X-Git-Url: https://arthur.barton.de/gitweb/?a=blobdiff_plain;f=python.d%2Fweb_log.chart.py;h=c8b43595e10b0a0b7e1563fc13e6d9eac15b152e;hb=e3c740c2bd0a54ab43af91338d4bcf3ef2574282;hp=fe1f140af09350d4fcde20c0cfbfacb6dca45817;hpb=e5f7b43e7ae28bfed31ab0c25359de684685ff86;p=netdata.git diff --git a/python.d/web_log.chart.py b/python.d/web_log.chart.py index fe1f140a..c8b43595 100644 --- a/python.d/web_log.chart.py +++ b/python.d/web_log.chart.py @@ -7,63 +7,72 @@ import re import bisect from os import access, R_OK from os.path import getsize -from collections import defaultdict, namedtuple +from collections import namedtuple from copy import deepcopy -try: - from itertools import zip_longest -except ImportError: - from itertools import izip_longest as zip_longest priority = 60000 retries = 60 -ORDER = ['response_codes', 'response_time', 'requests_per_url', 'http_method', 'requests_per_ipproto', 'bandwidth', 'clients', 'clients_all'] +ORDER = ['response_statuses', 'response_codes', 'bandwidth', 'response_time', 'requests_per_url', 'http_method', + 'requests_per_ipproto', 'clients', 'clients_all'] CHARTS = { 'response_codes': { 'options': [None, 'Response Codes', 'requests/s', 'responses', 'web_log.response_codes', 'stacked'], 'lines': [ - ['2xx', '2xx', 'absolute'], - ['5xx', '5xx', 'absolute'], - ['3xx', '3xx', 'absolute'], - ['4xx', '4xx', 'absolute'], - ['1xx', '1xx', 'absolute'], - ['0xx', 'other', 'absolute'], - ['unmatched', 'unmatched', 'absolute'] + ['2xx', '2xx', 'incremental'], + ['5xx', '5xx', 'incremental'], + ['3xx', '3xx', 'incremental'], + ['4xx', '4xx', 'incremental'], + ['1xx', '1xx', 'incremental'], + ['0xx', 'other', 'incremental'], + ['unmatched', 'unmatched', 'incremental'] ]}, 'bandwidth': { 'options': [None, 'Bandwidth', 'KB/s', 'bandwidth', 'web_log.bandwidth', 'area'], 'lines': [ - ['resp_length', 'received', 'absolute', 1, 1024], - ['bytes_sent', 'sent', 'absolute', -1, 1024] + ['resp_length', 'received', 'incremental', 1, 1024], + ['bytes_sent', 'sent', 'incremental', -1, 1024] ]}, 'response_time': { 'options': [None, 'Processing Time', 'milliseconds', 'timings', 'web_log.response_time', 'area'], 'lines': [ - ['resp_time_min', 'min', 'absolute', 1, 1], - ['resp_time_max', 'max', 'absolute', 1, 1], - ['resp_time_avg', 'avg', 'absolute', 1, 1] + ['resp_time_min', 'min', 'incremental', 1, 1000], + ['resp_time_max', 'max', 'incremental', 1, 1000], + ['resp_time_avg', 'avg', 'incremental', 1, 1000] ]}, 'clients': { - 'options': [None, 'Current Poll Unique Client IPs', 'unique ips', 'unique clients', 'web_log.clients', 'stacked'], + 'options': [None, 'Current Poll Unique Client IPs', 'unique ips', 'clients', 'web_log.clients', 'stacked'], 'lines': [ - ['unique_cur_ipv4', 'ipv4', 'absolute', 1, 1], - ['unique_cur_ipv6', 'ipv6', 'absolute', 1, 1] + ['unique_cur_ipv4', 'ipv4', 'incremental', 1, 1], + ['unique_cur_ipv6', 'ipv6', 'incremental', 1, 1] ]}, 'clients_all': { - 'options': [None, 'All Time Unique Client IPs', 'unique ips', 'unique clients', 'web_log.clients_all', 'stacked'], + 'options': [None, 'All Time Unique Client IPs', 'unique ips', 'clients', 'web_log.clients_all', 'stacked'], 'lines': [ ['unique_tot_ipv4', 'ipv4', 'absolute', 1, 1], ['unique_tot_ipv6', 'ipv6', 'absolute', 1, 1] ]}, 'http_method': { - 'options': [None, 'Requests Per HTTP Method', 'requests/s', 'requests', 'web_log.http_method', 'stacked'], + 'options': [None, 'Requests Per HTTP Method', 'requests/s', 'http methods', 'web_log.http_method', 'stacked'], 'lines': [ + ['GET', 'GET', 'incremental', 1, 1] ]}, 'requests_per_ipproto': { - 'options': [None, 'Requests Per IP Protocol', 'requests/s', 'requests', 'web_log.requests_per_ipproto', 'stacked'], + 'options': [None, 'Requests Per IP Protocol', 'requests/s', 'ip protocols', 'web_log.requests_per_ipproto', + 'stacked'], 'lines': [ - ['req_ipv4', 'ipv4', 'absolute', 1, 1], - ['req_ipv6', 'ipv6', 'absolute', 1, 1] + ['req_ipv4', 'ipv4', 'incremental', 1, 1], + ['req_ipv6', 'ipv6', 'incremental', 1, 1] + ]}, + 'response_statuses': { + 'options': [None, 'Response Statuses', 'requests/s', 'responses', 'web_log.response_statuses', + 'stacked'], + 'lines': [ + ['successful_requests', 'success', 'incremental', 1, 1], + ['server_errors', 'error', 'incremental', 1, 1], + ['redirects', 'redirect', 'incremental', 1, 1], + ['bad_requests', 'bad', 'incremental', 1, 1], + ['other_requests', 'other', 'incremental', 1, 1] ]} } @@ -72,34 +81,44 @@ NAMED_URL_PATTERN = namedtuple('URL_PATTERN', ['description', 'pattern']) class Service(LogService): def __init__(self, configuration=None, name=None): + """ + :param configuration: + :param name: + # self._get_data = None # will be assigned in 'check' method. + # self.order = None # will be assigned in 'create_*_method' method. + # self.definitions = None # will be assigned in 'create_*_method' method. + """ LogService.__init__(self, configuration=configuration, name=name) # Variables from module configuration file + self.type = self.configuration.get('type', 'web_access') self.log_path = self.configuration.get('path') - self.detailed_response_codes = self.configuration.get('detailed_response_codes', True) - self.all_time = self.configuration.get('all_time', True) self.url_pattern = self.configuration.get('categories') # dict - self.regex = None - # sorted list of unique IPs - self.unique_all_time = list() - # dict for values that should not be zeroed every poll - self.storage = {'unique_tot_ipv4': 0, 'unique_tot_ipv6': 0} - # if there is no new logs this dict + self.storage returned to netdata - self.data = {'bytes_sent': 0, 'resp_length': 0, 'resp_time_min': 0, - 'resp_time_max': 0, 'resp_time_avg': 0, 'unique_cur_ipv4': 0, - 'unique_cur_ipv6': 0, '2xx': 0, '5xx': 0, '3xx': 0, '4xx': 0, - '1xx': 0, '0xx': 0, 'unmatched': 0, 'req_ipv4': 0, 'req_ipv6': 0} + self.custom_log_format = self.configuration.get('custom_log_format') # dict + # Instance variables + self.regex = None # will be assigned in 'find_regex' or 'find_regex_custom' method + self.data = {'bytes_sent': 0, 'resp_length': 0, 'resp_time_min': 0, 'resp_time_max': 0, + 'resp_time_avg': 0, 'unique_cur_ipv4': 0, 'unique_cur_ipv6': 0, '2xx': 0, + '5xx': 0, '3xx': 0, '4xx': 0, '1xx': 0, '0xx': 0, 'unmatched': 0, 'req_ipv4': 0, + 'req_ipv6': 0, 'unique_tot_ipv4': 0, 'unique_tot_ipv6': 0, 'successful_requests': 0, + 'redirects': 0, 'bad_requests': 0, 'server_errors': 0, 'other_requests': 0, 'GET': 0} def check(self): + """ + :return: bool + + 1. "log_path" is specified in the module configuration file + 2. "log_path" must be readable by netdata user and must exist + 3. "log_path' must not be empty. We need at least 1 line to find appropriate pattern to parse + 4. other checks depends on log "type" + """ if not self.log_path: self.error('log path is not specified') return False - # log_path must be readable if not access(self.log_path, R_OK): self.error('%s not readable or not exist' % self.log_path) return False - # log_path file should not be empty if not getsize(self.log_path): self.error('%s is empty' % self.log_path) return False @@ -111,77 +130,220 @@ class Service(LogService): logs.seek(-2, 1) if logs.tell() == 0: break - last_line = logs.readline().decode(encoding='utf-8') + last_line = logs.readline() + + try: + last_line = last_line.decode() + except UnicodeDecodeError: + try: + last_line = last_line.decode(encoding='utf-8') + except (TypeError, UnicodeDecodeError) as error: + self.error(str(error)) + return False + + if self.type == 'web_access': + self.unique_all_time = list() # sorted list of unique IPs + self.detailed_response_codes = self.configuration.get('detailed_response_codes', True) + self.all_time = self.configuration.get('all_time', True) + + # Custom_log_format or predefined log format. + if self.custom_log_format: + match_dict, error = self.find_regex_custom(last_line) + else: + match_dict, error = self.find_regex(last_line) + + # "match_dict" is None if there are any problems + if match_dict is None: + self.error(str(error)) + return False - # Parse last line - parsed_line, regex_name = self.find_regex(last_line) - if not parsed_line: - self.error('Can\'t parse output') + # self.url_pattern check + if self.url_pattern: + self.url_pattern = check_req_per_url_pattern('rpu', self.url_pattern) + + self.create_access_charts(match_dict) # Create charts + self._get_data = self._get_access_data # _get_data assignment + else: + self.error('Not implemented') return False - self.create_charts(parsed_line[0], regex_name) - if len(parsed_line[0]) == 5: - self.info('Not all data collected. You need to modify LogFormat.') + # Double check + if not self.regex: + self.error('That can not happen, but it happened. "regex" is None') + + self.info('Collected data: %s' % list(match_dict.keys())) return True + def find_regex_custom(self, last_line): + """ + :param last_line: str: literally last line from log file + :return: tuple where: + [0]: dict or None: match_dict or None + [1]: str: error description + + We are here only if "custom_log_format" is in logs. We need to make sure: + 1. "custom_log_format" is a dict + 2. "pattern" in "custom_log_format" and pattern is instance + 3. if "time_multiplier" is in "custom_log_format" it must be instance + + If all parameters is ok we need to make sure: + 1. Pattern search is success + 2. Pattern search contains named subgroups (?P) (= "match_dict") + + If pattern search is success we need to make sure: + 1. All mandatory keys ['address', 'code', 'bytes_sent', 'method', 'url'] are in "match_dict" + + If this is True we need to make sure: + 1. All mandatory key values from "match_dict" have the correct format + ("code" is integer, "method" is uppercase word, etc) + + If non mandatory keys in "match_dict" we need to make sure: + 1. All non mandatory key values from match_dict ['resp_length', 'resp_time'] have the correct format + ("resp_length" is integer or "-", "resp_time" is integer or float) + + """ + if not is_dict(self.custom_log_format): + return find_regex_return(msg='Custom log: "custom_log_format" is not a ') + + pattern = self.custom_log_format.get('pattern') + if not (pattern and isinstance(pattern, str)): + return find_regex_return(msg='Custom log: "pattern" option is not specified or type is not ') + + resp_time_func = self.custom_log_format.get('time_multiplier') or 0 + + if not isinstance(resp_time_func, int): + return find_regex_return(msg='Custom log: "time_multiplier" is not an integer') + + try: + regex = re.compile(pattern) + except re.error as error: + return find_regex_return(msg='Pattern compile error: %s' % str(error)) + + match = regex.search(last_line) + if match: + match_dict = match.groupdict() or None + else: + return find_regex_return(msg='Custom log: pattern search FAILED') + + if match_dict is None: + find_regex_return(msg='Custom log: search OK but contains no named subgroups' + ' (you need to use ?P)') + else: + mandatory_dict = {'address': r'[\da-f.:]+', + 'code': r'[1-9]\d{2}', + 'method': r'[A-Z]+', + 'bytes_sent': r'\d+|-'} + optional_dict = {'resp_length': r'\d+', + 'resp_time': r'[\d.]+'} + + mandatory_values = set(mandatory_dict) - set(match_dict) + if mandatory_values: + return find_regex_return(msg='Custom log: search OK but some mandatory keys (%s) are missing' + % list(mandatory_values)) + else: + for key in mandatory_dict: + if not re.search(mandatory_dict[key], match_dict[key]): + return find_regex_return(msg='Custom log: can\'t parse "%s": %s' + % (key, match_dict[key])) + + optional_values = set(optional_dict) & set(match_dict) + for key in optional_values: + if not re.search(optional_dict[key], match_dict[key]): + return find_regex_return(msg='Custom log: can\'t parse "%s": %s' + % (key, match_dict[key])) + + dot_in_time = '.' in match_dict.get('resp_time', '') + if dot_in_time: + self.resp_time_func = lambda time: time * (resp_time_func or 1000000) + else: + self.resp_time_func = lambda time: time * (resp_time_func or 1) + + self.regex = regex + return find_regex_return(match_dict=match_dict) + def find_regex(self, last_line): """ :param last_line: str: literally last line from log file - :return: parsed line, regex name (the one that matches) OR None, None - It's sad but different web servers has different logs formats - We need to find appropriate regex for current log file - All logic is do a regex search through the string for all patterns + :return: tuple where: + [0]: dict or None: match_dict or None + [1]: str: error description + We need to find appropriate pattern for current log file + All logic is do a regex search through the string for all predefined patterns until we find something or fail. """ # REGEX: 1.IPv4 address 2.HTTP method 3. URL 4. Response code # 5. Bytes sent 6. Response length 7. Response process time - default = re.compile(r'([\da-f.:]+)' - r' -.*?"([A-Z]+)' - r' (.*?)"' - r' ([1-9]\d{2})' - r' (\d+)') - - apache_extended = re.compile(r'([\da-f.:]+)' - r' -.*?"([A-Z]+)' - r' (.*?)"' - r' ([1-9]\d{2})' - r' (\d+)' - r' (\d+)' - r' (\d+) ') - - nginx_extended = re.compile(r'([\da-f.:]+)' - r' -.*?"([A-Z]+)' - r' (.*?)"' - r' ([1-9]\d{2})' - r' (\d+)' - r' (\d+)' - r' ([\d.]+) ') - - regex_function = zip([apache_extended, nginx_extended, default], - [lambda x: x, lambda x: x * 1000, lambda x: x], - ['apache_extended', 'nginx_extended', 'default']) - - for regex, function, name in regex_function: - if regex.search(last_line): + acs_default = re.compile(r'(?P
[\da-f.:]+)' + r' -.*?"(?P[A-Z]+)' + r' (?P.*?)"' + r' (?P[1-9]\d{2})' + r' (?P\d+|-)') + + acs_apache_ext_insert = re.compile(r'(?P
[\da-f.:]+)' + r' -.*?"(?P[A-Z]+)' + r' (?P.*?)"' + r' (?P[1-9]\d{2})' + r' (?P\d+|-)' + r' (?P\d+)' + r' (?P\d+) ') + + acs_apache_ext_append = re.compile(r'(?P
[\da-f.:]+)' + r' -.*?"(?P[A-Z]+)' + r' (?P.*?)"' + r' (?P[1-9]\d{2})' + r' (?P\d+|-)' + r' .*?' + r' (?P\d+)' + r' (?P\d+)' + r'(?: |$)') + + acs_nginx_ext_insert = re.compile(r'(?P
[\da-f.:]+)' + r' -.*?"(?P[A-Z]+)' + r' (?P.*?)"' + r' (?P[1-9]\d{2})' + r' (?P\d+)' + r' (?P\d+)' + r' (?P\d\.\d+) ') + + acs_nginx_ext_append = re.compile(r'(?P
[\da-f.:]+)' + r' -.*?"(?P[A-Z]+)' + r' (?P.*?)"' + r' (?P[1-9]\d{2})' + r' (?P\d+)' + r' .*?' + r' (?P\d+)' + r' (?P\d\.\d+)') + + def func_usec(time): + return time + + def func_sec(time): + return time * 1000000 + + r_regex = [acs_apache_ext_insert, acs_apache_ext_append, acs_nginx_ext_insert, + acs_nginx_ext_append, acs_default] + r_function = [func_usec, func_usec, func_sec, func_sec, func_usec] + regex_function = zip(r_regex, r_function) + + match_dict = dict() + for regex, function in regex_function: + match = regex.search(last_line) + if match: self.regex = regex self.resp_time_func = function - regex_name = name + match_dict = match.groupdict() break - if self.regex: - return self.regex.findall(last_line), regex_name - else: - return None, None + return find_regex_return(match_dict=match_dict or None, + msg='Unknown log format. You need to use "custom_log_format" feature.') - def create_charts(self, parsed_line, regex_name): + def create_access_charts(self, match_dict): """ - :param parsed_line: list: re.findall result. - :param regex_name: str: regex name from 'find_regex' method. Ex.: 'apache_extended', 'nginx_extended' + :param match_dict: dict: regex.search.groupdict(). Ex. {'address': '127.0.0.1', 'code': '200', 'method': 'GET'} :return: - Create additional charts depending on the 'find_regex' result (parsed_line) and configuration file - 1. 'time_response' chart is removed if there is no 'time_response' in logs. - 2. We need to change divisor for 'response_time' chart for apache (time in microseconds in logs) - 3. Other stuff is just remove/add chart depending on yes/no in conf + Create additional charts depending on the 'match_dict' keys and configuration file options + 1. 'time_response' chart is removed if there is no 'resp_time' in match_dict. + 2. Other stuff is just remove/add chart depending on yes/no in conf """ def find_job_name(override_name, name): """ @@ -201,19 +363,15 @@ class Service(LogService): job_name = find_job_name(self.override_name, self.name) self.detailed_chart = 'CHART %s.detailed_response_codes ""' \ - ' "Response Codes" requests/s responses' \ - ' web_log.detailed_resp stacked 1 %s\n' % (job_name, self.update_every) + ' "Detailed Response Codes" requests/s responses' \ + ' web_log.detailed_response_codes stacked 1 %s\n' % (job_name, self.update_every) self.http_method_chart = 'CHART %s.http_method' \ - ' "" "HTTP Methods" requests/s requests' \ - ' web_log.http_method stacked 2 %s\n' % (job_name, self.update_every) + ' "" "Requests Per HTTP Method" requests/s "http methods"' \ + ' web_log.http_method stacked 2 %s\n' \ + 'DIMENSION GET GET incremental\n' % (job_name, self.update_every) - if 'apache' in regex_name: - self.definitions['response_time']['lines'][0][4] = 1000 - self.definitions['response_time']['lines'][1][4] = 1000 - self.definitions['response_time']['lines'][2][4] = 1000 - - # Remove 'request_time' chart from ORDER if request_time not in logs - if len(parsed_line) < 7: + # Remove 'request_time' chart from ORDER if resp_time not in match_dict + if 'resp_time' not in match_dict: self.order.remove('response_time') # Remove 'clients_all' chart from ORDER if specified in the configuration if not self.all_time: @@ -222,31 +380,32 @@ class Service(LogService): if self.detailed_response_codes: self.order.append('detailed_response_codes') self.definitions['detailed_response_codes'] = {'options': [None, 'Detailed Response Codes', 'requests/s', - 'responses', 'web_log.detailed_resp', 'stacked'], + 'responses', 'web_log.detailed_response_codes', + 'stacked'], 'lines': []} # Add 'requests_per_url' chart if specified in the configuration if self.url_pattern: - self.url_pattern = [NAMED_URL_PATTERN(description=k, pattern=re.compile(v)) for k, v in self.url_pattern.items()] self.definitions['requests_per_url'] = {'options': [None, 'Requests Per Url', 'requests/s', - 'requests', 'web_log.url_pattern', 'stacked'], - 'lines': [['other_url', 'other', 'absolute']]} + 'urls', 'web_log.requests_per_url', 'stacked'], + 'lines': [['rpu_other', 'other', 'incremental']]} for elem in self.url_pattern: - self.definitions['requests_per_url']['lines'].append([elem.description, elem.description, 'absolute']) + self.definitions['requests_per_url']['lines'].append([elem.description, elem.description[4:], + 'incremental']) self.data.update({elem.description: 0}) - self.data.update({'other_url': 0}) + self.data.update({'rpu_other': 0}) else: self.order.remove('requests_per_url') def add_new_dimension(self, dimension, line_list, chart_string, key): """ :param dimension: str: response status code. Ex.: '202', '499' - :param line_list: list: Ex.: ['202', '202', 'Absolute'] + :param line_list: list: Ex.: ['202', '202', 'incremental'] :param chart_string: Current string we need to pass to netdata to rebuild the chart :param key: str: CHARTS dict key (chart name). Ex.: 'response_time' :return: str: new chart string = previous + new dimensions """ - self.storage.update({dimension: 0}) + self.data.update({dimension: 0}) # SET method check if dim in _dimensions self._dimensions.append(dimension) # UPDATE method do SET only if dim in definitions @@ -256,7 +415,7 @@ class Service(LogService): print(chart) return chart - def _get_data(self): + def _get_access_data(self): """ Parse new log lines :return: dict OR None @@ -269,92 +428,84 @@ class Service(LogService): request_time, unique_current = list(), list() request_counter = {'count': 0, 'sum': 0} - to_netdata = dict() - to_netdata.update(self.data) - default_dict = defaultdict(lambda: 0) - + ip_address_counter = {'unique_cur_ip': 0} for line in raw: match = self.regex.search(line) if match: - match_dict = dict(zip_longest('address method url code sent resp_length resp_time'.split(), match.groups())) + match_dict = match.groupdict() try: code = ''.join([match_dict['code'][0], 'xx']) - to_netdata[code] += 1 + self.data[code] += 1 except KeyError: - to_netdata['0xx'] += 1 + self.data['0xx'] += 1 # detailed response code if self.detailed_response_codes: - self._get_data_detailed_response_codes(match_dict['code'], default_dict) + self._get_data_detailed_response_codes(match_dict['code']) + # response statuses + self._get_data_statuses(match_dict['code']) # requests per url if self.url_pattern: - self._get_data_per_url(match_dict['url'], default_dict) + self._get_data_per_url(match_dict['url']) # requests per http method - self._get_data_http_method(match_dict['method'], default_dict) - - to_netdata['bytes_sent'] += int(match_dict['sent']) - - if match_dict['resp_length'] and match_dict['resp_time']: - to_netdata['resp_length'] += int(match_dict['resp_length']) + self._get_data_http_method(match_dict['method']) + # bandwidth sent + bytes_sent = match_dict['bytes_sent'] if '-' not in match_dict['bytes_sent'] else 0 + self.data['bytes_sent'] += int(bytes_sent) + # request processing time and bandwidth received + if 'resp_length' in match_dict: + self.data['resp_length'] += int(match_dict['resp_length']) + if 'resp_time' in match_dict: resp_time = self.resp_time_func(float(match_dict['resp_time'])) bisect.insort_left(request_time, resp_time) request_counter['count'] += 1 request_counter['sum'] += resp_time # requests per ip proto - if '.' in match_dict['address']: - proto = 'ipv4' - to_netdata['req_ipv4'] += 1 - else: - proto = 'ipv6' - to_netdata['req_ipv6'] += 1 + proto = 'ipv4' if '.' in match_dict['address'] else 'ipv6' + self.data['req_' + proto] += 1 # unique clients ips if address_not_in_pool(self.unique_all_time, match_dict['address'], - self.storage['unique_tot_ipv4'] + self.storage['unique_tot_ipv6']): - self.storage['unique_tot_' + proto] += 1 - if address_not_in_pool(unique_current, match_dict['address'], - to_netdata['unique_cur_ipv4'] + to_netdata['unique_cur_ipv6']): - to_netdata['unique_cur_' + proto] += 1 + self.data['unique_tot_ipv4'] + self.data['unique_tot_ipv6']): + self.data['unique_tot_' + proto] += 1 + if address_not_in_pool(unique_current, match_dict['address'], ip_address_counter['unique_cur_ip']): + self.data['unique_cur_' + proto] += 1 + ip_address_counter['unique_cur_ip'] += 1 else: - to_netdata['unmatched'] += 1 + self.data['unmatched'] += 1 + # timings if request_time: - to_netdata['resp_time_min'] = request_time[0] - to_netdata['resp_time_avg'] = float(request_counter['sum']) / request_counter['count'] - to_netdata['resp_time_max'] = request_time[-1] + self.data['resp_time_min'] += int(request_time[0]) + self.data['resp_time_avg'] += int(round(float(request_counter['sum']) / request_counter['count'])) + self.data['resp_time_max'] += int(request_time[-1]) + return self.data - to_netdata.update(self.storage) - to_netdata.update(default_dict) - return to_netdata - - def _get_data_detailed_response_codes(self, code, default_dict): + def _get_data_detailed_response_codes(self, code): """ :param code: str: CODE from parsed line. Ex.: '202, '499' - :param default_dict: defaultdict :return: Calls add_new_dimension method If the value is found for the first time """ - if code not in self.storage: + if code not in self.data: chart_string_copy = self.detailed_chart - self.detailed_chart = self.add_new_dimension(code, [code, code, 'absolute'], + self.detailed_chart = self.add_new_dimension(code, [code, code, 'incremental'], chart_string_copy, 'detailed_response_codes') - default_dict[code] += 1 + self.data[code] += 1 - def _get_data_http_method(self, method, default_dict): + def _get_data_http_method(self, method): """ :param method: str: METHOD from parsed line. Ex.: 'GET', 'POST' - :param default_dict: defaultdict :return: Calls add_new_dimension method If the value is found for the first time """ - if method not in self.storage: + if method not in self.data: chart_string_copy = self.http_method_chart - self.http_method_chart = self.add_new_dimension(method, [method, method, 'absolute'], + self.http_method_chart = self.add_new_dimension(method, [method, method, 'incremental'], chart_string_copy, 'http_method') - default_dict[method] += 1 + self.data[method] += 1 - def _get_data_per_url(self, url, default_dict): + def _get_data_per_url(self, url): """ :param url: str: URL from parsed line - :param default_dict: defaultdict :return: Scan through string looking for the first location where patterns produce a match for all user defined patterns @@ -362,20 +513,36 @@ class Service(LogService): match = None for elem in self.url_pattern: if elem.pattern.search(url): - default_dict[elem.description] += 1 + self.data[elem.description] += 1 match = True break if not match: - default_dict['other_url'] += 1 + self.data['rpu_other'] += 1 + + def _get_data_statuses(self, code): + """ + :param code: str: response status code. Ex.: '202', '499' + :return: + """ + code_class = code[0] + if code_class == '2' or code == '304' or code_class == '1': + self.data['successful_requests'] += 1 + elif code_class == '3': + self.data['redirects'] += 1 + elif code_class == '4': + self.data['bad_requests'] += 1 + elif code_class == '5': + self.data['server_errors'] += 1 + else: + self.data['other_requests'] += 1 def address_not_in_pool(pool, address, pool_size): """ :param pool: list of ip addresses :param address: ip address - :param pool_size: current size of pool - :return: True if address not pool and False address in pool - If address not in pool function add address to pool. + :param pool_size: current pool size + :return: True if address not in pool. False if address in pool. """ index = bisect.bisect_left(pool, address) if index < pool_size: @@ -387,3 +554,61 @@ def address_not_in_pool(pool, address, pool_size): else: bisect.insort_left(pool, address) return True + + +def find_regex_return(match_dict=None, msg='Generic error message'): + """ + :param match_dict: dict: re.search.groupdict() or None + :param msg: str: error description + :return: tuple: + """ + return match_dict, msg + + +def check_req_per_url_pattern(string, url_pattern): + """ + :param string: str: + :param url_pattern: dict: ex. {'dim1': 'pattern1>', 'dim2': ''} + :return: list of named tuples or None: + We need to make sure all patterns are valid regular expressions + """ + if not is_dict(url_pattern): + return None + + result = list() + + def is_valid_pattern(pattern): + """ + :param pattern: str + :return: re.compile(pattern) or None + """ + if not isinstance(pattern, str): + return False + else: + try: + compile_pattern = re.compile(pattern) + except re.error: + return False + else: + return compile_pattern + + for dimension, regex in url_pattern.items(): + valid_pattern = is_valid_pattern(regex) + if isinstance(dimension, str) and valid_pattern: + result.append(NAMED_URL_PATTERN(description='_'.join([string, dimension]), pattern=valid_pattern)) + + return result or None + + +def is_dict(obj): + """ + :param obj: dict: + :return: True or False + obj can be or + """ + try: + obj.keys() + except AttributeError: + return False + else: + return True