iwla/iwla.py

979 lines
35 KiB
Python
Raw Normal View History

2020-10-30 14:42:56 +01:00
#!/usr/bin/env python3
2014-12-18 19:54:31 +01:00
# -*- coding: utf-8 -*-
#
# Copyright Grégory Soutadé 2015
# This file is part of iwla
# iwla is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# iwla is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with iwla. If not, see <http://www.gnu.org/licenses/>.
#
2014-11-18 20:18:53 +01:00
import os
2014-12-03 10:55:32 +01:00
import shutil
import sys
2014-11-18 20:18:53 +01:00
import re
import time
2020-10-30 14:42:56 +01:00
import pickle
2014-11-19 19:34:16 +01:00
import gzip
2014-11-22 19:23:56 +01:00
import importlib
2014-12-03 10:55:32 +01:00
import argparse
2014-12-15 22:30:49 +01:00
import logging
import gettext
2014-11-28 16:26:11 +01:00
from calendar import monthrange
2015-01-02 19:27:57 +01:00
from datetime import date, datetime
2014-11-19 19:45:41 +01:00
2014-11-24 21:37:37 +01:00
import default_conf as conf
2014-11-24 17:13:59 +01:00
from iplugin import *
from display import *
"""
Main class IWLA
Parse Log, compute them, call plugins and produce output
For now, only HTTP log are valid
Plugin requirements :
None
Conf values needed :
analyzed_filename
domain_name
locales_path
2020-04-15 14:44:11 +02:00
compress_output_files
excluded_ip
2023-06-14 09:21:11 +02:00
excluded_domain_name
Output files :
DB_ROOT/meta.db
DB_ROOT/year/month/iwla.db
OUTPUT_ROOT/index.html
2014-12-31 14:22:46 +01:00
OUTPUT_ROOT/year/_stats.html
OUTPUT_ROOT/year/month/index.html
Statistics creation :
meta :
last_time
start_analysis_time
stats =>
year =>
month =>
viewed_bandwidth
not_viewed_bandwidth
viewed_pages
viewed_hits
nb_visits
nb_visitors
month_stats :
viewed_bandwidth
not_viewed_bandwidth
viewed_pages
viewed_hits
nb_visits
days_stats :
day =>
viewed_bandwidth
not_viewed_bandwidth
viewed_pages
viewed_hits
nb_visits
nb_visitors
visits :
remote_ip =>
remote_addr
remote_ip
viewed_pages{0..31} # 0 contains total
viewed_hits{0..31} # 0 contains total
not_viewed_pages{0..31}
not_viewed_hits{0..31}
bandwidth{0..31}
last_access
requests =>
[fields_from_format_log]
extract_request =>
http_method
http_uri
http_version
extract_uri
extract_parameters*
extract_referer* =>
extract_uri
extract_parameters*
robot
hit_only
is_page
keep_requests
valid_visitors:
month_stats without robot and hit only visitors (if not conf.count_hit_only_visitors)
Statistics update :
None
Statistics deletion :
None
"""
2014-11-21 14:46:12 +01:00
class IWLA(object):
ANALYSIS_CLASS = 'HTTP'
API_VERSION = 1
2023-01-28 09:45:13 +01:00
IWLA_VERSION = '0.7'
2014-11-21 14:46:12 +01:00
def __init__(self, logLevel, args):
2014-11-21 14:46:12 +01:00
self.meta_infos = {}
self.analyse_started = False
self.current_analysis = {}
2015-01-02 19:27:57 +01:00
self.start_time = 0
2014-11-21 14:46:12 +01:00
self.cache_plugins = {}
2014-11-30 19:05:17 +01:00
self.display = DisplayHTMLBuild(self)
2014-11-21 14:46:12 +01:00
self.valid_visitors = None
self.args = args
2014-11-21 14:46:12 +01:00
2017-08-23 20:11:17 +02:00
self.log_format_extracted = re.sub(r'([^\$?\w])', r'\\\g<1>', conf.log_format)
2014-11-21 14:46:12 +01:00
self.log_format_extracted = re.sub(r'\$(\w+)', '(?P<\g<1>>.+)', self.log_format_extracted)
self.http_request_extracted = re.compile(r'(?P<http_method>\S+) (?P<http_uri>\S+) (?P<http_version>\S+)')
self.log_re = re.compile(self.log_format_extracted)
self.uri_re = re.compile(r'(?P<extract_uri>[^\?#]+)(\?(?P<extract_parameters>[^#]+))?(#.*)?')
2014-12-17 21:06:48 +01:00
self.domain_name_re = re.compile(r'.*%s' % conf.domain_name)
2022-06-23 21:11:43 +02:00
self.slash_re = re.compile(r'//')
self.protocol_re = re.compile(r'^.*://')
2020-04-09 09:33:59 +02:00
self.excluded_ip = []
for ip in conf.excluded_ip:
self.excluded_ip += [re.compile(ip)]
2023-06-14 09:21:11 +02:00
self.excluded_domain_name = []
for domain_name in conf.excluded_domain_name:
self.excluded_domain_name += [re.compile(domain_name)]
self.plugins = [(conf.PRE_HOOK_DIRECTORY , conf.pre_analysis_hooks),
(conf.POST_HOOK_DIRECTORY , conf.post_analysis_hooks),
(conf.DISPLAY_HOOK_DIRECTORY , conf.display_hooks)]
2014-11-21 14:46:12 +01:00
2014-12-16 20:23:33 +01:00
logging.basicConfig(format='%(name)s %(message)s', level=logLevel)
self.logger = logging.getLogger(self.__class__.__name__)
if self.args.dry_run:
2017-05-25 21:03:46 +02:00
self.logger.info('==> Start (DRY RUN)')
else:
self.logger.info('==> Start')
try:
2020-10-30 14:42:56 +01:00
t = gettext.translation('iwla', localedir=conf.locales_path, languages=[conf.locale])
self.logger.info('\tUsing locale %s' % (conf.locale))
except IOError:
t = gettext.NullTranslations()
self.logger.info('\tUsing default locale en_EN')
2020-10-30 14:42:56 +01:00
self._ = t.gettext
2014-12-17 19:00:42 +01:00
def getVersion(self):
return IWLA.IWLA_VERSION
def getConfValue(self, key, default=None):
2014-11-24 21:42:57 +01:00
if not key in dir(conf):
return default
else:
return conf.__dict__[key]
2014-11-21 14:46:12 +01:00
def _clearVisits(self):
self.current_analysis = {
'days_stats' : {},
'month_stats' : {},
'visits' : {}
}
self.valid_visitors = None
return self.current_analysis
def getDaysStats(self):
return self.current_analysis['days_stats']
2014-11-21 16:56:58 +01:00
def getMonthStats(self):
2014-11-21 14:46:12 +01:00
return self.current_analysis['month_stats']
def getCurrentVisits(self):
2014-11-21 14:46:12 +01:00
return self.current_analysis['visits']
def getSortedCurrentVisits(self):
visits = self.current_analysis['visits'].values()
return sorted(visits, key=lambda hit: hit['last_access'])
2014-11-21 14:46:12 +01:00
def getValidVisitors(self):
2014-11-21 16:56:58 +01:00
return self.valid_visitors
def getDisplay(self):
return self.display
2014-11-21 14:46:12 +01:00
2014-11-25 16:59:29 +01:00
def getCurTime(self):
return self.meta_infos['last_time']
2014-11-26 19:53:00 +01:00
def getStartAnalysisTime(self):
return self.meta_infos['start_analysis_time']
def isValidForCurrentAnalysis(self, request):
cur_time = self.meta_infos['start_analysis_time']
2014-11-27 14:11:47 +01:00
# Analyse not started
if not cur_time: return False
return (time.mktime(cur_time) < time.mktime(request['time_decoded']))
2014-11-27 13:07:14 +01:00
def hasBeenViewed(self, request):
return int(request['status']) in conf.viewed_http_codes
def getCurDisplayPath(self, filename):
2014-11-27 14:11:47 +01:00
cur_time = self.meta_infos['last_time']
2014-12-12 13:24:47 +01:00
return os.path.join(str(cur_time.tm_year), '%02d' % (cur_time.tm_mon), filename)
2014-11-27 14:11:47 +01:00
2014-11-30 19:05:17 +01:00
def getResourcesPath(self):
return conf.resources_path
def getCSSPath(self):
return conf.css_path
2014-11-21 14:46:12 +01:00
def _clearMeta(self):
self.meta_infos = {
'last_time' : None,
'start_analysis_time' : None
2014-11-21 14:46:12 +01:00
}
return self.meta_infos
def _clearDisplay(self):
2016-02-04 20:46:12 +01:00
self.display.clear()
2020-10-30 14:42:56 +01:00
return self.display
2014-11-21 14:46:12 +01:00
def getDBFilename(self, time):
2014-12-12 13:24:47 +01:00
return os.path.join(conf.DB_ROOT, str(time.tm_year), '%02d' % (time.tm_mon), conf.DB_FILENAME)
2014-11-21 14:46:12 +01:00
2015-05-23 16:38:39 +02:00
def _openDB(self, filename, prot='r'):
if self.args.dont_compress:
return open(filename, prot)
else:
return gzip.open(filename, prot)
2014-11-21 14:46:12 +01:00
def _serialize(self, obj, filename):
if self.args.dry_run: return
self.logger.info("==> Serialize to %s" % (filename))
2014-11-21 14:46:12 +01:00
base = os.path.dirname(filename)
if not os.path.exists(base):
os.makedirs(base)
# Make a backup in case of something fails
if os.path.exists(filename):
shutil.copy(filename, filename + '.bak')
2015-05-23 16:38:39 +02:00
with open(filename + '.tmp', 'wb+') as f, self._openDB(filename, 'w') as fzip:
2020-10-30 14:42:56 +01:00
pickle.dump(obj, f)
2014-11-21 16:56:58 +01:00
f.seek(0)
2014-12-19 18:06:02 +01:00
fzip.write(f.read())
os.fsync(fzip)
2014-11-19 19:34:16 +01:00
os.remove(filename + '.tmp')
if os.path.exists(filename + '.bak'):
os.remove(filename + '.bak')
2014-11-19 08:01:12 +01:00
2014-11-21 14:46:12 +01:00
def _deserialize(self, filename):
if not os.path.exists(filename):
return None
2017-05-07 16:55:05 +02:00
res = None
2015-05-23 16:38:39 +02:00
with self._openDB(filename) as f:
2020-10-30 14:42:56 +01:00
res = pickle.load(f)
2017-05-07 16:55:05 +02:00
return res
2014-11-19 08:01:12 +01:00
def _callPlugins(self, target_root, *args):
2014-12-16 20:23:33 +01:00
self.logger.info('==> Call plugins (%s)' % (target_root))
for (root, plugins) in self.plugins:
if root != target_root: continue
for p in plugins:
mod = self.cache_plugins.get(root + '.' + p, None)
if mod:
2014-12-16 20:23:33 +01:00
self.logger.info('\t%s' % (p))
mod.hook(*args)
2014-11-19 08:01:12 +01:00
2014-11-21 14:46:12 +01:00
def isPage(self, request):
self.logger.debug("Is page %s" % (request))
2014-11-24 21:37:37 +01:00
for e in conf.pages_extensions:
2014-11-21 14:46:12 +01:00
if request.endswith(e):
self.logger.debug("True")
2014-11-21 14:46:12 +01:00
return True
# No extension -> page
if not '.' in request.split('/')[-1]:
self.logger.debug("True")
return True
self.logger.debug("False")
2014-11-21 14:46:12 +01:00
return False
2014-11-18 20:18:53 +01:00
def isMultimediaFile(self, request):
self.logger.debug("Is multimedia %s" % (request))
for e in conf.multimedia_files:
if request.lower().endswith(e):
self.logger.debug("True")
return True
self.logger.debug("False")
return False
2016-01-18 07:33:48 +01:00
def isValidVisitor(self, hit):
if hit['robot']: return False
if not conf.count_hit_only_visitors and not hit['viewed_pages'][0]:
2016-01-18 07:33:48 +01:00
return False
return True
2016-09-25 20:36:51 +02:00
def isRobot(self, hit):
# By default robot is None
return hit['robot'] == True
2016-09-25 20:36:51 +02:00
2014-11-21 14:46:12 +01:00
def _appendHit(self, hit):
remote_ip = hit['remote_ip']
2014-11-21 14:46:12 +01:00
if not remote_ip: return
2014-11-28 16:02:04 +01:00
2020-04-09 09:33:59 +02:00
for ip in self.excluded_ip:
if ip.match(remote_ip):
2020-04-09 09:33:59 +02:00
return
# Redirected page/hit
if int(hit['status']) in (301, 302, 307, 308):
return
if not remote_ip in self.current_analysis['visits'].keys():
2014-11-25 16:22:07 +01:00
self._createVisitor(hit)
2020-10-30 14:42:56 +01:00
super_hit = self.current_analysis['visits'][remote_ip]
# Don't keep all requests for robots
if not super_hit['robot']:
super_hit['requests'].append(hit)
day = self.meta_infos['last_time'].tm_mday
if self.hasBeenViewed(hit):
super_hit['bandwidth'][day] = super_hit['bandwidth'].get(day, 0) + int(hit['body_bytes_sent'])
super_hit['bandwidth'][0] += int(hit['body_bytes_sent'])
2020-10-30 14:42:56 +01:00
super_hit['last_access'] = self.meta_infos['last_time']
2014-11-21 14:46:12 +01:00
request = hit['extract_request']
uri = request.get('extract_uri', request['http_uri'])
2014-11-21 14:46:12 +01:00
hit['is_page'] = self.isPage(uri)
if super_hit['robot'] or\
2020-10-30 14:42:56 +01:00
not self.hasBeenViewed(hit):
2014-11-21 14:46:12 +01:00
page_key = 'not_viewed_pages'
hit_key = 'not_viewed_hits'
else:
page_key = 'viewed_pages'
hit_key = 'viewed_hits'
if hit['is_page']:
super_hit[page_key][day] = super_hit[page_key].get(day, 0) + 1
super_hit[page_key][0] += 1
2014-11-21 14:46:12 +01:00
else:
super_hit[hit_key][day] = super_hit[hit_key].get(day, 0) + 1
super_hit[hit_key][0] += 1
2014-11-21 14:46:12 +01:00
2014-11-25 16:22:07 +01:00
def _createVisitor(self, hit):
2014-11-21 14:46:12 +01:00
super_hit = self.current_analysis['visits'][hit['remote_addr']] = {}
super_hit['remote_addr'] = hit['remote_addr']
2014-11-26 16:17:16 +01:00
super_hit['remote_ip'] = hit['remote_addr']
super_hit['viewed_pages'] = {0:0}
super_hit['viewed_hits'] = {0:0}
super_hit['not_viewed_pages'] = {0:0}
super_hit['not_viewed_hits'] = {0:0}
super_hit['bandwidth'] = {0:0}
2014-11-21 14:46:12 +01:00
super_hit['last_access'] = self.meta_infos['last_time']
super_hit['requests'] = []
super_hit['robot'] = None
2014-11-21 14:46:12 +01:00
super_hit['hit_only'] = 0
def _normalizeURI(self, uri, removeFileSlash=True):
if uri == '/': return uri
2022-06-23 21:11:43 +02:00
# Remove protocol
uri = self.protocol_re.sub('', uri)
# Remove double /
uri = self.slash_re.sub('/', uri)
if removeFileSlash:
while len(uri) > 1 and uri[-1] == '/':
uri = uri[:-1]
return uri
def _normalizeParameters(self, parameters):
# No parameters
if parameters == '?': return None
return parameters
2014-11-21 14:46:12 +01:00
def _decodeHTTPRequest(self, hit):
if not 'request' in hit.keys(): return False
groups = self.http_request_extracted.match(hit['request'])
if groups:
2017-08-23 20:11:17 +02:00
hit['extract_request'] = groups.groupdict("")
2014-11-21 14:46:12 +01:00
uri_groups = self.uri_re.match(hit['extract_request']['http_uri'])
if uri_groups:
2017-08-23 20:11:17 +02:00
d = uri_groups.groupdict("")
hit['extract_request']['extract_uri'] = self._normalizeURI(d['extract_uri'])
2014-11-21 14:46:12 +01:00
if 'extract_parameters' in d.keys():
parameters = self._normalizeParameters(d['extract_parameters'])
if parameters:
hit['extract_request']['extract_parameters'] = parameters
2014-11-21 14:46:12 +01:00
else:
2014-12-16 20:23:33 +01:00
self.logger.warning("Bad request extraction %s" % (hit['request']))
2014-11-21 14:46:12 +01:00
return False
2014-11-26 16:17:16 +01:00
if hit['http_referer']:
referer_groups = self.uri_re.match(hit['http_referer'])
if referer_groups:
2017-08-23 20:11:17 +02:00
hit['extract_referer'] = referer_groups.groupdict("")
hit['extract_referer']['extract_uri'] = self._normalizeURI(hit['extract_referer']['extract_uri'])
hit['extract_referer']['extract_parameters'] = self._normalizeParameters(hit['extract_referer']['extract_parameters'])
hit['remote_ip'] = hit['remote_addr']
2014-11-21 14:46:12 +01:00
return True
def _decodeTime(self, hit):
2014-12-10 21:41:22 +01:00
try:
hit['time_decoded'] = time.strptime(hit['time_local'], conf.time_format)
2020-10-30 14:42:56 +01:00
except ValueError as e:
2014-12-10 21:41:22 +01:00
if sys.version_info < (3, 2):
2020-10-30 14:42:56 +01:00
# Try without UTC value at the end (%z not recognized)
2014-12-10 21:41:22 +01:00
gmt_offset_str = hit['time_local'][-5:]
gmt_offset_hours = int(gmt_offset_str[1:3])*60*60
gmt_offset_minutes = int(gmt_offset_str[3:5])*60
gmt_offset = gmt_offset_hours + gmt_offset_minutes
hit['time_decoded'] = time.strptime(hit['time_local'][:-6], conf.time_format[:-3])
2014-12-31 18:00:10 +01:00
# if gmt_offset_str[0] == '-':
# hit['time_decoded'] = time.localtime(time.mktime(hit['time_decoded'])+gmt_offset)
# else:
# hit['time_decoded'] = time.localtime(time.mktime(hit['time_decoded'])-gmt_offset)
2014-12-10 21:41:22 +01:00
else:
raise e
2014-11-26 19:53:00 +01:00
return hit['time_decoded']
2014-11-21 14:46:12 +01:00
def getDisplayIndex(self):
cur_time = self.meta_infos['last_time']
filename = self.getCurDisplayPath('index.html')
2014-11-21 14:46:12 +01:00
2014-11-21 16:56:58 +01:00
return self.display.getPage(filename)
2014-11-21 14:46:12 +01:00
2014-12-03 21:58:55 +01:00
def _generateDisplayDaysStats(self):
2014-11-21 14:46:12 +01:00
cur_time = self.meta_infos['last_time']
2014-12-19 17:50:45 +01:00
title = createCurTitle(self, self._('Statistics'))
filename = self.getCurDisplayPath('index.html')
2014-12-16 20:23:33 +01:00
self.logger.info('==> Generate display (%s)' % (filename))
page = self.display.createPage(title, filename, conf.css_path)
2014-12-31 14:22:46 +01:00
link = DisplayHTMLRaw(self, '<iframe src="../_stats.html"></iframe>')
page.appendBlock(link)
2014-11-21 14:46:12 +01:00
months_name = ['', self._('Jan'), self._('Feb'), self._('Mar'), self._('Apr'), self._('May'), self._('June'), self._('Jul'), self._('Aug'), self._('Sep'), self._('Oct'), self._('Nov'), self._('Dec')]
_, nb_month_days = monthrange(cur_time.tm_year, cur_time.tm_mon)
days = self.display.createBlock(DisplayHTMLBlockTableWithGraph, self._('By day'), [self._('Day'), self._('Visits'), self._('Pages'), self._('Hits'), self._('Bandwidth'), self._('Not viewed Bandwidth')], None, nb_month_days, range(1,6), [4, 5])
2014-12-15 20:43:43 +01:00
days.setColsCSSClass(['', 'iwla_visit', 'iwla_page', 'iwla_hit', 'iwla_bandwidth', 'iwla_bandwidth'])
2014-11-21 14:46:12 +01:00
nb_visits = 0
2014-11-28 16:26:11 +01:00
nb_days = 0
2014-12-04 19:15:15 +01:00
for i in range(1, nb_month_days+1):
month = months_name[int(time.strftime('%m', cur_time), 10)]
day = '%d<br/>%s' % (i, month)
full_day = '%02d %s %d' % (i, month, cur_time.tm_year)
2014-11-28 16:26:11 +01:00
if i in self.current_analysis['days_stats'].keys():
stats = self.current_analysis['days_stats'][i]
2014-12-15 20:43:43 +01:00
row = [full_day, stats['nb_visits'], stats['viewed_pages'], stats['viewed_hits'],
stats['viewed_bandwidth'], stats['not_viewed_bandwidth']]
2014-12-15 20:43:43 +01:00
nb_visits += stats['nb_visits']
2014-11-28 16:26:11 +01:00
nb_days += 1
else:
2014-12-02 21:53:20 +01:00
row = [full_day, 0, 0, 0, 0, 0]
2014-11-21 16:56:58 +01:00
days.appendRow(row)
viewed_bandwidth = row[4]
not_viewed_bandwidth = row[5]
days.setCellValue(i-1, 4, viewed_bandwidth)
days.setCellValue(i-1, 5, not_viewed_bandwidth)
2014-12-02 21:53:20 +01:00
days.appendShortTitle(day)
2014-12-04 19:15:15 +01:00
adate = date(cur_time.tm_year, cur_time.tm_mon, i)
2014-12-02 21:53:20 +01:00
week_day = adate.weekday()
2014-12-02 21:16:27 +01:00
if week_day == 5 or week_day == 6:
2014-12-04 19:15:15 +01:00
days.setRowCSSClass(i-1, 'iwla_weekend')
2014-12-02 21:53:20 +01:00
if adate == date.today():
css = days.getCellCSSClass(i-1, 0)
2014-12-02 21:53:20 +01:00
if css: css = '%s %s' % (css, 'iwla_curday')
else: css = 'iwla_curday'
2014-12-04 19:15:15 +01:00
days.setCellCSSClass(i-1, 0, css)
2014-11-21 14:46:12 +01:00
stats = self.current_analysis['month_stats']
row = [0, nb_visits, stats['viewed_pages'], stats['viewed_hits'], stats['viewed_bandwidth'], stats['not_viewed_bandwidth']]
if nb_days:
2020-12-09 13:24:29 +01:00
average_row = list(map(lambda v: int(v/nb_days), row))
2014-11-21 14:46:12 +01:00
else:
2020-12-09 13:24:29 +01:00
average_row = list(map(lambda v: 0, row))
2014-11-21 14:46:12 +01:00
average_row[0] = self._('Average')
days.appendRow(average_row)
2014-11-18 20:18:53 +01:00
row[0] = self._('Total')
2014-11-21 16:56:58 +01:00
days.appendRow(row)
page.appendBlock(days)
self.display.addPage(page)
2014-11-20 08:18:31 +01:00
2014-12-03 21:58:55 +01:00
def _generateDisplayMonthStats(self, page, year, month_stats):
2014-12-04 19:15:15 +01:00
cur_time = time.localtime()
2014-12-19 17:50:45 +01:00
months_name = ['', self._('Jan'), self._('Feb'), self._('Mar'), self._('Apr'), self._('May'), self._('June'), self._('Jul'), self._('Aug'), self._('Sep'), self._('Oct'), self._('Nov'), self._('Dec')]
title = '%s %d' % (self._('Summary'), year)
cols = [self._('Month'), self._('Visitors'), self._('Visits'), self._('Pages'), self._('Hits'), self._('Bandwidth'), self._('Not viewed Bandwidth')]
graph_cols=range(1,6)
months = self.display.createBlock(DisplayHTMLBlockTableWithGraph, title, cols, None, 12, graph_cols, [5, 6])
months.setColsCSSClass(['', 'iwla_visitor', 'iwla_visit', 'iwla_page', 'iwla_hit', 'iwla_bandwidth', 'iwla_bandwidth'])
2014-12-03 21:58:55 +01:00
total = [0] * len(cols)
2014-12-04 19:15:15 +01:00
for i in range(1, 13):
month = '%s<br/>%d' % (months_name[i], year)
full_month = '%s %d' % (months_name[i], year)
link_month = '<a target="_top" href="/%d/%02d/index.html">%s</a>' % (year, i, full_month)
2014-12-03 21:58:55 +01:00
if i in month_stats.keys():
stats = month_stats[i]
row = [link_month, stats['nb_visitors'], stats['nb_visits'], stats['viewed_pages'], stats['viewed_hits'],
stats['viewed_bandwidth'], stats['not_viewed_bandwidth']]
for j in range(1,7):
2014-12-04 19:15:15 +01:00
total[j] += row[j]
2014-12-03 21:58:55 +01:00
else:
row = [full_month, 0, 0, 0, 0, 0, 0]
2014-12-03 21:58:55 +01:00
months.appendRow(row)
months.appendShortTitle(month)
2014-12-04 19:15:15 +01:00
if year == cur_time.tm_year and i == cur_time.tm_mon:
css = months.getCellCSSClass(i-1, 0)
if css: css = '%s %s' % (css, 'iwla_curday')
else: css = 'iwla_curday'
months.setCellCSSClass(i-1, 0, css)
2014-12-03 21:58:55 +01:00
total[0] = self._('Total')
2014-12-03 21:58:55 +01:00
months.appendRow(total)
page.appendBlock(months)
2014-12-31 14:22:46 +01:00
filename = '%d/_stats.html' % (year)
page_ = self.display.createPage(u'', filename, conf.css_path)
page_.appendBlock(months)
2014-12-31 14:22:46 +01:00
page_.build(conf.DISPLAY_ROOT, False)
months.resetHTML()
2014-12-31 14:22:46 +01:00
2014-12-03 21:58:55 +01:00
def _generateDisplayWholeMonthStats(self):
title = '%s %s' % (self._('Statistics for'), conf.domain_name)
2014-12-03 21:58:55 +01:00
filename = 'index.html'
2014-12-16 20:23:33 +01:00
self.logger.info('==> Generate main page (%s)' % (filename))
2014-12-03 21:58:55 +01:00
page = self.display.createPage(title, filename, conf.css_path)
2014-12-03 21:58:55 +01:00
2015-01-02 19:27:57 +01:00
last_update = u'<b>%s</b> %s<br />' % (self._(u'Last update'), time.strftime('%02d %b %Y %H:%M', time.localtime()))
page.appendBlock(self.display.createBlock(DisplayHTMLRaw, last_update))
2015-01-02 19:27:57 +01:00
duration = datetime.now() - self.start_time
duration = time.gmtime(duration.seconds)
time_analysis = u'<b>%s</b> ' % (self._('Time analysis'))
if duration.tm_hour:
time_analysis += u'%d %s, ' % (duration.tm_hour, self._(u'hours'))
time_analysis += u'%d %s and %d %s<br />' % (duration.tm_min, self._(u'minutes'), duration.tm_sec, self._(u'seconds'))
page.appendBlock(self.display.createBlock(DisplayHTMLRaw, time_analysis))
2014-12-04 21:47:11 +01:00
for year in sorted(self.meta_infos['stats'].keys(), reverse=True):
2014-12-03 21:58:55 +01:00
self._generateDisplayMonthStats(page, year, self.meta_infos['stats'][year])
self.display.addPage(page)
def _compressFile(self, root, filename):
2014-12-15 21:28:25 +01:00
path = os.path.join(root, filename)
gz_path = path + '.gz'
2014-12-16 20:23:33 +01:00
self.logger.debug('Compress %s => %s' % (path, gz_path))
2014-12-15 21:28:25 +01:00
if not os.path.exists(gz_path) or\
os.stat(path).st_mtime > os.stat(gz_path).st_mtime:
if self.args.dry_run: return
2014-12-19 18:06:02 +01:00
with open(path, 'rb') as f_in, gzip.open(gz_path, 'wb') as f_out:
f_out.write(f_in.read())
2014-12-15 21:28:25 +01:00
def _compressFiles(self, root):
2014-12-15 21:28:25 +01:00
if not conf.compress_output_files: return
for rootdir, subdirs, files in os.walk(root, followlinks=True):
for f in files:
for ext in conf.compress_output_files:
if f.endswith(ext):
self._compressFile(rootdir, f)
2014-12-15 21:28:25 +01:00
break
2020-10-30 14:42:56 +01:00
2014-11-21 14:46:12 +01:00
def _generateDisplay(self):
if self.args.disable_display: return
2014-12-03 21:58:55 +01:00
self._generateDisplayDaysStats()
self._callPlugins(conf.DISPLAY_HOOK_DIRECTORY)
2014-12-03 21:58:55 +01:00
self._generateDisplayWholeMonthStats()
if self.args.dry_run: return
2014-11-24 21:37:37 +01:00
self.display.build(conf.DISPLAY_ROOT)
self._compressFiles(conf.DISPLAY_ROOT)
2014-11-20 08:18:31 +01:00
def _createEmptyStats(self):
2014-11-21 14:46:12 +01:00
stats = {}
stats['viewed_bandwidth'] = 0
stats['not_viewed_bandwidth'] = 0
stats['viewed_pages'] = 0
stats['viewed_hits'] = 0
2014-12-15 20:43:43 +01:00
stats['nb_visits'] = 0
2014-11-18 20:18:53 +01:00
2014-11-21 14:46:12 +01:00
return stats
2014-11-21 10:41:29 +01:00
2014-11-21 14:46:12 +01:00
def _generateMonthStats(self):
self._clearDisplay()
2014-11-21 14:46:12 +01:00
visits = self.current_analysis['visits']
stats = self._createEmptyStats()
for (day, stat) in self.current_analysis['days_stats'].items():
for k in stats.keys():
stats[k] += stat[k]
2014-12-15 20:43:43 +01:00
2014-12-01 21:13:35 +01:00
duplicated_stats = {k:v for (k,v) in stats.items()}
2014-11-21 14:46:12 +01:00
cur_time = self.meta_infos['last_time']
2014-12-16 20:23:33 +01:00
self.logger.info("== Stats for %d/%02d ==" % (cur_time.tm_year, cur_time.tm_mon))
self.logger.info(stats)
2014-11-21 14:46:12 +01:00
if not 'month_stats' in self.current_analysis.keys():
self.current_analysis['month_stats'] = stats
else:
for (k,v) in stats.items():
self.current_analysis['month_stats'][k] = v
2014-11-21 14:46:12 +01:00
self.valid_visitors = {}
for (k,v) in visits.items():
2016-01-18 07:33:48 +01:00
if self.isValidVisitor(v):
self.valid_visitors[k] = v
2014-12-15 20:43:43 +01:00
duplicated_stats['nb_visitors'] = stats['nb_visitors'] = len(self.valid_visitors.keys())
2014-12-03 21:58:55 +01:00
self._callPlugins(conf.POST_HOOK_DIRECTORY)
if self.args.display_only:
2020-12-09 13:24:29 +01:00
if not 'stats' in self.meta_infos.keys():
self.meta_infos['stats'] = {}
2016-08-20 13:11:22 +02:00
self._generateDisplay()
return
for (k,v) in visits.items():
# Keep at least one request (for referers...)
if not v.get('keep_requests', conf.keep_requests):
if len(v['requests']) > 1:
v['requests'] = [v['requests'][0]]
2014-11-21 14:46:12 +01:00
path = self.getDBFilename(cur_time)
self._serialize(self.current_analysis, path)
2014-12-01 21:13:35 +01:00
# Save month stats
2014-12-03 21:58:55 +01:00
year = cur_time.tm_year
month = cur_time.tm_mon
2014-12-01 21:13:35 +01:00
if not 'stats' in self.meta_infos.keys():
self.meta_infos['stats'] = {}
if not year in self.meta_infos['stats'].keys():
self.meta_infos['stats'][year] = {}
self.meta_infos['stats'][year][month] = duplicated_stats
meta_path = os.path.join(conf.DB_ROOT, conf.META_FILENAME)
self._serialize(self.meta_infos, meta_path)
2014-12-03 21:58:55 +01:00
self._generateDisplay()
2014-11-21 14:46:12 +01:00
def _generateDayStats(self):
self._callPlugins(conf.PRE_HOOK_DIRECTORY)
2016-08-20 13:11:22 +02:00
2014-11-21 14:46:12 +01:00
visits = self.current_analysis['visits']
cur_time = self.meta_infos['last_time']
stats = self._createEmptyStats()
day = cur_time.tm_mday
for (k, super_hit) in visits.items():
if super_hit['last_access'].tm_mday != day:
continue
if super_hit['robot']:
stats['not_viewed_bandwidth'] += super_hit['bandwidth'].get(day, 0)
continue
stats['viewed_bandwidth'] += super_hit['bandwidth'].get(day, 0)
stats['viewed_hits'] += super_hit['viewed_hits'].get(day, 0)
stats['viewed_pages'] += super_hit['viewed_pages'].get(day, 0)
if ((conf.count_hit_only_visitors and super_hit['viewed_hits'].get(day, 0)) or\
super_hit['viewed_pages'].get(day, 0)):
2014-12-15 20:43:43 +01:00
stats['nb_visits'] += 1
2014-11-21 14:46:12 +01:00
2014-12-16 20:23:33 +01:00
self.logger.info("== Stats for %d/%02d/%02d ==" % (cur_time.tm_year, cur_time.tm_mon, cur_time.tm_mday))
self.logger.info(stats)
2014-11-21 14:46:12 +01:00
self.current_analysis['days_stats'][cur_time.tm_mday] = stats
def _newHit(self, hit):
2014-12-17 21:06:48 +01:00
if not self.domain_name_re.match(hit['server_name']):
self.logger.debug("Not in domain %s" % (hit))
2014-12-17 21:06:48 +01:00
return False
2023-06-14 09:21:11 +02:00
for domain_name in self.excluded_domain_name:
if domain_name.match(hit['server_name']):
self.logger.debug("Domain name %s excluded" % (hit['server_name']))
return False
2014-11-26 19:53:00 +01:00
t = self._decodeTime(hit)
2014-11-21 14:46:12 +01:00
cur_time = self.meta_infos['last_time']
if cur_time == None:
self.current_analysis = self._deserialize(self.getDBFilename(t)) or self._clearVisits()
self.analyse_started = True
else:
if not self.analyse_started and\
2020-10-30 14:42:56 +01:00
time.mktime(t) <= time.mktime(cur_time):
self.logger.debug("Not in time")
2014-12-12 13:18:12 +01:00
return False
self.analyse_started = True
if t < cur_time: # Don't accept past hits
return False
2014-11-21 14:46:12 +01:00
if cur_time.tm_mon != t.tm_mon:
self._generateDayStats()
2014-11-21 14:46:12 +01:00
self._generateMonthStats()
self.current_analysis = self._deserialize(self.getDBFilename(t)) or self._clearVisits()
elif cur_time.tm_mday != t.tm_mday:
self._generateDayStats()
self.meta_infos['last_time'] = t
2014-11-26 19:53:00 +01:00
if not self.meta_infos['start_analysis_time']:
self.meta_infos['start_analysis_time'] = t
2014-11-21 14:46:12 +01:00
if not self._decodeHTTPRequest(hit): return False
if hit['extract_request']['http_method'] not in ['GET', 'POST']:
return False
2014-11-21 14:46:12 +01:00
for k in hit.keys():
2014-11-28 16:02:04 +01:00
if hit[k] == '-' or hit[k] == '*':
hit[k] = ''
2014-12-17 21:06:48 +01:00
2014-11-21 14:46:12 +01:00
self._appendHit(hit)
return True
def _reset(self):
reset_time = time.strptime(self.args.reset, '%m/%Y')
self.logger.info('Reset time')
self.logger.info(reset_time)
self.meta_infos['last_time'] = reset_time
cur_time = time.localtime()
year = reset_time.tm_year
while year < cur_time.tm_year:
db_path = os.path.join(conf.DB_ROOT, str(year))
if os.path.exists(db_path): shutil.rmtree(db_path)
output_path = os.path.join(conf.DISPLAY_ROOT, str(year))
if os.path.exists(output_path): shutil.rmtree(output_path)
year += 1
month = reset_time.tm_mon
while month <= cur_time.tm_mon:
db_path = os.path.join(conf.DB_ROOT, str(year), '%02d' % (month))
if os.path.exists(db_path): shutil.rmtree(db_path)
output_path = os.path.join(conf.DISPLAY_ROOT, str(year), '%02d' % (month))
if os.path.exists(output_path): shutil.rmtree(output_path)
month += 1
def start(self, _file):
2015-01-02 19:27:57 +01:00
self.start_time = datetime.now()
meta_path = os.path.join(conf.DB_ROOT, conf.META_FILENAME)
if os.path.exists(meta_path):
self.logger.info('==> Load previous database')
2015-01-02 19:27:57 +01:00
self.meta_infos = self._deserialize(meta_path) or self._clearMeta()
2014-12-10 21:41:22 +01:00
if self.meta_infos['last_time']:
if args.reset:
self._reset()
2014-12-16 20:23:33 +01:00
self.logger.info('Last time')
self.logger.info(self.meta_infos['last_time'])
2014-11-21 14:46:12 +01:00
self.current_analysis = self._deserialize(self.getDBFilename(self.meta_infos['last_time'])) or self._clearVisits()
else:
self._clearVisits()
2014-11-26 19:53:00 +01:00
self.meta_infos['start_analysis_time'] = None
self.cache_plugins = preloadPlugins(self.plugins, self)
2014-12-16 20:23:33 +01:00
self.logger.info('==> Analysing log')
2014-11-25 16:22:07 +01:00
2014-12-03 10:55:32 +01:00
for l in _file:
# print "line " + l
2014-11-21 14:46:12 +01:00
groups = self.log_re.match(l)
2014-11-21 14:46:12 +01:00
2014-12-03 10:55:32 +01:00
if groups:
2017-08-23 20:11:17 +02:00
self._newHit(groups.groupdict(""))
2014-12-03 10:55:32 +01:00
else:
2014-12-16 20:23:33 +01:00
self.logger.warning("No match for %s" % (l))
2014-11-21 16:56:58 +01:00
#break
2014-11-21 14:46:12 +01:00
if self.analyse_started:
self._generateDayStats()
self._generateMonthStats()
2014-11-26 19:53:00 +01:00
del self.meta_infos['start_analysis_time']
2014-11-21 14:46:12 +01:00
else:
2014-12-16 20:23:33 +01:00
self.logger.info('==> Analyse not started : nothing new')
2014-11-21 14:46:12 +01:00
def displayOnly(self, start_time):
self.start_time = datetime.now()
meta_path = os.path.join(conf.DB_ROOT, conf.META_FILENAME)
if os.path.exists(meta_path):
self.logger.info('==> Load previous database')
self.meta_infos = self._deserialize(meta_path) or self._clearMeta()
self.meta_infos['last_time'] = time.strptime(start_time, '%m/%Y')
if self.meta_infos['last_time']:
self.logger.info('Last time')
self.logger.info(self.meta_infos['last_time'])
self.current_analysis = self._deserialize(self.getDBFilename(self.meta_infos['last_time'])) or self._clearVisits()
else:
self._clearVisits()
self.meta_infos['start_analysis_time'] = None
self.cache_plugins = preloadPlugins(self.plugins, self)
self.logger.info('==> Analysing log')
self._generateDayStats()
self._generateMonthStats()
class FileIter(object):
def __init__(self, filenames):
self.filenames = [f for f in filenames.split(',') if f]
for f in self.filenames:
if not os.path.exists(f):
2020-10-30 14:42:56 +01:00
print('No such file \'%s\'' % (f))
sys.exit(-1)
self.cur_file = None
self._openNextFile()
def __iter__(self):
return self
def __next__(self):
return self.next()
def _openNextFile(self):
if self.cur_file:
self.cur_file.close()
self.cur_file = None
if not self.filenames:
raise StopIteration()
filename = self.filenames.pop(0)
if filename.endswith('gz'):
self.cur_file = gzip.open(filename, 'rt')
else:
self.cur_file = open(filename, 'rt')
def next(self):
l = self.cur_file.readline()
if not l:
self._openNextFile()
l = self.cur_file.readline()
return l[:-1]
2014-11-21 16:56:58 +01:00
if __name__ == '__main__':
2014-12-03 10:55:32 +01:00
parser = argparse.ArgumentParser(description='Intelligent Web Log Analyzer')
parser.add_argument('-c', '--config-file', dest='config_file',
default='conf.py', type=str,
help='Config file to use (default conf.py)')
parser.add_argument('-C', '--clean-output', dest='clean_output', action='store_true',
2014-12-03 10:55:32 +01:00
default=False,
help='Clean output before starting')
parser.add_argument('-i', '--stdin', dest='stdin', action='store_true',
default=False,
help='Read data from stdin instead of conf.analyzed_filename')
2014-12-14 15:46:01 +01:00
parser.add_argument('-f', '--file', dest='file',
2015-07-13 13:09:32 +02:00
help='Analyse this log file, multiple files can be specified (comma separated). gz files are accepted')
2014-12-14 15:46:01 +01:00
2014-12-15 22:30:49 +01:00
parser.add_argument('-d', '--log-level', dest='loglevel',
2014-12-16 20:23:33 +01:00
default='INFO', type=str,
help='Loglevel in %s, default : %s' % (['DEBUG', 'INFO', 'WARNING', 'ERROR', 'CRITICAL'], 'INFO'))
2014-12-15 22:30:49 +01:00
parser.add_argument('-r', '--reset', dest='reset',
2015-05-23 08:53:06 +02:00
default=False,
help='Reset analysis to a specific date (month/year)')
2015-05-23 16:38:39 +02:00
parser.add_argument('-z', '--dont-compress', dest='dont_compress', action='store_true',
default=False,
2015-07-13 13:09:32 +02:00
help='Don\'t compress databases (bigger but faster, not compatible with compressed databases)')
2015-05-23 16:38:39 +02:00
parser.add_argument('-p', '--display-only', dest='display_only',
default='', type=str,
help='Only generate display for a specific date (month/year)')
parser.add_argument('-P', '--disable-display', dest='disable_display', action='store_true',
default=False,
help='Don\'t generate display')
2016-08-20 13:11:22 +02:00
2017-05-25 21:03:46 +02:00
parser.add_argument('-D', '--dry-run', dest='dry_run', action='store_true',
default=False,
help='Process log but don\'t write files (database and HTML) to disk')
2014-12-03 10:55:32 +01:00
args = parser.parse_args()
if args.config_file.endswith('.py'):
args.config_file = args.config_file[:-3]
user_conf = importlib.import_module(args.config_file)
# Load user conf
for (k,v) in user_conf.__dict__.items():
if k.endswith('_append'):
new_k = k[:-7]
if new_k in dir(conf):
if type(conf.__dict__[new_k]) == list:
if type(v) == list:
conf.__dict__[new_k] += v
else:
conf.__dict__[new_k].append(v)
else:
print("Error %s is not a list" % (new_k))
else:
print("Error %s doesn't exists in default conf" % (new_k))
else:
conf.__dict__.update({k:v})
2017-05-25 21:03:46 +02:00
if args.clean_output and not args.dry_run:
2014-12-03 10:55:32 +01:00
if os.path.exists(conf.DB_ROOT): shutil.rmtree(conf.DB_ROOT)
if os.path.exists(conf.DISPLAY_ROOT): shutil.rmtree(conf.DISPLAY_ROOT)
2014-12-16 20:23:33 +01:00
loglevel = getattr(logging, args.loglevel.upper(), None)
if not isinstance(loglevel, int):
2014-12-15 22:30:49 +01:00
raise ValueError('Invalid log level: %s' % (args.loglevel))
2014-12-17 19:00:42 +01:00
iwla = IWLA(loglevel, args)
2014-12-15 22:30:49 +01:00
required_conf = ['analyzed_filename', 'domain_name']
if not validConfRequirements(required_conf, iwla, 'Main Conf'):
sys.exit(0)
if args.display_only:
iwla.displayOnly(args.display_only)
2014-12-03 10:55:32 +01:00
else:
if args.stdin:
iwla.start(sys.stdin)
else:
filename = args.file or conf.analyzed_filename
iwla.start(FileIter(filename))