soutade/iwla
1
0
Fork 0
Code Issues Pull Requests Releases 2 Activity

Basically seems to work

Browse Source
This commit is contained in:
Gregory Soutade 2014-11-20 14:09:01 +01:00
parent 7bd5a42962
commit f593cc78d9
3 changed files with 9 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
hooks/pre_analysis/H001_soutade.py
View File

@ -13,6 +13,8 @@ def hook(hits):
for p in super_hit['pages']: for p in super_hit['pages']:
if not p['is_page']: continue if not p['is_page']: continue
if int(p['status']) != 200: continue
if logo_re.match(p['extract_request']['extract_uri']): if logo_re.match(p['extract_request']['extract_uri']):
p['is_page'] = False p['is_page'] = False
if super_hit['viewed_pages']: if super_hit['viewed_pages']:

3
hooks/pre_analysis/H002_robot.py
View File

@ -19,6 +19,9 @@ def hook(hits):
if not super_hit['viewed_hits']: if not super_hit['viewed_hits']:
super_hit['robot'] = 1 super_hit['robot'] = 1
continue continue
elif not super_hit['viewed_pages']:
# Hit only
super_hit['hit_only'] = 1
for hit in super_hit['pages']: for hit in super_hit['pages']:
# 3) /robots.txt read # 3) /robots.txt read

2
iwla.py
View File

@ -161,6 +161,7 @@ def createUser(hit):
super_hit['last_access'] = meta_visit['last_time'] super_hit['last_access'] = meta_visit['last_time']
super_hit['pages'] = []; super_hit['pages'] = [];
super_hit['robot'] = isRobot(hit); super_hit['robot'] = isRobot(hit);
super_hit['hit_only'] = 0;
appendHit(hit) appendHit(hit)
def isRobot(hit): def isRobot(hit):
@ -276,6 +277,7 @@ def generateStats(visits):
print "[%s] =>\t%d/%d" % (k, super_hit['viewed_pages'], super_hit['viewed_hits']) print "[%s] =>\t%d/%d" % (k, super_hit['viewed_pages'], super_hit['viewed_hits'])
if not super_hit['hit_only']:
stats['nb_visitors'] += 1 stats['nb_visitors'] += 1
stats['viewed_bandwidth'] += super_hit['bandwith'] stats['viewed_bandwidth'] += super_hit['bandwith']
stats['viewed_pages'] += super_hit['viewed_pages'] stats['viewed_pages'] += super_hit['viewed_pages']