merge pull request #487 from CIRCL/crawler_v2

Browse source 
fix: [crawler] error catcher
This commit is contained in:
Thirion Aurélien 2020-04-01 14:59:24 +02:00 committed by GitHub
commit 00573c9401
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
1 changed files with 4 additions and 4 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

8
bin/torcrawler/TorSplashCrawler.py
View file

@ -156,7 +156,7 @@ class TorSplashCrawler():
self.parse, self.parse,
errback=self.errback_catcher, errback=self.errback_catcher,
endpoint='execute', endpoint='execute',
meta={'father': self.original_item}, meta={'father': self.original_item, 'current_url': self.start_urls},
args=l_cookies args=l_cookies
) )
@ -217,7 +217,7 @@ class TorSplashCrawler():
self.parse, self.parse,
errback=self.errback_catcher, errback=self.errback_catcher,
endpoint='execute', endpoint='execute',
meta={'father': item_id}, meta={'father': item_id, 'current_url': link.url},
args=l_cookies args=l_cookies
) )
@ -227,7 +227,7 @@ class TorSplashCrawler():
if failure.check(ResponseNeverReceived): if failure.check(ResponseNeverReceived):
request = failure.request request = failure.request
url= response.data['last_url'] url= request.meta['current_url']
father = request.meta['father'] father = request.meta['father']
self.logger.error('Splash, ResponseNeverReceived for %s, retry in 10s ...', url) self.logger.error('Splash, ResponseNeverReceived for %s, retry in 10s ...', url)
@ -242,7 +242,7 @@ class TorSplashCrawler():
errback=self.errback_catcher, errback=self.errback_catcher,
endpoint='execute', endpoint='execute',
cache_args=['lua_source'], cache_args=['lua_source'],
meta={'father': father}, meta={'father': father, 'current_url': url},
args=self.build_request_arg(response.cookiejar) args=self.build_request_arg(response.cookiejar)
) )