Git-Mirrors/brozzler
1
0
Fork 0
mirror of https://github.com/internetarchive/brozzler.git synced 2025-07-01 18:16:43 -04:00
Code Issues Projects Releases Packages Wiki Activity Actions 1

bugfix

Browse source
This commit is contained in:
Vangelis Banos 2017-07-31 10:36:35 +00:00
parent 9c81a7bbda
commit 6259d03be1
1 changed files with 13 additions and 9 deletions
Download patch file Download diff file Expand all files Collapse all files

22
brozzler/worker.py
View file

@ -342,6 +342,8 @@ class BrozzlerWorker:
self.logger.error( self.logger.error(
'youtube_dl raised exception on %s', page, 'youtube_dl raised exception on %s', page,
exc_info=True) exc_info=True)
else:
ydl_spy = False
if self._needs_browsing(page, ydl_spy): if self._needs_browsing(page, ydl_spy):
self.logger.info('needs browsing: %s', page) self.logger.info('needs browsing: %s', page)
@ -437,19 +439,21 @@ class BrozzlerWorker:
'proxy error fetching %s' % page.url) from e 'proxy error fetching %s' % page.url) from e
def _needs_browsing(self, page, brozzler_spy): def _needs_browsing(self, page, brozzler_spy):
final_bounces = brozzler_spy.final_bounces(page.url) if brozzler_spy:
if not final_bounces: final_bounces = brozzler_spy.final_bounces(page.url)
return True if not final_bounces:
for txn in final_bounces:
if txn['response_headers'].get_content_type() in [
'text/html', 'application/xhtml+xml']:
return True return True
for txn in final_bounces:
if txn['response_headers'].get_content_type() in [
'text/html', 'application/xhtml+xml']:
return True
return False return False
def _already_fetched(self, page, brozzler_spy): def _already_fetched(self, page, brozzler_spy):
for txn in brozzler_spy.final_bounces(page.url): if brozzler_spy:
if (txn['method'] == 'GET' and txn['status_code'] == 200): for txn in brozzler_spy.final_bounces(page.url):
return True if (txn['method'] == 'GET' and txn['status_code'] == 200):
return True
return False return False
def brozzle_site(self, browser, site): def brozzle_site(self, browser, site):