Newer
Older
#
# Piwik - Open source web analytics
#
# @link http://piwik.org
# @license http://www.gnu.org/licenses/gpl-3.0.html GPL v3 or later
# @version $Id$
# For more info see: http://piwik.org/log-analytics/
import ConfigParser
import datetime
import fnmatch
import gzip
import hashlib
import inspect
import itertools
import logging
import optparse
import os
import os.path
import threading
import time
import urllib
import urllib2
Thomas Steur
a validé
import subprocess
try:
import json
except ImportError:
try:
import simplejson as json
except ImportError:
if sys.version_info < (2, 6):
print >> sys.stderr, 'simplejson (http://pypi.python.org/pypi/simplejson/) is required.'
sys.exit(1)
##
## Constants.
##
STATIC_EXTENSIONS = (
'gif jpg jpeg png bmp ico svg ttf eot woff class swf css js xml robots.txt'
).split()
mattpiwik
a validé
'7z aac arc arj asf asx avi bin csv deb dmg doc exe flv gz gzip hqx '
'jar mpg mp2 mp3 mp4 mpeg mov movie msi msp odb odf odg odp '
'ods odt ogg ogv pdf phps ppt qt qtm ra ram rar rpm sea sit tar tbz '
mattab
a validé
'bz2 tbz tgz torrent txt wav wma wmv wpd xls xml xsd z zip'
# A good source is: http://phpbb-bots.blogspot.com/
EXCLUDED_USER_AGENTS = (
'bot-',
'bot/',
'ccooter/',
mattpiwik
a validé
'crawl',
mattpiwik
a validé
'feed',
'ia_archiver',
'java/',
mattpiwik
a validé
'libwww',
'robot',
'surveybot',
'twiceler',
'voilabot',
'yahoo',
'yandex',
)
PIWIK_MAX_ATTEMPTS = 3
PIWIK_DELAY_AFTER_FAILURE = 2
PIWIK_EXPECTED_IMAGE = base64.b64decode(
'R0lGODlhAQABAIAAAAAAAAAAACH5BAEAAAAALAAAAAABAAEAAAICRAEAOw=='
)
##
## Formats.
##
class BaseFormatException(Exception): pass
class BaseFormat(object):
def __init__(self, name):
self.date_format = '%d/%b/%Y:%H:%M:%S'
def check_format(self, file):
line = file.readline()
file.seek(0)
diosmosis
a validé
return self.check_format_line(line)
def check_format_line(self, line):
return False
class JsonFormat(BaseFormat):
def __init__(self, name):
self.json = None
self.date_format = '%Y-%m-%dT%H:%M:%S'
def check_format_line(self, line):
try:
return True
except:
return False
def match(self, line):
try:
self.json = json.loads(line)
return self
except:
self.json = None
return None
# Some ugly patchs ...
if key == 'generation_time_milli':
self.json[key] = int(self.json[key] * 1000)
# Patch date format ISO 8601
elif key == 'date':
tz = self.json[key][19:]
self.json['timezone'] = tz.replace(':', '')
self.json[key] = self.json[key][:19]
try:
return self.json[key]
except KeyError:
raise BaseFormatException()
def get_all(self,):
return self.json
class RegexFormat(BaseFormat):
def __init__(self, name, regex, date_format=None):
if regex is not None:
self.regex = re.compile(regex)
if date_format is not None:
self.date_format = date_format
diosmosis
a validé
def check_format_line(self, line):
diosmosis
a validé
def match(self,line):
self.matched = self.regex.match(line)
return self.matched
def get(self, key):
try:
return self.matched.group(key)
except IndexError:
raise BaseFormatException()
def get_all(self,):
return self.matched.groupdict()
diosmosis
a validé
class IisFormat(RegexFormat):
diosmosis
a validé
def __init__(self):
super(IisFormat, self).__init__('iis', None, '%Y-%m-%d %H:%M:%S')
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
def check_format(self, file):
line = file.readline()
if not line.startswith('#Software: Microsoft Internet Information Services '):
file.seek(0)
return
# Skip the next 2 lines.
for i in xrange(2):
file.readline()
# Parse the 4th line (regex)
full_regex = []
line = file.readline()
fields = {
'date': '(?P<date>^\d+[-\d+]+',
'time': '[\d+:]+)',
'cs-uri-stem': '(?P<path>/\S*)',
'cs-uri-query': '(?P<query_string>\S*)',
'c-ip': '(?P<ip>[\d*.]*)',
'cs(User-Agent)': '(?P<user_agent>\S+)',
'cs(Referer)': '(?P<referrer>\S+)',
'sc-status': '(?P<status>\d+)',
'sc-bytes': '(?P<length>\S+)',
'cs-host': '(?P<host>\S+)',
}
# Skip the 'Fields: ' prefix.
line = line[9:]
for field in line.split():
try:
regex = fields[field]
except KeyError:
regex = '\S+'
full_regex.append(regex)
diosmosis
a validé
self.regex = re.compile(' '.join(full_regex))
diosmosis
a validé
start_pos = file.tell()
nextline = file.readline()
file.seek(start_pos)
return self.check_format_line(nextline)
_HOST_PREFIX = '(?P<host>[\w\-\.]*)(?::\d+)? '
_COMMON_LOG_FORMAT = (
'(?P<ip>\S+) \S+ \S+ \[(?P<date>.*?) (?P<timezone>.*?)\] '
'"\S+ (?P<path>.*?) \S+" (?P<status>\S+) (?P<length>\S+)'
)
_NCSA_EXTENDED_LOG_FORMAT = (_COMMON_LOG_FORMAT +
' "(?P<referrer>.*?)" "(?P<user_agent>.*?)"'
)
_S3_LOG_FORMAT = (
'\S+ (?P<host>\S+) \[(?P<date>.*?) (?P<timezone>.*?)\] (?P<ip>\S+) '
'\S+ \S+ \S+ \S+ "\S+ (?P<path>.*?) \S+" (?P<status>\S+) \S+ (?P<length>\S+) '
diosmosis
a validé
'\S+ \S+ \S+ "(?P<referrer>.*?)" "(?P<user_agent>.*?)"'
_ICECAST2_LOG_FORMAT = ( _NCSA_EXTENDED_LOG_FORMAT +
' (?P<session_time>\S+)'
)
FORMATS = {
'common': RegexFormat('common', _COMMON_LOG_FORMAT),
'common_vhost': RegexFormat('common_vhost', _HOST_PREFIX + _COMMON_LOG_FORMAT),
'ncsa_extended': RegexFormat('ncsa_extended', _NCSA_EXTENDED_LOG_FORMAT),
'common_complete': RegexFormat('common_complete', _HOST_PREFIX + _NCSA_EXTENDED_LOG_FORMAT),
'iis': IisFormat(),
's3': RegexFormat('s3', _S3_LOG_FORMAT),
'icecast2': RegexFormat('icecast2', _ICECAST2_LOG_FORMAT),
'nginx_json': JsonFormat('nginx_json'),
##
## Code.
##
class Configuration(object):
"""
Stores all the configuration options by reading sys.argv and parsing,
if needed, the config.inc.php.
It has 2 attributes: options and filenames.
"""
class Error(Exception):
pass
def _create_parser(self):
"""
Initialize and return the OptionParser instance.
"""
option_parser = optparse.OptionParser(
usage='Usage: %prog [options] log_file [ log_file [...] ]',
"log_file is the path to a server access log file (uncompressed, .gz, .bz2, or specify - to read from stdin). "
" By default, the script will try to produce clean reports and will exclude bots, static files, discard http error and redirects, etc. This is customizable, see below.",
epilog="About Piwik Server Log Analytics: http://piwik.org/log-analytics/ "
" Found a bug? Please create a ticket in http://dev.piwik.org/ "
" Please send your suggestions or successful user story to hello@piwik.org "
)
option_parser.add_option(
'--debug', '-d', dest='debug', action='count', default=0,
help="Enable debug output (specify multiple times for more verbose)",
)
option_parser.add_option(
'--url', dest='piwik_url',
help="REQUIRED Piwik base URL, eg. http://example.com/piwik/ or http://analytics.example.net",
)
option_parser.add_option(
'--dry-run', dest='dry_run',
action='store_true', default=False,
help="Perform a trial run with no tracking data being inserted into Piwik",
)
option_parser.add_option(
'--show-progress', dest='show_progress',
action='store_true', default=os.isatty(sys.stdout.fileno()),
help="Print a progress report X seconds (default: 1, use --show-progress-delay to override)"
)
option_parser.add_option(
'--show-progress-delay', dest='show_progress_delay',
type='int', default=1,
help="Change the default progress delay"
)
option_parser.add_option(
'--add-sites-new-hosts', dest='add_sites_new_hosts',
action='store_true', default=False,
help="When a hostname is found in the log file, but not matched to any website "
"in Piwik, automatically create a new website in Piwik with this hostname to "
"import the logs"
)
option_parser.add_option(
'--idsite', dest='site_id',
help= ("When specified, "
"data in the specified log files will be tracked for this Piwik site ID."
" The script will not auto-detect the website based on the log line hostname (new websites will not be automatically created).")
)
option_parser.add_option(
'--idsite-fallback', dest='site_id_fallback',
help="Default Piwik site ID to use if the hostname doesn't match any "
"known Website's URL. New websites will not be automatically created. "
" Used only if --add-sites-new-hosts or --idsite are not set",
default_config = os.path.abspath(
os.path.join(os.path.dirname(__file__),
'../../config/config.ini.php'),
)
'--config', dest='config_file', default=default_config,
"This is only used when --login and --password is not used. "
"fetch the Super User token_auth from the config file. "
)
)
option_parser.add_option(
'--login', dest='login',
help="You can manually specify the Piwik Super User login"
)
option_parser.add_option(
'--password', dest='password',
help="You can manually specify the Piwik Super User password"
)
option_parser.add_option(
'--token-auth', dest='piwik_token_auth',
help="Piwik Super User token_auth, 32 characters hexadecimal string, found in Piwik > API",
'--hostname', dest='hostnames', action='append', default=[],
help="Accepted hostname (requests with other hostnames will be excluded). "
option_parser.add_option(
'--exclude-path', dest='excluded_paths', action='append', default=[],
help="Paths to exclude. Can be specified multiple times"
)
option_parser.add_option(
'--exclude-path-from', dest='exclude_path_from',
help="Each line from this file is a path to exclude"
)
option_parser.add_option(
'--include-path', dest='included_paths', action='append', default=[],
help="Paths to include. Can be specified multiple times. If not specified, all paths are included."
)
option_parser.add_option(
'--include-path-from', dest='include_path_from',
help="Each line from this file is a path to include"
)
option_parser.add_option(
'--useragent-exclude', dest='excluded_useragents',
action='append', default=[],
help="User agents to exclude (in addition to the standard excluded "
"user agents). Can be specified multiple times",
)
option_parser.add_option(
'--enable-static', dest='enable_static',
action='store_true', default=False,
help="Track static files (images, css, js, etc.)"
option_parser.add_option(
'--enable-bots', dest='enable_bots',
action='store_true', default=False,
help="Track bots. All bot visits will have a Custom Variable set with name='Bot' and value='$Bot_user_agent_here$'"
option_parser.add_option(
'--enable-http-errors', dest='enable_http_errors',
action='store_true', default=False,
help="Track HTTP errors (status code 4xx or 5xx)"
)
option_parser.add_option(
'--enable-http-redirects', dest='enable_http_redirects',
action='store_true', default=False,
help="Track HTTP redirects (status code 3xx except 304)"
)
option_parser.add_option(
'--enable-reverse-dns', dest='reverse_dns',
action='store_true', default=False,
help="Enable reverse DNS, used to generate the 'Providers' report in Piwik. "
"Disabled by default, as it impacts performance"
'--strip-query-string', dest='strip_query_string',
help="Strip the query string from the URL"
option_parser.add_option(
'--query-string-delimiter', dest='query_string_delimiter', default='?',
help="The query string delimiter (default: %default)"
)
'--log-format-name', dest='log_format_name', default=None,
help=("Access log format to detect (supported are: %s). "
"When not specified, the log format will be autodetected by trying all supported log formats."
% ', '.join(sorted(FORMATS.iterkeys())))
)
option_parser.add_option(
'--log-format-regex', dest='log_format_regex', default=None,
help="Access log regular expression. For an example of a supported Regex, see the source code of this file. "
"Overrides --log-format-name"
option_parser.add_option(
'--log-hostname', dest='log_hostname', default=None,
help="Force this hostname for a log format that doesn't incldude it. All hits "
"will seem to came to this host"
)
'--skip', dest='skip', default=0, type='int',
help="Skip the n first lines to start parsing/importing data at a given line for the specified log file",
)
option_parser.add_option(
'--recorders', dest='recorders', default=1, type='int',
help="Number of simultaneous recorders (default: %default). "
"It should be set to the number of CPU cores in your server. "
"You can also experiment with higher values which may increase performance until a certain point",
)
option_parser.add_option(
benakamoorthi
a validé
'--recorder-max-payload-size', dest='recorder_max_payload_size', default=200, type='int',
help="Maximum number of log entries to record in one tracking request (default: %default). "
)
option_parser.add_option(
'--replay-tracking', dest='replay_tracking',
action='store_true', default=False,
help="Replay piwik.php requests found in custom logs (only piwik.php requests expected). \nSee http://piwik.org/faq/how-to/faq_17033/"
option_parser.add_option(
'--output', dest='output',
help="Redirect output (stdout and stderr) to the specified file"
option_parser.add_option(
'--encoding', dest='encoding', default='utf8',
help="Log files encoding (default: %default)"
)
benakamoorthi
a validé
option_parser.add_option(
'--disable-bulk-tracking', dest='use_bulk_tracking',
default=True, action='store_false',
help="Disables use of bulk tracking so recorders record one hit at a time."
)
option_parser.add_option(
'--debug-force-one-hit-every-Ns', dest='force_one_action_interval', default=False, type='float',
help="Debug option that will force each recorder to record one hit every N secs."
)
option_parser.add_option(
'--invalidate-dates', dest='invalidate_dates', default=None,
help="Invalidate reports for the specified dates (format: YYYY-MM-DD,YYYY-MM-DD,...). "
"By default, all dates found in the logs will be invalidated.",
)
option_parser.add_option(
'--force-lowercase-path', dest='force_lowercase_path', default=False, action='store_true',
help="Make URL path lowercase so paths with the same letters but different cases are "
"treated the same."
)
option_parser.add_option(
'--enable-testmode', dest='enable_testmode', default=False, action='store_true',
help="If set, it will try to get the token_auth from the piwik_tests directory"
)
return option_parser
def _parse_args(self, option_parser):
"""
Parse the command line args and create self.options and self.filenames.
"""
self.options, self.filenames = option_parser.parse_args(sys.argv[1:])
if self.options.output:
sys.stdout = sys.stderr = open(self.options.output, 'a+', 0)
if not self.filenames:
print(option_parser.format_help())
sys.exit(1)
# Configure logging before calling logging.{debug,info}.
logging.basicConfig(
format='%(asctime)s: [%(levelname)s] %(message)s',
level=logging.DEBUG if self.options.debug >= 1 else logging.INFO,
self.options.excluded_useragents = [s.lower() for s in self.options.excluded_useragents]
if self.options.exclude_path_from:
paths = [path.strip() for path in open(self.options.exclude_path_from).readlines()]
self.options.excluded_paths.extend(path for path in paths if len(path) > 0)
if self.options.excluded_paths:
logging.debug('Excluded paths: %s', ' '.join(self.options.excluded_paths))
if self.options.include_path_from:
paths = [path.strip() for path in open(self.options.include_path_from).readlines()]
self.options.included_paths.extend(path for path in paths if len(path) > 0)
if self.options.included_paths:
logging.debug('Included paths: %s', ' '.join(self.options.included_paths))
logging.debug('Accepted hostnames: %s', ', '.join(self.options.hostnames))
else:
logging.debug('Accepted hostnames: all')
self.format = RegexFormat('custom', self.options.log_format_regex)
self.format = FORMATS[self.options.log_format_name]
fatal_error('invalid log format: %s' % self.options.log_format_name)
if not self.options.piwik_url:
fatal_error('no URL given for Piwik')
if not (self.options.piwik_url.startswith('http://') or self.options.piwik_url.startswith('https://')):
self.options.piwik_url = 'http://' + self.options.piwik_url
logging.debug('Piwik URL is: %s', self.options.piwik_url)
if not self.options.piwik_token_auth:
try:
self.options.piwik_token_auth = self._get_token_auth()
except Piwik.Error, e:
fatal_error(e)
logging.debug('Authentication token token_auth is: %s', self.options.piwik_token_auth)
if self.options.recorders < 1:
self.options.recorders = 1
def __init__(self):
self._parse_args(self._create_parser())
def _get_token_auth(self):
"""
If the token auth is not specified in the options, get it from Piwik.
"""
# Get superuser login/password from the options.
logging.debug('No token-auth specified')
if self.options.login and self.options.password:
piwik_login = self.options.login
piwik_password = hashlib.md5(self.options.password).hexdigest()
else:
# Fallback to the given (or default) configuration file, then
# get the token from the API.
logging.debug(
'No credentials specified, reading them from "%s"',
self.options.config_file,
)
config_file = ConfigParser.RawConfigParser()
success = len(config_file.read(self.options.config_file)) > 0
if not success:
fatal_error(
"couldn't open the configuration file, "
"required to get the authentication token"
)
Thomas Steur
a validé
updatetokenfile = os.path.abspath(
os.path.join(os.path.dirname(__file__),
'../../misc/cron/updatetoken.php'),
process = "php " + updatetokenfile
if self.options.enable_testmode:
process = process + " --testmode"
filename = subprocess.check_output(process, shell=True);
Thomas Steur
a validé
credentials = open(filename, 'r').readline()
credentials = credentials.split('\t')
return credentials[1]
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
def get_resolver(self):
if self.options.site_id:
logging.debug('Resolver: static')
return StaticResolver(self.options.site_id)
else:
logging.debug('Resolver: dynamic')
return DynamicResolver()
class Statistics(object):
"""
Store statistics about parsed logs and recorded entries.
Can optionally print statistics on standard output every second.
"""
class Counter(object):
"""
Simple integers cannot be used by multithreaded programs. See:
http://stackoverflow.com/questions/6320107/are-python-ints-thread-safe
"""
def __init__(self):
# itertools.count's implementation in C does not release the GIL and
# therefore is thread-safe.
self.counter = itertools.count(1)
self.value = 0
def increment(self):
self.value = self.counter.next()
def advance(self, n):
for i in range(n):
self.increment()
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
def __str__(self):
return str(int(self.value))
def __init__(self):
self.time_start = None
self.time_stop = None
self.piwik_sites = set() # sites ID
self.piwik_sites_created = [] # (hostname, site ID)
self.piwik_sites_ignored = set() # hostname
self.count_lines_parsed = self.Counter()
self.count_lines_recorded = self.Counter()
# Do not match the regexp.
self.count_lines_invalid = self.Counter()
# No site ID found by the resolver.
self.count_lines_no_site = self.Counter()
# Hostname filtered by config.options.hostnames
self.count_lines_hostname_skipped = self.Counter()
# Static files.
self.count_lines_static = self.Counter()
# Ignored user-agents.
self.count_lines_skipped_user_agent = self.Counter()
# Ignored HTTP erors.
self.count_lines_skipped_http_errors = self.Counter()
# Ignored HTTP redirects.
self.count_lines_skipped_http_redirects = self.Counter()
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
# Downloads
self.count_lines_downloads = self.Counter()
# Misc
self.dates_recorded = set()
self.monitor_stop = False
def set_time_start(self):
self.time_start = time.time()
def set_time_stop(self):
self.time_stop = time.time()
def _compute_speed(self, value, start, end):
delta_time = end - start
if value == 0:
return 0
if delta_time == 0:
return 'very high!'
else:
return value / delta_time
def _round_value(self, value, base=100):
return round(value * base) / base
def _indent_text(self, lines, level=1):
"""
Return an indented text. 'lines' can be a list of lines or a single
line (as a string). One level of indentation is 4 spaces.
"""
prefix = ' ' * (4 * level)
if isinstance(lines, basestring):
return prefix + lines
else:
return '\n'.join(
prefix + line
for line in lines
)
def print_summary(self):
print '''
Logs import summary
-------------------
%(count_lines_recorded)d requests imported successfully
%(count_lines_downloads)d requests were downloads
%(total_lines_ignored)d requests ignored:
%(count_lines_invalid)d invalid log lines
%(count_lines_skipped_user_agent)d requests done by bots, search engines, ...
%(count_lines_skipped_http_errors)d HTTP errors
%(count_lines_skipped_http_redirects)d HTTP redirects
%(count_lines_static)d requests to static resources (css, js, ...)
%(count_lines_no_site)d requests did not match any known site
%(count_lines_hostname_skipped)d requests did not match any requested hostname
Website import summary
----------------------
%(count_lines_recorded)d requests imported to %(total_sites)d sites
%(total_sites_existing)d sites already existed
%(total_sites_created)d sites were created:
%(sites_created)s
%(total_sites_ignored)d distinct hostnames did not match any existing site:
%(sites_ignored)s
Performance summary
-------------------
Total time: %(total_time)d seconds
Requests imported per second: %(speed_recording)s requests per second
''' % {
'count_lines_recorded': self.count_lines_recorded.value,
'count_lines_downloads': self.count_lines_downloads.value,
'total_lines_ignored': sum([
self.count_lines_invalid.value,
self.count_lines_skipped_user_agent.value,
self.count_lines_skipped_http_errors.value,
self.count_lines_static.value,
self.count_lines_no_site.value,
self.count_lines_hostname_skipped.value,
]),
'count_lines_invalid': self.count_lines_invalid.value,
'count_lines_skipped_user_agent': self.count_lines_skipped_user_agent.value,
'count_lines_skipped_http_errors': self.count_lines_skipped_http_errors.value,
'count_lines_skipped_http_redirects': self.count_lines_skipped_http_redirects.value,
'count_lines_static': self.count_lines_static.value,
'count_lines_no_site': self.count_lines_no_site.value,
'count_lines_hostname_skipped': self.count_lines_hostname_skipped.value,
'total_sites': len(self.piwik_sites),
'total_sites_existing': len(self.piwik_sites - set(site_id for hostname, site_id in self.piwik_sites_created)),
'total_sites_created': len(self.piwik_sites_created),
'sites_created': self._indent_text(
['%s (ID: %d)' % (hostname, site_id) for hostname, site_id in self.piwik_sites_created],
level=3,
),
'total_sites_ignored': len(self.piwik_sites_ignored),
'sites_ignored': self._indent_text(
self.piwik_sites_ignored, level=3,
),
'sites_ignored_tips': '''
TIPs:
- if one of these hosts is an alias host for one of the websites
in Piwik, you can add this host as an "Alias URL" in Settings > Websites.
- use --add-sites-new-hosts if you wish to automatically create
one website for each of these hosts in Piwik rather than discarding
these requests.
- use --idsite-fallback to force all these log lines with a new hostname
to be recorded in a specific idsite (for example for troubleshooting/visualizing the data)
- use --idsite to force all lines in the specified log files
to be all recorded in the specified idsite
- or you can also manually create a new Website in Piwik with the URL set to this hostname
''' if self.piwik_sites_ignored else '',
'total_time': self.time_stop - self.time_start,
'speed_recording': self._round_value(self._compute_speed(
self.count_lines_recorded.value,
self.time_start, self.time_stop,
)),
}
##
## The monitor is a thread that prints a short summary each second.
##
def _monitor(self):
latest_total_recorded = 0
while not self.monitor_stop:
current_total = stats.count_lines_recorded.value
benakamoorthi
a validé
time_elapsed = time.time() - self.time_start
benakamoorthi
a validé
print '%d lines parsed, %d lines recorded, %d records/sec (avg), %d records/sec (current)' % (
stats.count_lines_parsed.value,
current_total,
benakamoorthi
a validé
current_total / time_elapsed if time_elapsed != 0 else 0,
(current_total - latest_total_recorded) / config.options.show_progress_delay,
)
latest_total_recorded = current_total
time.sleep(config.options.show_progress_delay)
def start_monitor(self):
t = threading.Thread(target=self._monitor)
t.daemon = True
t.start()
def stop_monitor(self):
self.monitor_stop = True
class Piwik(object):
"""
Make requests to Piwik.
"""
class Error(Exception):
pass
@staticmethod
def _call(path, args, headers=None, url=None, data=None):
"""
Make a request to the Piwik site. It is up to the caller to format
arguments, to embed authentication, etc.
"""
if url is None:
url = config.options.piwik_url
headers = headers or {}
if data is None:
# If Content-Type isn't defined, PHP do not parse the request's body.
headers['Content-type'] = 'application/x-www-form-urlencoded'
data = urllib.urlencode(args)
elif not isinstance(data, basestring) and headers['Content-type'] == 'application/json':
data = json.dumps(data)
request = urllib2.Request(url + path, data, headers)
response = urllib2.urlopen(request)
result = response.read()
response.close()
return result
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
@staticmethod
def _call_api(method, **kwargs):
"""
Make a request to the Piwik API taking care of authentication, body
formatting, etc.
"""
args = {
'module' : 'API',
'format' : 'json',
'method' : method,
}
# token_auth, by default, is taken from config.
token_auth = kwargs.pop('_token_auth', None)
if token_auth is None:
token_auth = config.options.piwik_token_auth
if token_auth:
args['token_auth'] = token_auth
url = kwargs.pop('_url', None)
if kwargs:
args.update(kwargs)
# Convert lists into appropriate format.
# See: http://dev.piwik.org/trac/wiki/API/Reference#PassinganArrayParameter
# Warning: we have to pass the parameters in order: foo[0], foo[1], foo[2]
# and not foo[1], foo[0], foo[2] (it will break Piwik otherwise.)
final_args = []
for key, value in args.iteritems():
if isinstance(value, (list, tuple)):
for index, obj in enumerate(value):
final_args.append(('%s[%d]' % (key, index), obj))
else:
final_args.append((key, value))
res = Piwik._call('/', final_args, url=url)
try:
return json.loads(res)
except ValueError:
raise urllib2.URLError('Piwik returned an invalid response: ' + res[:truncate_after])
@staticmethod
def _call_wrapper(func, expected_response, on_failure, *args, **kwargs):
"""
Try to make requests to Piwik at most PIWIK_FAILURE_MAX_RETRY times.
"""
errors = 0
while True:
try:
response = func(*args, **kwargs)
if expected_response is not None and response != expected_response:
if on_failure is not None:
error_message = on_failure(response, kwargs.get('data'))
else:
truncated_response = (response[:truncate_after] + '..') if len(response) > truncate_after else response
error_message = "didn't receive the expected response. Response was %s " % truncated_response
raise urllib2.URLError(error_message)
except (urllib2.URLError, httplib.HTTPException, ValueError), e:
logging.debug('Error when connecting to Piwik: %s', e)
errors += 1
if errors == PIWIK_MAX_ATTEMPTS:
if isinstance(e, urllib2.HTTPError):
# See Python issue 13211.
message = e.msg
elif isinstance(e, urllib2.URLError):
message = e.reason
else:
message = str(e)
raise Piwik.Error(message)
else:
time.sleep(PIWIK_DELAY_AFTER_FAILURE)
@classmethod
def call(cls, path, args, expected_content=None, headers=None, data=None, on_failure=None):
return cls._call_wrapper(cls._call, expected_content, on_failure, path, args, headers,
@classmethod
def call_api(cls, method, **kwargs):
return cls._call_wrapper(cls._call_api, None, None, method, **kwargs)
##
## Resolvers.
##
## A resolver is a class that turns a hostname into a Piwik site ID.
##
class StaticResolver(object):
"""
Always return the same site ID, specified in the configuration.
"""
def __init__(self, site_id):
self.site_id = site_id
# Go get the main URL
sites = piwik.call_api(
'SitesManager.getSiteFromId', idSite=self.site_id
)
try:
site = sites[0]
except (IndexError, KeyError):
benakamoorthi
a validé
logging.debug('response for SitesManager.getSiteFromId: %s', str(sites))
fatal_error(
"cannot get the main URL of this site: invalid site ID: %s" % site_id
)
if site.get('result') == 'error':
fatal_error(
"cannot get the main URL of this site: %s" % site.get('message')
)
self._main_url = site['main_url']
stats.piwik_sites.add(self.site_id)
def resolve(self, hit):
return (self.site_id, self._main_url)
def check_format(self, format):
pass
class DynamicResolver(object):
"""
Use Piwik API to determine the site ID.
"""
benakamoorthi
a validé
_add_site_lock = threading.Lock()
def __init__(self):
self._cache = {}
if config.options.replay_tracking:
# get existing sites
self._cache['sites'] = piwik.call_api('SitesManager.getAllSites')
benakamoorthi
a validé
def _get_site_id_from_hit_host(self, hit):
main_url = 'http://' + hit.host
benakamoorthi
a validé
return piwik.call_api(
'SitesManager.getSitesIdFromSiteUrl',
url=main_url,
)