def get_404(log):
'''
Find the set of all requests with status 404 in the given `log` sequence
>>> get_404('access-log')
[]
'''
l = []
for line in apachelog.lines_from_dir(log, 'www'):
#print line
for i in apachelog.apache_log(line):
a = i['status']
b = i['request']
if a == '404':
#print b
if b not in l:
l.append(b)
return l
def get_largest(log):
'''
Find the largest data transfer in the given `log` sequence
>>> get_largest('access-log')
(4919642, '/dynamic/ffcache.zip')
'''
size = 0
f = 0
for line in apachelog.lines_from_dir(log, 'www'):
for i in apachelog.apache_log(line):
try:
a = int(i["bytes"])
b = i["request"]
except:
continue
if a > size:
size = a
f = b
return size, f
# realtime404.py
#
# Print all 404 requests as they happen in the log
from apachelog import apache_log
from follow import follow
logfile = open("run/foo/access-log")
loglines = follow(logfile)
log = apache_log(loglines)
r404 = (r for r in log if r["status"] == 404)
for r in r404:
print(r["host"], r["datetime"], r["request"])
def main():
'''Main function'''
lines = apachelog.lines_from_dir('access-log*', 'www')
log = apachelog.apache_log(lines)
for r in sorted(get_404(log)):
print r
def main():
'''Main function'''
lines = apachelog.lines_from_dir('access-log*', 'www')
log = apachelog.apache_log(lines)
print 'Total: %d' % get_downloads_count(log, '/ply/ply-2.3.tar.gz')
# gentrace.py
#
# Trace a generator by printing items received
def trace(source):
for item in source:
print(item)
yield item
# Example use
if __name__ == "__main__":
from apachelog import apache_log
lines = open("access-log")
log = trace(apache_log(lines))
r404 = (r for r in log if r["status"] == 404)
for r in r404:
pass
def run(self):
self.target(self.generate())
# Example use:
if __name__ == "__main__":
from follow import follow
from apachelog import apache_log
from broadcast import broadcast
def find_404(log):
r404 = (r for r in log if r["status"] == 404)
for r in r404:
print(r["status"], r["datetime"], r["request"])
def bytes_transferred(log):
total = 0
for r in log:
total = r["bytes"]
print("Total bytes", total)
c1 = ConsumerThread(find_404)
c1.start()
c2 = ConsumerThread(bytes_transferred)
c2.start()
lines = follow(open("run/foo/access-log")) # Foolow a log
log = apache_log(lines) # Turn into records
broadcast(log, [c1, c2]) # Broadcast to consumers
def main():
'''Main function'''
lines = apachelog.lines_from_dir('access-log*', 'www')
log = apachelog.apache_log(lines)
print '%d %s' % get_largest(log)