def get_field_name(self, value, full=True): try: field_ = self.all_fields[value] except KeyError: raise BadArgumentError( value, msg='Unknown field "%s"' % value ) if not field_['is_returned']: # Returning this field is not allowed. raise BadArgumentError( value, msg='Field "%s" is not allowed to be returned' % value ) field_name = '%s.%s' % ( field_['namespace'], field_['in_database_name'] ) if full and field_['has_full_version']: # If the param has a full version, that means what matters # is the full string, and not its individual terms. field_name += '.full' return field_name
def get_signatures(self, **kwargs): """Return top crashers by signatures. See http://socorro.readthedocs.org/en/latest/middleware.html#tcbs """ filters = [ ("product", None, "str"), ("version", None, "str"), ("crash_type", "all", "str"), ("to_date", datetimeutil.utc_now(), "datetime"), ("duration", datetime.timedelta(7), "timedelta"), ("os", None, "str"), ("limit", 100, "int"), ("date_range_type", None, "str") ] params = external_common.parse_arguments(filters, kwargs) params.logger = logger # what the twoPeriodTopCrasherComparison() function does is that it # makes a start date from taking the to_date - duration if params.duration > datetime.timedelta(30): raise BadArgumentError('Duration too long. Max 30 days.') with self.get_connection() as connection: return tcbs.twoPeriodTopCrasherComparison(connection, params)
def get(self, **kwargs): filters = [ ("report_date", None, "datetime"), ("report_type", None, "str"), ("product", None, "str"), ("version", None, "str"), ("signature", None, "str"), ("platform", None, "str"), ("min_crashes", 10, "int"), ("min_baseline_diff", 0.005, "float"), ] params = external_common.parse_arguments(filters, kwargs) hits = [] if params['report_type'] == 'interesting-addons': hits = self.interesting_addons(params) elif params['report_type'] == 'interesting-modules': hits = self.interesting_modules(params) elif params['report_type'] == 'interesting-addons-with-version': hits = self.interesting_addons_with_version(params) elif params['report_type'] == 'interesting-modules-with-version': hits = self.interesting_modules_with_version(params) elif params['report_type'] == 'core-counts': hits = self.core_counts(params) else: raise BadArgumentError('report_type', received=report_type) return {'hits': hits, 'total': len(hits)}
def test_mapping(self, mapping): """Verify that a mapping is correct. This function does so by first creating a new, temporary index in elasticsearch using the mapping. It then takes some recent crash reports that are in elasticsearch and tries to insert them in the temporary index. Any failure in any of those steps will raise an exception. If any is raised, that means the mapping is incorrect in some way (either it doesn't validate against elasticsearch's rules, or is not compatible with the data we currently store). If no exception is raised, the mapping is likely correct. This function is to be used in any place that can change the `storage_mapping` field in any Super Search Field. Methods `create_field` and `update_field` use it, see above. """ temp_index = 'socorro_mapping_test' es_connection = self.get_connection() # Import at runtime to avoid dependency circle. from socorro.external.es.index_creator import IndexCreator index_creator = IndexCreator(self.config) try: index_creator.create_index( temp_index, mapping, ) now = datetimeutil.utc_now() last_week = now - datetime.timedelta(days=7) current_indices = self.generate_list_of_indexes(last_week, now) crashes_sample = es_connection.search( index=current_indices, doc_type=self.config.elasticsearch.elasticsearch_doctype, size=self.config.elasticsearch.mapping_test_crash_number, ) crashes = [x['_source'] for x in crashes_sample['hits']['hits']] for crash in crashes: es_connection.index( index=temp_index, doc_type=self.config.elasticsearch.elasticsearch_doctype, body=crash, ) except elasticsearch.exceptions.ElasticsearchException as e: raise BadArgumentError( 'storage_mapping', msg='Indexing existing data in Elasticsearch failed with the ' 'new mapping. Error is: %s' % str(e), ) finally: try: index_creator.get_index_client().delete(temp_index) except elasticsearch.exceptions.NotFoundError: # If the index does not exist (if the index creation failed # for example), we don't need to do anything. pass
def fix_date_parameter(self, parameters): """Correct the date parameter. If there is no date parameter, set default values. Otherwise, make sure there is exactly one lower bound value and one greater bound value. """ default_date_range = datetime.timedelta( days=self.config.search_default_date_range) maximum_date_range = datetime.timedelta( days=self.config.search_maximum_date_range) if not parameters.get('date'): now = datetimeutil.utc_now() lastweek = now - default_date_range parameters['date'] = [] parameters['date'].append( SearchParam('date', lastweek, '>=', 'datetime')) parameters['date'].append( SearchParam('date', now, '<=', 'datetime')) else: lower_than = None greater_than = None for param in parameters['date']: if ('<' in param.operator and (not lower_than or (lower_than and lower_than.value > param.value))): lower_than = param if ('>' in param.operator and (not greater_than or (greater_than and greater_than.value < param.value))): greater_than = param # Remove all the existing parameters so we have exactly # one lower value and one greater value parameters['date'] = [] if not lower_than: # add a lower than that is now lower_than = SearchParam('date', datetimeutil.utc_now(), '<=', 'datetime') if not greater_than: # add a greater than that is lower_than minus the date range greater_than = SearchParam( 'date', lower_than.value - default_date_range, '>=', 'datetime') # Verify the date range is not too big. delta = lower_than.value - greater_than.value if delta > maximum_date_range: raise BadArgumentError( 'date', msg='Date range is bigger than %s days' % self.config.webapi.search_maximum_date_range) parameters['date'].append(lower_than) parameters['date'].append(greater_than)
def get(self, **kwargs): filters = [ ("report_types", None, ["list", "str"]), ("report_type", None, "str"), ("signature", None, "str"), ("start_date", None, "datetime"), ("end_date", None, "datetime"), ("versions", None, ["list", "str"]), ] params = external_common.parse_arguments(filters, kwargs) if not params.get('report_types') and params.get('report_type'): # to support the legacy case individual_report = True report_types = [params['report_type']] else: individual_report = False report_types = params['report_types'] # check that all the report types are recognized for report_type in report_types: query_params = report_type_sql.get(report_type, {}) known_report_types = ('products', 'distinct_install', 'exploitability', 'devices', 'graphics') if (report_type not in known_report_types and 'first_col' not in query_params): raise BadArgumentError(report_type) products = [] versions = [] # Get information about the versions util_service = Util(config=self.context) versions_info = util_service.versions_info(**params) if versions_info: for i, elem in enumerate(versions_info): products.append(versions_info[elem]["product_name"]) versions.append(str(versions_info[elem]["version_string"])) # This MUST be a tuple otherwise it gets cast to an array params['product'] = tuple(products) params['version'] = tuple(versions) all_results = {} with self.get_connection() as connection: for report_type in report_types: result_cols, query_string, query_parameters = self._get_query( report_type, params) sql_results = self.query(query_string, params=query_parameters, connection=connection) results = [dict(zip(result_cols, row)) for row in sql_results] all_results[report_type] = results if individual_report: return all_results.values()[0] else: return {'reports': all_results}
def get_adu_by_signature(self, **kwargs): """Return a list of ADUs and crash counts by signature and ADU date """ now = datetimeutil.utc_now().date() lastweek = now - datetime.timedelta(weeks=1) filters = [ ("start_date", lastweek, "date"), ("end_date", now, "date"), ("signature", None, "str"), ("channel", None, "str"), ("product_name", None, "str"), ] params = external_common.parse_arguments(filters, kwargs) for param in ("start_date", "end_date", "signature", "channel"): if not params[param]: raise MissingArgumentError(param) if (params.end_date - params.start_date) > datetime.timedelta(days=365): raise BadArgumentError('Duration too long. Max 365 days.') sql_query = """ SELECT product_name, signature, adu_date::TEXT, build_date::TEXT, buildid::TEXT, crash_count, adu_count, os_name, channel FROM crash_adu_by_build_signature WHERE adu_date BETWEEN %(start_date)s AND %(end_date)s AND product_name = %(product_name)s AND channel = %(channel)s AND signature = %(signature)s ORDER BY buildid """ error_message = ( "Failed to retrieve crash ADU by build signature from PostgreSQL") results = self.query(sql_query, params, error_message=error_message) fields = [ 'product_name', 'signature', 'adu_date', 'build_date', 'buildid', 'crash_count', 'adu_count', 'os_name', 'channel' ] crashes = [dict(zip(fields, row)) for row in results] return {"hits": crashes, "total": len(crashes)}
def get(self, **kwargs): '''Return the result of a custom query. ''' params = external_common.parse_arguments(self.filters, kwargs) if not params.query: raise MissingArgumentError('query') try: query = json.loads(params.query) except ValueError: raise BadArgumentError( 'query', msg="Invalid JSON value for parameter 'query'" ) es = pyelasticsearch.ElasticSearch( urls=self.config.elasticsearch_urls, timeout=self.config.elasticsearch_timeout_extended, ) # Set indices. indices = [] if not params.indices: # By default, use the last two indices. today = utc_now() last_week = today - datetime.timedelta(days=7) indices = self.generate_list_of_indexes(last_week, today) elif len(params.indices) == 1 and params.indices[0] == 'ALL': # If we want all indices, just do nothing. pass else: indices = params.indices search_args = {} if indices: search_args['index'] = indices search_args['doc_type'] = self.config.elasticsearch_doctype try: results = es.search( query, **search_args ) except ElasticHttpNotFoundError, e: missing_index = re.findall(BAD_INDEX_REGEX, e.error)[0] raise ResourceNotFound( "elasticsearch index '%s' does not exist" % missing_index )
def post(self, **kwargs): """Return a list of signatures-to-bug_ids or bug_ids-to-signatures associations. """ params = external_common.parse_arguments(self.filters, kwargs) if not params['signatures'] and not params['bug_ids']: raise MissingArgumentError('specify one of signatures or bug_ids') elif params['signatures'] and params['bug_ids']: raise BadArgumentError('specify only one of signatures or bug_ids') sql_params = [] if params['signatures']: sql_params.append(tuple(params.signatures)) sql = """/* socorro.external.postgresql.bugs.Bugs.get */ SELECT ba.signature, bugs.id FROM bugs JOIN bug_associations AS ba ON bugs.id = ba.bug_id WHERE EXISTS( SELECT 1 FROM bug_associations WHERE bug_associations.bug_id = bugs.id AND signature IN %s ) """ elif params['bug_ids']: sql_params.append(tuple(params.bug_ids)) sql = """/* socorro.external.postgresql.bugs.Bugs.get */ SELECT ba.signature, bugs.id FROM bugs JOIN bug_associations AS ba ON bugs.id = ba.bug_id WHERE bugs.id IN %s """ error_message = "Failed to retrieve bug associations from PostgreSQL" results = self.query(sql, sql_params, error_message=error_message) bugs = [] for row in results: bug = dict(zip(("signature", "id"), row)) bugs.append(bug) return {"hits": bugs, "total": len(bugs)}
def get(self, **kwargs): filters = [ ("backfill_type", None, "str"), ("reports_clean", True, "bool"), ("check_period", '01:00:00', "str"), ("table_name", None, "str"), ("update_day", None, "datetime"), ("start_date", None, "datetime"), ("end_date", None, "datetime"), ] params = external_common.parse_arguments(filters, kwargs) if not params.backfill_type: raise MissingArgumentError('backfill_type') date_param = ['update_day', 'start_date', 'end_date'] for i in date_param: if i in kwargs: params[i] = str(params[i].date()) try: query = 'SELECT backfill_%(backfill_type)s (%(params)s); ' required_params = BACKFILL_PARAMETERS[params.backfill_type] query_params = [(i, params[i]) for i in required_params] query_params_str = ', '.join('%(' + str(i[0]) + ')s' for i in query_params) query = query % { 'backfill_type': params.backfill_type, 'params': query_params_str } except: raise BadArgumentError(kwargs['backfill_type']) error_message = "Failed to retrieve backfill %s from PostgreSQL" error_message = error_message % kwargs['backfill_type'] results = self.query(query, params, error_message=error_message) return results
def get_list(self, **kwargs): """ List all crashes with a given signature and return them. Both `from_date` and `to_date` (and their aliases `from` and `to`) are required and can not be greater than 30 days apart. Optional arguments: see SearchCommon.get_parameters() """ # aliases if "from" in kwargs and "from_date" not in kwargs: kwargs["from_date"] = kwargs.get("from") if "to" in kwargs and "to_date" not in kwargs: kwargs["to_date"] = kwargs.get("to") if not kwargs.get('from_date'): raise MissingArgumentError('from_date') if not kwargs.get('to_date'): raise MissingArgumentError('to_date') from_date = datetimeutil.datetimeFromISOdateString(kwargs['from_date']) to_date = datetimeutil.datetimeFromISOdateString(kwargs['to_date']) span_days = (to_date - from_date).days if span_days > 30: raise BadArgumentError( 'Span between from_date and to_date can not be more than 30') # start with the default sort_order = {'key': 'date_processed', 'direction': 'DESC'} if 'sort' in kwargs: sort_order['key'] = kwargs.pop('sort') _recognized_sort_orders = ( 'date_processed', 'uptime', 'user_comments', 'uuid', 'uuid_text', 'product', 'version', 'build', 'signature', 'url', 'os_name', 'os_version', 'cpu_name', 'cpu_info', 'address', 'reason', 'last_crash', 'install_age', 'hangid', 'process_type', 'release_channel', 'install_time', 'duplicate_of', ) if sort_order['key'] not in _recognized_sort_orders: raise BadArgumentError( '%s is not a recognized sort order key' % sort_order['key']) sort_order['direction'] = 'ASC' if 'reverse' in kwargs: if kwargs.pop('reverse'): sort_order['direction'] = 'DESC' include_raw_crash = kwargs.get('include_raw_crash') or False params = search_common.get_parameters(kwargs) if not params["signature"]: raise MissingArgumentError('signature') params["terms"] = params["signature"] params["search_mode"] = "is_exactly" # Default mode falls back to starts_with for postgres if params["plugin_search_mode"] == "default": params["plugin_search_mode"] = "starts_with" # Limiting to a signature if params["terms"]: params["terms"] = self.prepare_terms(params["terms"], params["search_mode"]) # Searching for terms in plugins if params["report_process"] == "plugin" and params["plugin_terms"]: params["plugin_terms"] = " ".join(params["plugin_terms"]) params["plugin_terms"] = self.prepare_terms( params["plugin_terms"], params["plugin_search_mode"]) # Get information about the versions util_service = Util(config=self.context) params["versions_info"] = util_service.versions_info(**params) # Parsing the versions params["versions_string"] = params["versions"] (params["versions"], params["products"]) = self.parse_versions(params["versions"], params["products"]) if hasattr(self.context, 'webapi'): context = self.context.webapi else: # old middleware context = self.context # Changing the OS ids to OS names for i, elem in enumerate(params["os"]): for platform in context.platforms: if platform["id"][:3] == elem[:3]: params["os"][i] = platform["name"] # Creating the parameters for the sql query sql_params = {} # Preparing the different parts of the sql query sql_select = """ SELECT r.date_processed, r.uptime, r.user_comments, r.uuid::uuid, r.uuid as uuid_text, r.product, r.version, r.build, r.signature, r.url, r.os_name, r.os_version, r.cpu_name, r.cpu_info, r.address, r.reason, r.last_crash, r.install_age, r.hangid, r.process_type, r.release_channel, (r.client_crash_date - (r.install_age * INTERVAL '1 second')) AS install_time """ if include_raw_crash: pass else: sql_select += """ , rd.duplicate_of """ wrapped_select = """ WITH report_slice AS ( %s ), dupes AS ( SELECT report_slice.uuid, rd.duplicate_of FROM reports_duplicates rd JOIN report_slice ON report_slice.uuid_text = rd.uuid WHERE rd.date_processed BETWEEN %%(from_date)s AND %%(to_date)s ) SELECT rs.*, dupes.duplicate_of, rc.raw_crash FROM report_slice rs LEFT OUTER JOIN dupes USING (uuid) LEFT OUTER JOIN raw_crashes rc ON rs.uuid = rc.uuid AND rc.date_processed BETWEEN %%(from_date)s AND %%(to_date)s """ sql_from = self.build_reports_sql_from(params) if not include_raw_crash: sql_from = """%s LEFT OUTER JOIN reports_duplicates rd ON r.uuid = rd.uuid """ % sql_from sql_where, sql_params = self.build_reports_sql_where( params, sql_params, self.context) sql_order = """ ORDER BY %(key)s %(direction)s """ % sort_order sql_limit, sql_params = self.build_reports_sql_limit( params, sql_params) # Assembling the query if include_raw_crash: sql_query = "\n".join( ("/* socorro.external.postgresql.report.Report.list */", sql_select, sql_from, sql_where, sql_order, sql_limit)) else: sql_query = "\n".join( ("/* socorro.external.postgresql.report.Report.list */", sql_select, sql_from, sql_where, sql_order, sql_limit)) # Query for counting the results sql_count_query = "\n".join( ("/* socorro.external.postgresql.report.Report.list */", "SELECT count(*)", sql_from, sql_where)) # Querying the DB with self.get_connection() as connection: total = self.count( sql_count_query, sql_params, error_message="Failed to count crashes from reports.", connection=connection) # No need to call Postgres if we know there will be no results if total: if include_raw_crash: sql_query = wrapped_select % sql_query results = self.query( sql_query, sql_params, error_message="Failed to retrieve crashes from reports", connection=connection) else: results = [] # Transforming the results into what we want fields = ( "date_processed", "uptime", "user_comments", "uuid", "uuid", # the uuid::text one "product", "version", "build", "signature", "url", "os_name", "os_version", "cpu_name", "cpu_info", "address", "reason", "last_crash", "install_age", "hangid", "process_type", "release_channel", "install_time", "duplicate_of", ) if include_raw_crash: fields += ("raw_crash", ) crashes = [] for row in results: crash = dict(zip(fields, row)) if include_raw_crash and crash['raw_crash']: crash['raw_crash'] = json.loads(crash['raw_crash']) for i in crash: try: crash[i] = datetimeutil.date_to_string(crash[i]) except TypeError: pass crashes.append(crash) return {"hits": crashes, "total": total}
def get(self, **kwargs): filters = [ ("report_type", None, "str"), ("signature", None, "str"), ("start_date", None, "datetime"), ("end_date", None, "datetime"), ("versions", None, ["list", "str"]), ] params = external_common.parse_arguments(filters, kwargs) products = [] versions = [] # Get information about the versions util_service = Util(config=self.context) versions_info = util_service.versions_info(**params) if versions_info: for i, elem in enumerate(versions_info): products.append(versions_info[elem]["product_name"]) versions.append(str(versions_info[elem]["version_string"])) # This MUST be a tuple otherwise it gets cast to an array params['product'] = tuple(products) params['version'] = tuple(versions) if params['product'] and params['report_type'] is not 'products': product_list = ' AND product_name IN %s ' else: product_list = '' if params['version'] and params['report_type'] is not 'products': version_list = ' AND version_string IN %s ' else: version_list = '' query_params = report_type_sql.get(params['report_type'], {}) if (params['report_type'] not in ('products', 'distinct_install', 'exploitability', 'devices', 'graphics') and 'first_col' not in query_params): raise BadArgumentError('report type') self.connection = self.database.connection() cursor = self.connection.cursor() if params['report_type'] == 'products': result_cols = ['product_name', 'version_string', 'report_count', 'percentage'] query_string = """ WITH crashes as ( SELECT product_name as category , version_string , SUM(report_count) as report_count FROM signature_summary_products JOIN signatures USING (signature_id) WHERE signatures.signature = %s AND report_date >= %s AND report_date < %s GROUP BY product_name, version_string ), totals as ( SELECT category , version_string , report_count , SUM(report_count) OVER () as total_count FROM crashes ) SELECT category , version_string , report_count , round((report_count * 100::numeric)/total_count,3)::TEXT as percentage FROM totals ORDER BY report_count DESC""" query_parameters = (params['signature'], params['start_date'], params['end_date']) elif params['report_type'] == 'distinct_install': result_cols = ['product_name', 'version_string', 'crashes', 'installations'] query_string = """ SELECT product_name , version_string , SUM(crash_count) AS crashes , SUM(install_count) AS installations FROM signature_summary_installations JOIN signatures USING (signature_id) WHERE signatures.signature = %s AND report_date >= %s AND report_date < %s """ query_string += product_list query_string += version_list query_string += """ GROUP BY product_name, version_string ORDER BY crashes DESC """ query_parameters = ( params['signature'], params['start_date'], params['end_date'] ) if product_list: query_parameters += (params['product'],) if version_list: query_parameters += (params['version'],) elif params['report_type'] == 'exploitability': # Note, even if params['product'] is something we can't use # that in this query result_cols = [ 'report_date', 'null_count', 'none_count', 'low_count', 'medium_count', 'high_count', ] query_string = """ SELECT cast(report_date as TEXT), SUM(null_count), SUM(none_count), SUM(low_count), SUM(medium_count), SUM(high_count) FROM exploitability_reports JOIN signatures USING (signature_id) WHERE signatures.signature = %s AND report_date >= %s AND report_date < %s """ query_string += product_list query_string += version_list query_string += """ GROUP BY report_date ORDER BY report_date DESC """ query_parameters = ( params['signature'], params['start_date'], params['end_date'], ) if product_list: query_parameters += (params['product'],) if version_list: query_parameters += (params['version'],) elif params['report_type'] == 'devices': result_cols = [ 'cpu_abi', 'manufacturer', 'model', 'version', 'report_count', 'percentage', ] query_string = """ WITH crashes as ( SELECT android_devices.android_cpu_abi as cpu_abi, android_devices.android_manufacturer as manufacturer, android_devices.android_model as model, android_devices.android_version as version, SUM(report_count) as report_count FROM signature_summary_device JOIN signatures USING (signature_id) JOIN android_devices ON signature_summary_device.android_device_id = android_devices.android_device_id WHERE signatures.signature = %s AND report_date >= %s AND report_date < %s """ query_string += product_list query_string += version_list query_string += """ GROUP BY android_devices.android_cpu_abi, android_devices.android_manufacturer, android_devices.android_model, android_devices.android_version ), totals as ( SELECT cpu_abi, manufacturer, model, version, report_count, SUM(report_count) OVER () as total_count FROM crashes ) SELECT cpu_abi, manufacturer, model, version, report_count, round((report_count * 100::numeric)/total_count,3)::TEXT as percentage FROM totals ORDER BY report_count DESC """ query_parameters = ( params['signature'], params['start_date'], params['end_date'], ) if product_list: query_parameters += (params['product'],) if version_list: query_parameters += (params['version'],) elif params['report_type'] == 'graphics': result_cols = [ 'vendor_hex', 'adapter_hex', 'vendor_name', 'adapter_name', 'report_count', 'percentage', ] query_string = """ WITH crashes as ( SELECT graphics_device.vendor_hex as vendor_hex, graphics_device.adapter_hex as adapter_hex, graphics_device.vendor_name as vendor_name, graphics_device.adapter_name as adapter_name, SUM(report_count) as report_count FROM signature_summary_graphics JOIN signatures USING (signature_id) JOIN graphics_device ON signature_summary_graphics.graphics_device_id = graphics_device.graphics_device_id WHERE signatures.signature = %s AND report_date >= %s AND report_date < %s """ query_string += product_list query_string += version_list query_string += """ GROUP BY graphics_device.graphics_device_id ), totals as ( SELECT vendor_hex, adapter_hex, vendor_name, adapter_name, report_count, SUM(report_count) OVER () as total_count FROM crashes ) SELECT vendor_hex, adapter_hex, vendor_name, adapter_name, report_count, round((report_count * 100::numeric)/total_count,3)::TEXT as percentage FROM totals ORDER BY report_count DESC """ query_parameters = ( params['signature'], params['start_date'], params['end_date'], ) if product_list: query_parameters += (params['product'],) if version_list: query_parameters += (params['version'],) elif params['report_type'] in report_type_columns: result_cols = ['category', 'report_count', 'percentage'] query_string = """ WITH crashes AS ( SELECT """ query_string += report_type_columns[params['report_type']] query_string += """ AS category , sum(report_count) AS report_count FROM signature_summary_""" query_string += params['report_type'] query_string += """ JOIN signatures USING (signature_id) WHERE signatures.signature = %s AND report_date >= %s AND report_date < %s """ query_string += product_list query_string += version_list query_string += """ GROUP BY category ), totals AS ( SELECT category , report_count , sum(report_count) OVER () as total_count FROM crashes ) SELECT category , report_count , round((report_count * 100::numeric)/total_count,3)::TEXT as percentage FROM totals ORDER BY report_count DESC """ query_parameters = ( params['signature'], params['start_date'], params['end_date'] ) if product_list: query_parameters += (params['product'],) if version_list: query_parameters += (params['version'],) sql_results = db.execute(cursor, query_string, query_parameters) results = [] for row in sql_results: newrow = dict(zip(result_cols, row)) results.append(newrow) # Closing the connection here because we're not using # the parent class' query() self.connection.close() return results
def post(self, **kwargs): try: data = json.loads(kwargs['data']) if data is None: raise BadArgumentError('POST data sent was null') except AttributeError: raise MissingArgumentError('No POST data sent') except ValueError: raise BadArgumentError('Posted data not valid JSON') except TypeError: # happens if kwargs['data'] is None raise BadArgumentError('POST data sent was empty') # make an upsert for each thing and rollback if any failed upsert = """ WITH update_graphics_device AS ( UPDATE graphics_device SET adapter_name = %(adapter_name)s, vendor_name = %(vendor_name)s WHERE vendor_hex = %(vendor_hex)s AND adapter_hex = %(adapter_hex)s RETURNING 1 ), insert_graphics_device AS ( INSERT INTO graphics_device (vendor_hex, adapter_hex, vendor_name, adapter_name) SELECT %(vendor_hex)s AS vendor_hex, %(adapter_hex)s AS adapter_hex, %(vendor_name)s AS vendor_name, %(adapter_name)s AS adapter_name WHERE NOT EXISTS ( SELECT * FROM graphics_device WHERE vendor_hex = %(vendor_hex)s AND adapter_hex = %(adapter_hex)s LIMIT 1 ) RETURNING 2 ) SELECT * FROM update_graphics_device UNION ALL SELECT * FROM insert_graphics_device """ with self.get_connection() as connection: try: for row in data: self.query(upsert, row, connection=connection) connection.commit() return True except (psycopg2.Error, KeyError): # KeyErrors happen if any of the rows don't have # all the required keys connection.rollback() return False
def get(self, **kwargs): """Return a list of results and facets based on parameters. The list of accepted parameters (with types and default values) is in socorro.lib.search_common.SearchBase """ # Filter parameters and raise potential errors. params = self.get_parameters(**kwargs) # Find the indexes to use to optimize the elasticsearch query. indexes = self.get_indexes(params['date']) # Create and configure the search object. search = SuperS().es( urls=self.config.elasticsearch_urls, timeout=self.config.elasticsearch_timeout, ) search = search.indexes(*indexes) search = search.doctypes(self.config.elasticsearch_doctype) # Create filters. filters = F() for field, sub_params in params.items(): sub_filters = F() for param in sub_params: if param.name.startswith('_'): if param.name == '_results_offset': results_from = param.value[0] elif param.name == '_results_number': results_number = param.value[0] # Don't use meta parameters in the query. continue field_data = self.all_fields[param.name] name = '%s.%s' % (field_data['namespace'], field_data['in_database_name']) if param.data_type in ('date', 'datetime'): param.value = datetimeutil.date_to_string(param.value) elif param.data_type == 'enum': param.value = [x.lower() for x in param.value] elif param.data_type == 'str' and not param.operator: param.value = [x.lower() for x in param.value] args = {} if not param.operator: # contains one of the terms if len(param.value) == 1: val = param.value[0] if not isinstance(val, basestring) or (isinstance( val, basestring) and ' ' not in val): args[name] = val # If the term contains white spaces, we want to perform # a phrase query. Thus we do nothing here and let this # value be handled later. else: args['%s__in' % name] = param.value elif param.operator == '=': # is exactly if field_data['has_full_version']: name = '%s.full' % name args[name] = param.value elif param.operator == '>': # greater than args['%s__gt' % name] = param.value elif param.operator == '<': # lower than args['%s__lt' % name] = param.value elif param.operator == '>=': # greater than or equal to args['%s__gte' % name] = param.value elif param.operator == '<=': # lower than or equal to args['%s__lte' % name] = param.value elif param.operator == '__null__': # is null args['%s__missing' % name] = param.value if args: if param.operator_not: new_filter = ~F(**args) else: new_filter = F(**args) if param.data_type == 'enum': sub_filters |= new_filter else: sub_filters &= new_filter continue # These use a wildcard and thus need to be in a query # instead of a filter. operator_wildcards = { '~': '*%s*', # contains '$': '%s*', # starts with '^': '*%s' # ends with } if param.operator in operator_wildcards: if field_data['has_full_version']: name = '%s.full' % name args['%s__wildcard' % name] = \ operator_wildcards[param.operator] % param.value args['must_not'] = param.operator_not elif not param.operator: # This is phrase that was passed down. args['%s__match_phrase' % name] = param.value[0] if args: search = search.query(**args) else: # If we reach this point, that means the operator is # not supported, and we should raise an error about that. raise NotImplementedError('Operator %s is not supported' % param.operator) filters &= sub_filters search = search.filter(filters) # Pagination. results_to = results_from + results_number search = search[results_from:results_to] # Create facets. processed_filters = search._process_filters(filters.filters) for param in params['_facets']: for value in param.value: try: field_ = self.all_fields[value] except KeyError: # That is not a known field, we can't facet on it. raise BadArgumentError( value, msg='Unknown field "%s", cannot facet on it' % value) field_name = '%s.%s' % (field_['namespace'], field_['in_database_name']) if field_['has_full_version']: # If the param has a full version, that means what matters # is the full string, and not its individual terms. field_name += '.full' args = { value: { 'terms': { 'field': field_name, 'size': self.config.facets_max_number, }, 'facet_filter': processed_filters, } } search = search.facet_raw(**args) # Query and compute results. hits = [] fields = [ '%s.%s' % (x['namespace'], x['in_database_name']) for x in self.all_fields.values() if x['is_returned'] ] if params['_return_query'][0].value[0]: # Return only the JSON query that would be sent to elasticsearch. return { 'query': search._build_query(), 'indices': indexes, } # We call elasticsearch with a computed list of indices, based on # the date range. However, if that list contains indices that do not # exist in elasticsearch, an error will be raised. We thus want to # remove all failing indices until we either have a valid list, or # an empty list in which case we return no result. while True: try: for hit in search.values_dict(*fields): hits.append(self.format_field_names(hit)) total = search.count() facets = search.facet_counts() break # Yay! Results! except ElasticHttpNotFoundError, e: missing_index = re.findall(BAD_INDEX_REGEX, e.error)[0] if missing_index in indexes: del indexes[indexes.index(missing_index)] else: # Wait what? An error caused by an index that was not # in the request? That should never happen, but in case # it does, better know it. raise if indexes: # Update the list of indices and try again. search = search.indexes(*indexes) else: # There is no index left in the list, return an empty # result. hits = [] total = 0 facets = {} break
def get_exploitability(self, **kwargs): """Return a list of exploitable crash reports. See socorro.lib.external_common.parse_arguments() for all filters. """ now = datetimeutil.utc_now().date() lastweek = now - datetime.timedelta(weeks=1) filters = [ ("start_date", lastweek, "date"), ("end_date", now, "date"), ("product", None, "str"), ("version", None, "str"), ("page", None, "int"), ("batch", None, "int"), ] params = external_common.parse_arguments(filters, kwargs) sql_where = """ report_date BETWEEN %(start_date)s AND %(end_date)s AND null_count + none_count + low_count + medium_count + high_count > 4 """ if params.product: sql_where += " AND pv.product_name = %(product)s" if params.version: sql_where += " AND pv.version_string = %(version)s" inner_with_sql = """ SELECT signature, SUM(high_count) AS high_count, SUM(medium_count) AS medium_count, SUM(low_count) AS low_count, SUM(null_count) AS null_count, SUM(none_count) AS none_count, SUM(high_count) + SUM(medium_count) AS med_or_high FROM exploitability_reports JOIN product_versions AS pv USING (product_version_id) WHERE high_count + medium_count + null_count + none_count > 4 AND %s GROUP BY signature """ % (sql_where, ) count_sql_query = """ /* external.postgresql.crashes.Crashes.get_exploitability */ WITH sums AS ( %s ) SELECT count(signature) FROM sums """ % (inner_with_sql, ) results = self.query( count_sql_query, params, error_message="Failed to retrieve exploitable crashes count") total_crashes_count, = results[0] sql_query = """ /* external.postgresql.crashes.Crashes.get_exploitability */ WITH sums AS ( %s ) SELECT signature, high_count, medium_count, low_count, null_count, none_count FROM sums ORDER BY med_or_high DESC, signature ASC """ % (inner_with_sql, ) if params['page'] is not None: if params['page'] <= 0: raise BadArgumentError('page', params['page'], 'starts on 1') if params['batch'] is None: raise MissingArgumentError('batch') sql_query += """ LIMIT %(limit)s OFFSET %(offset)s """ params['limit'] = params['batch'] params['offset'] = params['batch'] * (params['page'] - 1) error_message = ( "Failed to retrieve exploitable crashes from PostgreSQL") results = self.query(sql_query, params, error_message=error_message) # Transforming the results into what we want crashes = [] for row in results: crash = dict( zip(("signature", "high_count", "medium_count", "low_count", "null_count", "none_count"), row)) crashes.append(crash) return {"hits": crashes, "total": total_crashes_count}
def get(self, **kwargs): """ Return product information, or version information for one or more product:version combinations """ filters = [ ("versions", None, ["list", "str"]), # for legacy, to be removed ("type", "desktop", "str"), ] params = external_common.parse_arguments(filters, kwargs) accepted_types = ("desktop", "webapp") if params.type not in accepted_types: raise BadArgumentError('type', params.type, accepted_types) if params.versions and params.versions[0]: return self._get_versions(params) if params.type == "desktop": sql = """ /* socorro.external.postgresql.products.Products.get */ SELECT product_name, version_string, start_date, end_date, throttle, is_featured, build_type, has_builds FROM product_info ORDER BY product_sort, version_sort DESC, channel_sort """ elif params.type == "webapp": sql = """ /* socorro.external.postgresql.products.Products.get */ SELECT product_name, version, NULL as start_date, NULL as end_date, 1.0 as throttle, FALSE as is_featured, build_type, FALSE as has_builds FROM bixie.raw_product_releases ORDER BY product_name, version DESC """ error_message = "Failed to retrieve products/versions from PostgreSQL" results = self.query(sql, error_message=error_message) products = [] versions_per_product = {} for row in results: version = dict( zip(( 'product', 'version', 'start_date', 'end_date', 'throttle', 'featured', 'release', 'has_builds', ), row)) try: version['end_date'] = datetimeutil.date_to_string( version['end_date']) except TypeError: pass try: version['start_date'] = datetimeutil.date_to_string( version['start_date']) except TypeError: pass version['throttle'] = float(version['throttle']) product = version['product'] if product not in products: products.append(product) if product not in versions_per_product: versions_per_product[product] = [version] else: versions_per_product[product].append(version) return { 'products': products, 'hits': versions_per_product, 'total': len(results) }
def get_parameters(self, **kwargs): parameters = {} fields = kwargs['_fields'] assert fields if fields: self.build_filters(fields) for param in self.filters: values = kwargs.get(param.name, param.default) if values in ('', []): # Those values are equivalent to None here. # Note that we cannot use bool(), because 0 is not equivalent # to None in our case. values = None if values is None and param.mandatory: raise MissingArgumentError(param.name) if values is None and param.default is not None: values = param.default # all values can be a list, so we make them all lists to simplify if values is not None and not isinstance(values, (list, tuple)): values = [values] if values is not None: # There should only be one parameter with no operator, and # we want to stack all values into it. That's why we want # to keep track of it. # Actually, we want _two_ parameters with no operator: one # for each possible value of "operator_not". no_operator_param = {True: None, False: None} for value in values: operator = None operator_not = False operators = OPERATORS_MAP.get(param.data_type, OPERATORS_MAP['default']) if isinstance(value, basestring): if value.startswith(OPERATOR_NOT): operator_not = True value = value[1:] for ope in operators: if value.startswith(ope): operator = ope value = value[len(ope):] break # ensure the right data type try: value = convert_to_type(value, param.data_type) except ValueError: raise BadArgumentError( param.name, msg='Bad value for parameter %s:' ' "%s" is not a valid %s' % (param.name, value, param.data_type)) if param.name not in parameters: parameters[param.name] = [] if not operator: if not no_operator_param[operator_not]: no_operator_param[operator_not] = SearchParam( param.name, [value], operator, param.data_type, operator_not) else: no_operator_param[operator_not].value.append(value) else: parameters[param.name].append( SearchParam(param.name, value, operator, param.data_type, operator_not)) for value in no_operator_param.values(): if value: parameters[value.name].append(value) self.fix_date_parameter(parameters) self.fix_process_type_parameter(parameters) self.fix_hang_type_parameter(parameters) self.fix_version_parameter(parameters) return parameters
def get_exploitability(self, **kwargs): """Return a list of exploitable crash reports. See socorro.lib.external_common.parse_arguments() for all filters. """ now = datetimeutil.utc_now().date() lastweek = now - datetime.timedelta(weeks=1) filters = [ ("start_date", lastweek, "date"), ("end_date", now, "date"), ("page", None, "int"), ("batch", None, "int"), ] params = external_common.parse_arguments(filters, kwargs) count_sql_query = """ /* external.postgresql.crashes.Crashes.get_exploitability */ SELECT COUNT(*) FROM exploitability_reports WHERE report_date BETWEEN %(start_date)s AND %(end_date)s """ results = self.query( count_sql_query, params, error_message="Failed to retrieve exploitable crashes count") total_crashes_count, = results[0] sql_query = """ /* external.postgresql.crashes.Crashes.get_exploitability */ SELECT signature, report_date, null_count, none_count, low_count, medium_count, high_count FROM exploitability_reports WHERE report_date BETWEEN %(start_date)s AND %(end_date)s ORDER BY report_date DESC """ if params['page'] is not None: if params['page'] <= 0: raise BadArgumentError('page', params['page'], 'starts on 1') if params['batch'] is None: raise MissingArgumentError('batch') sql_query += """ LIMIT %(limit)s OFFSET %(offset)s """ params['limit'] = params['batch'] params['offset'] = params['batch'] * (params['page'] - 1) error_message = "Failed to retrieve exploitable crashes from PostgreSQL" results = self.query(sql_query, params, error_message=error_message) # Transforming the results into what we want crashes = [] for row in results: crash = dict( zip(("signature", "report_date", "null_count", "none_count", "low_count", "medium_count", "high_count"), row)) crash["report_date"] = datetimeutil.date_to_string( crash["report_date"]) crashes.append(crash) return {"hits": crashes, "total": total_crashes_count}
def get(self, **kwargs): """ Return urls for signature """ filters = [ ("signature", None, "str"), ("start_date", None, "datetime"), ("end_date", None, "datetime"), ("products", None, ["list", "str"]), ("versions", None, ["list", "str"]), ] params = external_common.parse_arguments(filters, kwargs) #Because no parameters are optional, we need to loop through #all parameters to ensure each has been set and is not None missingParams = [] for param in params: if not params[param]: if param == 'versions': # force versions parameter to being 'ALL' if empty params[param] = 'ALL' continue missingParams.append(param) if len(missingParams) > 0: raise MissingArgumentError(", ".join(missingParams)) all_products_versions_sql = """ /* socorro.external.postgresql.signature_urls.SignatureURLs.get */ SELECT url, count(*) as crash_count FROM reports_clean JOIN reports_user_info USING ( UUID ) JOIN signatures USING ( signature_id ) WHERE reports_clean.date_processed BETWEEN %(start_date)s AND %(end_date)s AND reports_user_info.date_processed BETWEEN %(start_date)s AND %(end_date)s AND signature = %(signature)s AND url <> '' """ sql = """ /* socorro.external.postgresql.signature_urls.SignatureURLs.get */ SELECT url, count(*) as crash_count FROM reports_clean JOIN reports_user_info USING ( UUID ) JOIN signatures USING ( signature_id ) JOIN product_versions USING ( product_version_id ) WHERE reports_clean.date_processed BETWEEN %(start_date)s AND %(end_date)s AND reports_user_info.date_processed BETWEEN %(start_date)s AND %(end_date)s AND signature = %(signature)s AND url <> '' AND ( """ sql_group_order = """ GROUP BY url ORDER BY crash_count DESC LIMIT 100""" sql_params = { "start_date": params.start_date, "end_date": params.end_date, "signature": params.signature } # if this query is for all products the 'ALL' keyword will be # the only item in the products list and this will then also # be for all versions. if 'ALL' in params['products']: sql_query = " ".join((all_products_versions_sql, sql_group_order)) # if this query is for all versions the 'ALL' keyword will be # the only item in the versions list. elif 'ALL' in params['versions']: sql_products = " product_name IN %(products)s )" sql_params['products'] = tuple(params.products) sql_date_range_limit = """AND %(end_date)s BETWEEN product_versions.build_date AND product_versions.sunset_date""" sql_query = " ".join((sql, sql_products, sql_date_range_limit, sql_group_order)) else: products = [] (params["products_versions"], products) = self.parse_versions(params["versions"], []) if len(params["products_versions"]) == 0: raise BadArgumentError(", ".join(params["versions"])) versions_list = [] products_list = [] for x in range(0, len(params["products_versions"]), 2): products_list.append(params["products_versions"][x]) versions_list.append(params["products_versions"][x + 1]) product_version_list = [] for prod in params["products"]: versions = [] [versions.append(versions_list[i]) for i, x in enumerate(products_list) if x == prod] product_version_list.append(tuple(versions)) sql_product_version_ids = [ """( product_name = %%(product%s)s AND version_string IN %%(version%s)s ) """ % (x, x) for x in range(len(product_version_list))] sql_params = add_param_to_dict(sql_params, "version", product_version_list) sql_params = add_param_to_dict(sql_params, "product", params.products) sql_query = " ".join((sql, " OR ".join(sql_product_version_ids), " ) " + sql_group_order)) error_message = "Failed to retrieve urls for signature from PostgreSQL" results = self.query(sql_query, sql_params, error_message=error_message) urls = results.zipped() return { "hits": urls, "total": len(urls) }
def get(self, **kwargs): """Return a list of results and facets based on parameters. The list of accepted parameters (with types and default values) is in socorro.lib.search_common.SearchBase """ # Filter parameters and raise potential errors. params = self.get_parameters(**kwargs) # Find the indexes to use to optimize the elasticsearch query. indexes = self.get_indexes(params['date']) # Create and configure the search object. search = SuperS().es( urls=self.config.elasticsearch_urls, timeout=self.config.elasticsearch_timeout, ) search = search.indexes(indexes) search = search.doctypes(self.config.elasticsearch_doctype) # Create filters. filters = F() for field, sub_params in params.items(): for param in sub_params: name = PARAM_TO_FIELD_MAPPING.get(param.name, param.name) name = self.prefix_field_name(name) if name.startswith('_'): if name == '_results_offset': results_from = param.value[0] elif name == '_results_number': results_number = param.value[0] # Don't use meta parameters in the query. continue if param.data_type in ('date', 'datetime'): param.value = datetimeutil.date_to_string(param.value) elif param.data_type == 'enum': param.value = [x.lower() for x in param.value] elif param.data_type == 'str' and not param.operator: param.value = [x.lower() for x in param.value] args = {} if not param.operator: # contains one of the terms if len(param.value) == 1: args[name] = param.value[0] else: args['%s__in' % name] = param.value elif param.operator == '=': # is exactly if name in FIELDS_WITH_FULL_VERSION: name = '%s.full' % name args[name] = param.value elif param.operator == '>': # greater than args['%s__gt' % name] = param.value elif param.operator == '<': # lower than args['%s__lt' % name] = param.value elif param.operator == '>=': # greater than or equal to args['%s__gte' % name] = param.value elif param.operator == '<=': # lower than or equal to args['%s__lte' % name] = param.value elif param.operator == '__null__': # is null args['%s__missing' % name] = param.value if args: if param.operator_not: filters &= ~F(**args) else: filters &= F(**args) continue # These use a wildcard and thus need to be in a query # instead of a filter. operator_wildcards = { '~': '*%s*', # contains '$': '%s*', # starts with '^': '*%s' # ends with } if param.operator in operator_wildcards: if name in FIELDS_WITH_FULL_VERSION: name = '%s.full' % name args['%s__wildcard' % name] = \ operator_wildcards[param.operator] % param.value args['must_not'] = param.operator_not if args: search = search.query(**args) else: # If we reach this point, that means the operator is # not supported, and we should raise an error about that. raise NotImplementedError('Operator %s is not supported' % param.operator) search = search.filter(filters) # Pagination. results_to = results_from + results_number search = search[results_from:results_to] # Create facets. processed_filters = search._process_filters(filters.filters) for param in params['_facets']: for value in param.value: filter_ = self.get_filter(value) if not filter_: # That is not a known field, we can't facet on it. raise BadArgumentError( 'Unknown field "%s", cannot facet on it' % value) field_name = PARAM_TO_FIELD_MAPPING.get(value, value) field_name = self.prefix_field_name(field_name) if field_name in FIELDS_WITH_FULL_VERSION: # If the param has a full version, that means what matters # is the full string, and not its individual terms. field_name += '.full' args = { value: { 'terms': { 'field': field_name, 'size': self.config.facets_max_number, }, 'facet_filter': processed_filters, } } search = search.facet_raw(**args) # Query and compute results. hits = [] fields = ['processed_crash.%s' % x for x in PROCESSED_CRASH_FIELDS] for hit in search.values_dict(*fields): hits.append(self.format_field_names(hit)) return { 'hits': hits, 'total': search.count(), 'facets': search.facet_counts(), }
def get(self, **kwargs): """Return a list of results and aggregations based on parameters. The list of accepted parameters (with types and default values) is in the database and can be accessed with the super_search_fields service. """ # Require that the list of fields be passed. if not kwargs.get('_fields'): raise MissingArgumentError('_fields') self.all_fields = kwargs['_fields'] self._build_fields() # Filter parameters and raise potential errors. params = self.get_parameters(**kwargs) # Find the indices to use to optimize the elasticsearch query. indices = self.get_indices(params['date']) # Create and configure the search object. search = Search( using=self.get_connection(), index=indices, doc_type=self.config.elasticsearch.elasticsearch_doctype, ) # Create filters. filters = [] histogram_intervals = {} for field, sub_params in params.items(): sub_filters = None for param in sub_params: if param.name.startswith('_'): # By default, all param values are turned into lists, # even when they have and can have only one value. # For those we know there can only be one value, # so we just extract it from the made-up list. if param.name == '_results_offset': results_from = param.value[0] elif param.name == '_results_number': results_number = param.value[0] if results_number > 1000: raise BadArgumentError('_results_number too large') elif param.name == '_facets_size': facets_size = param.value[0] for f in self.histogram_fields: if param.name == '_histogram_interval.%s' % f: histogram_intervals[f] = param.value[0] # Don't use meta parameters in the query. continue field_data = self.all_fields[param.name] name = '%s.%s' % (field_data['namespace'], field_data['in_database_name']) if param.data_type in ('date', 'datetime'): param.value = datetimeutil.date_to_string(param.value) elif param.data_type == 'enum': param.value = [x.lower() for x in param.value] elif param.data_type == 'str' and not param.operator: param.value = [x.lower() for x in param.value] # Operators needing wildcards, and the associated value # transformation with said wildcards. operator_wildcards = { '~': '*%s*', # contains '$': '%s*', # starts with '^': '*%s' # ends with } # Operators needing ranges, and the associated Elasticsearch # comparison operator. operator_range = { '>': 'gt', '<': 'lt', '>=': 'gte', '<=': 'lte', } args = {} filter_type = 'term' filter_value = None if not param.operator: # contains one of the terms if len(param.value) == 1: val = param.value[0] if not isinstance(val, basestring) or ' ' not in val: # There's only one term and no white space, this # is a simple term filter. filter_value = val else: # If the term contains white spaces, we want to # perform a phrase query. filter_type = 'query' args = Q( 'simple_query_string', query=param.value[0], fields=[name], default_operator='and', ).to_dict() else: # There are several terms, this is a terms filter. filter_type = 'terms' filter_value = param.value elif param.operator == '=': # is exactly if field_data['has_full_version']: name = '%s.full' % name filter_value = param.value elif param.operator in operator_range: filter_type = 'range' filter_value = { operator_range[param.operator]: param.value } elif param.operator == '__null__': filter_type = 'missing' args['field'] = name elif param.operator in operator_wildcards: filter_type = 'query' # Wildcard operations are better applied to a non-analyzed # field (called "full") if there is one. if field_data['has_full_version']: name = '%s.full' % name q_args = {} q_args[name] = (operator_wildcards[param.operator] % param.value) query = Q('wildcard', **q_args) args = query.to_dict() if filter_value is not None: args[name] = filter_value if args: new_filter = F(filter_type, **args) if param.operator_not: new_filter = ~new_filter if sub_filters is None: sub_filters = new_filter elif filter_type == 'range': sub_filters &= new_filter else: sub_filters |= new_filter continue if sub_filters is not None: filters.append(sub_filters) search = search.filter(F('bool', must=filters)) # Restricting returned fields. fields = [] for param in params['_columns']: for value in param.value: if not value: continue field_name = self.get_field_name(value, full=False) fields.append(field_name) search = search.fields(fields) # Sorting. sort_fields = [] for param in params['_sort']: for value in param.value: if not value: continue # Values starting with a '-' are sorted in descending order. # In order to retrieve the database name of the field, we # must first remove the '-' part and add it back later. # Example: given ['product', '-version'], the results will be # sorted by ascending product and descending version. desc = False if value.startswith('-'): desc = True value = value[1:] field_name = self.get_field_name(value, full=False) if desc: # The underlying library understands that '-' means # sorting in descending order. field_name = '-' + field_name sort_fields.append(field_name) search = search.sort(*sort_fields) # Pagination. results_to = results_from + results_number search = search[results_from:results_to] # Create facets. for param in params['_facets']: self._add_second_level_aggs( param, search.aggs, facets_size, histogram_intervals, ) # Create sub-aggregations. for key in params: if not key.startswith('_aggs.'): continue fields = key.split('.')[1:] if fields[0] not in self.all_fields: continue base_bucket = self._get_fields_agg(fields[0], facets_size) sub_bucket = base_bucket for field in fields[1:]: # For each field, make a bucket, then include that bucket in # the latest one, and then make that new bucket the latest. if field in self.all_fields: tmp_bucket = self._get_fields_agg(field, facets_size) sub_bucket.bucket(field, tmp_bucket) sub_bucket = tmp_bucket for value in params[key]: self._add_second_level_aggs( value, sub_bucket, facets_size, histogram_intervals, ) search.aggs.bucket(fields[0], base_bucket) # Create histograms. for f in self.histogram_fields: key = '_histogram.%s' % f if params.get(key): histogram_bucket = self._get_histogram_agg( f, histogram_intervals) for param in params[key]: self._add_second_level_aggs( param, histogram_bucket, facets_size, histogram_intervals, ) search.aggs.bucket('histogram_%s' % f, histogram_bucket) # Query and compute results. hits = [] if params['_return_query'][0].value[0]: # Return only the JSON query that would be sent to elasticsearch. return { 'query': search.to_dict(), 'indices': indices, } # We call elasticsearch with a computed list of indices, based on # the date range. However, if that list contains indices that do not # exist in elasticsearch, an error will be raised. We thus want to # remove all failing indices until we either have a valid list, or # an empty list in which case we return no result. while True: try: results = search.execute() for hit in results: hits.append(self.format_fields(hit.to_dict())) total = search.count() aggregations = self.format_aggregations(results.aggregations) break # Yay! Results! except NotFoundError, e: missing_index = re.findall(BAD_INDEX_REGEX, e.error)[0] if missing_index in indices: del indices[indices.index(missing_index)] else: # Wait what? An error caused by an index that was not # in the request? That should never happen, but in case # it does, better know it. raise if indices: # Update the list of indices and try again. # Note: we need to first empty the list of indices before # updating it, otherwise the removed indices never get # actually removed. search = search.index().index(*indices) else: # There is no index left in the list, return an empty # result. hits = [] total = 0 aggregations = {} break
def get(self, **kwargs): filters = [ ('limit', 500, 'int'), ('offset', 0, 'int'), ] params = external_common.parse_arguments(filters, kwargs) if params['limit'] > 1000: raise BadArgumentError('Max limit is 1000') sql = """ /* socorro.external.postgresql.laglog.LagLog.get */ SELECT replica_name, lag, moment, master, AVG(lag) OVER ( PARTITION BY replica_name ORDER BY moment DESC ROWS BETWEEN 0 FOLLOWING AND 11 FOLLOWING ) AS average FROM lag_log ORDER BY moment DESC LIMIT %(limit)s OFFSET %(offset)s """ results = self.query(sql, params) averages = defaultdict(list) all = defaultdict(list) for row in results: replica_name = row[0] lag = row[1] moment = row[2] master = row[3] average = row[4] timestamp = calendar.timegm(moment.utctimetuple()) all[replica_name].append({ 'x': timestamp, 'y': lag, 'master': master }) if not average: average = 0 averages[replica_name].append({'x': timestamp, 'y': int(average)}) max_bytes_critical = self.context.laglog.max_bytes_critical max_bytes_warning = self.context.laglog.max_bytes_warning replicas = [] for name, rows in all.items(): message = None last_average = averages[name][0]['y'] last_value = rows[0]['y'] if last_average > max_bytes_critical: message = 'CRITICAL' elif last_average > max_bytes_warning: message = 'WARNING' rows.reverse() averages_rows = averages[name] averages_rows.reverse() replicas.append({ 'name': name, 'rows': rows, 'averages': averages_rows, 'message': message, 'last_average': last_average, 'last_value': last_value, }) replicas.sort(key=lambda x: x['name']) return {'replicas': replicas}
def get_paireduuid(self, **kwargs): """Return paired uuid given a uuid and an optional hangid. If a hangid is passed, then return only one result. Otherwise, return all found paired uuids. """ filters = [ ("uuid", None, "str"), ("hangid", None, "str"), ] params = external_common.parse_arguments(filters, kwargs) if not params.uuid: raise MissingArgumentError('uuid') try: crash_date = datetimeutil.uuid_to_date(params.uuid) except ValueError: raise BadArgumentError( 'uuid', 'Date could not be converted extract from %s' % (params.uuid, )) sql = """ /* socorro.external.postgresql.crashes.Crashes.get_paireduuid */ SELECT uuid FROM reports r WHERE r.uuid != %(uuid)s AND r.date_processed BETWEEN TIMESTAMP %(crash_date)s - CAST('1 day' AS INTERVAL) AND TIMESTAMP %(crash_date)s + CAST('1 day' AS INTERVAL) """ sql_params = {"uuid": params.uuid, "crash_date": crash_date} if params.hangid is not None: sql = """%s AND r.hangid = %%(hangid)s LIMIT 1 """ % sql sql_params["hangid"] = params.hangid else: sql = """%s AND r.hangid IN ( SELECT hangid FROM reports r2 WHERE r2.date_processed BETWEEN TIMESTAMP %%(crash_date)s - CAST('1 day' AS INTERVAL) AND TIMESTAMP %%(crash_date)s + CAST('1 day' AS INTERVAL) AND r2.uuid = %%(uuid)s ) """ % sql # Query the database error_message = "Failed to retrieve paired uuids from PostgreSQL" results = self.query(sql, sql_params, error_message=error_message) # Transforming the results into what we want uuids = [dict(zip(("uuid", ), row)) for row in results] return {"hits": uuids, "total": len(uuids)}
def get(self, **kwargs): self.context.logger.info('Running %s' % self.__class__.__name__) raise BadArgumentError('bad arg')
def create_field(self, **kwargs): """Create a new field in the database, to be used by supersearch and all Elasticsearch related services. """ filters = [ ('name', None, 'str'), ('data_validation_type', 'enum', 'str'), ('default_value', None, 'str'), ('description', None, 'str'), ('form_field_choices', None, ['list', 'str']), ('has_full_version', False, 'bool'), ('in_database_name', None, 'str'), ('is_exposed', False, 'bool'), ('is_returned', False, 'bool'), ('is_mandatory', False, 'bool'), ('query_type', 'enum', 'str'), ('namespace', None, 'str'), ('permissions_needed', None, ['list', 'str']), ('storage_mapping', None, 'json'), ] params = external_common.parse_arguments(filters, kwargs) mandatory_params = ('name', 'in_database_name') for param in mandatory_params: if not params[param]: raise MissingArgumentError(param) # Before making the change, make sure it does not break indexing. new_mapping = self.get_mapping(overwrite_mapping=params) # Try the mapping. If there is an error, an exception will be raised. # If an exception is raised, the new mapping will be rejected. self.test_mapping(new_mapping) es_connection = self.get_connection() try: es_connection.index( index=self.config.elasticsearch.elasticsearch_default_index, doc_type='supersearch_fields', body=params, id=params['name'], op_type='create', refresh=True, ) except elasticsearch.exceptions.ConflictError: # This field exists in the database, it thus cannot be created! raise BadArgumentError( 'name', msg='The field "%s" already exists in the database, ' 'impossible to create it. ' % params['name'], ) if params.get('storage_mapping'): # If we made a change to the storage_mapping, log that change. self.config.logger.info( 'elasticsearch mapping changed for field "%s", ' 'added new mapping "%s"', params['name'], params['storage_mapping'], ) return True
def get(self, **kwargs): """Return a list of results and aggregations based on parameters. The list of accepted parameters (with types and default values) is in the database and can be accessed with the super_search_fields service. """ # Filter parameters and raise potential errors. params = self.get_parameters(**kwargs) # Find the indices to use to optimize the elasticsearch query. indices = self.get_indices(params['date']) # Create and configure the search object. search = Search( using=self.get_connection(), index=indices, doc_type=self.config.elasticsearch.elasticsearch_doctype, ) # Create filters. filters = None for field, sub_params in params.items(): sub_filters = None for param in sub_params: if param.name.startswith('_'): if param.name == '_results_offset': results_from = param.value[0] elif param.name == '_results_number': results_number = param.value[0] # Don't use meta parameters in the query. continue field_data = self.all_fields[param.name] name = '%s.%s' % (field_data['namespace'], field_data['in_database_name']) if param.data_type in ('date', 'datetime'): param.value = datetimeutil.date_to_string(param.value) elif param.data_type == 'enum': param.value = [x.lower() for x in param.value] elif param.data_type == 'str' and not param.operator: param.value = [x.lower() for x in param.value] args = {} filter_type = 'term' filter_value = None if not param.operator: # contains one of the terms if len(param.value) == 1: val = param.value[0] if not isinstance(val, basestring) or (isinstance( val, basestring) and ' ' not in val): filter_value = val # If the term contains white spaces, we want to perform # a phrase query. Thus we do nothing here and let this # value be handled later. else: filter_type = 'terms' filter_value = param.value elif param.operator == '=': # is exactly if field_data['has_full_version']: name = '%s.full' % name filter_value = param.value elif param.operator == '>': # greater than filter_type = 'range' filter_value = {'gt': param.value} elif param.operator == '<': # lower than filter_type = 'range' filter_value = {'lt': param.value} elif param.operator == '>=': # greater than or equal to filter_type = 'range' filter_value = {'gte': param.value} elif param.operator == '<=': # lower than or equal to filter_type = 'range' filter_value = {'lte': param.value} elif param.operator == '__null__': # is null filter_type = 'missing' args['field'] = name if filter_value is not None: args[name] = filter_value if args: if param.operator_not: new_filter = ~F(filter_type, **args) else: new_filter = F(filter_type, **args) if sub_filters is None: sub_filters = new_filter elif param.data_type == 'enum': sub_filters |= new_filter else: sub_filters &= new_filter continue # These use a wildcard and thus need to be in a query # instead of a filter. operator_wildcards = { '~': '*%s*', # contains '$': '%s*', # starts with '^': '*%s' # ends with } if param.operator in operator_wildcards: if field_data['has_full_version']: name = '%s.full' % name query_type = 'wildcard' args[name] = (operator_wildcards[param.operator] % param.value) elif not param.operator: # This is a phrase that was passed down. query_type = 'simple_query_string' args['query'] = param.value[0] args['fields'] = [name] args['default_operator'] = 'and' if args: query = Q(query_type, **args) if param.operator_not: query = ~query search = search.query(query) else: # If we reach this point, that means the operator is # not supported, and we should raise an error about that. raise NotImplementedError('Operator %s is not supported' % param.operator) if filters is None: filters = sub_filters elif sub_filters is not None: filters &= sub_filters search = search.filter(filters) # Pagination. results_to = results_from + results_number search = search[results_from:results_to] # Create facets. for param in params['_facets']: for value in param.value: try: field_ = self.all_fields[value] except KeyError: # That is not a known field, we can't facet on it. raise BadArgumentError( value, msg='Unknown field "%s", cannot facet on it' % value) field_name = '%s.%s' % (field_['namespace'], field_['in_database_name']) if field_['has_full_version']: # If the param has a full version, that means what matters # is the full string, and not its individual terms. field_name += '.full' search.aggs.bucket(value, 'terms', field=field_name, size=self.config.facets_max_number) # Query and compute results. hits = [] fields = [ '%s.%s' % (x['namespace'], x['in_database_name']) for x in self.all_fields.values() if x['is_returned'] ] search = search.fields(*fields) if params['_return_query'][0].value[0]: # Return only the JSON query that would be sent to elasticsearch. return { 'query': search.to_dict(), 'indices': indices, } # We call elasticsearch with a computed list of indices, based on # the date range. However, if that list contains indices that do not # exist in elasticsearch, an error will be raised. We thus want to # remove all failing indices until we either have a valid list, or # an empty list in which case we return no result. while True: try: results = search.execute() for hit in results: hits.append(self.format_fields(hit.to_dict())) total = search.count() aggregations = self.format_aggregations(results.aggregations) break # Yay! Results! except NotFoundError, e: missing_index = re.findall(BAD_INDEX_REGEX, e.error)[0] if missing_index in indices: del indices[indices.index(missing_index)] else: # Wait what? An error caused by an index that was not # in the request? That should never happen, but in case # it does, better know it. raise if indices: # Update the list of indices and try again. # Note: we need to first empty the list of indices before # updating it, otherwise the removed indices never get # actually removed. search = search.index().index(*indices) else: # There is no index left in the list, return an empty # result. hits = [] total = 0 aggregations = {} break
def get_parameters(self, **kwargs): parameters = {} for param in self.filters: values = kwargs.get(param.name, param.default) if values in ('', []): # Those values are equivalent to None here. # Note that we cannot use bool(), because 0 is not equivalent # to None in our case. values = None if values is None and param.mandatory: raise MissingArgumentError(param.name) if values is None and param.default is not None: values = param.default # all values can be a list, so we make them all lists to simplify if values is not None and not isinstance(values, (list, tuple)): values = [values] if values is not None: no_operator_param = None for value in values: operator = None operator_not = False operators = OPERATORS_MAP.get(param.data_type, OPERATORS_MAP['default']) if isinstance(value, basestring): if value.startswith(OPERATOR_NOT): operator_not = True value = value[1:] for ope in operators: if value.startswith(ope): operator = ope value = value[len(ope):] break # ensure the right data type try: value = convert_to_type(value, param.data_type) except ValueError: raise BadArgumentError( param.name, msg='Bad value for parameter %s:' ' "%s" is not a valid %s' % (param.name, value, param.data_type)) if not param.name in parameters: parameters[param.name] = [] if not operator: if not no_operator_param: no_operator_param = SearchParam( param.name, [value], operator, param.data_type, operator_not) else: no_operator_param.value.append(value) else: parameters[param.name].append( SearchParam(param.name, value, operator, param.data_type, operator_not)) if no_operator_param: parameters[no_operator_param.name].append( no_operator_param) self.fix_date_parameter(parameters) self.fix_process_type_parameter(parameters) self.fix_hang_type_parameter(parameters) return parameters
def getListOfTopCrashersBySignature(connection, dbParams): """ Answers a generator of tcbs rows """ assertPairs = { 'startDate': (datetime.date, datetime.datetime), 'to_date': (datetime.date, datetime.datetime), 'product': basestring, 'version': basestring, 'limit': int } for param in assertPairs: if not isinstance(dbParams[param], assertPairs[param]): raise BadArgumentError(type(dbParams[param])) order_by = 'report_count' # default order field where = [''] # trick for the later join if dbParams['crash_type'] != 'all': where.append("process_type = %s" % (sqlutils.quote_value(dbParams['crash_type']), )) if dbParams['os']: abbreviated_os = dbParams['os'][0:3].lower() if abbreviated_os not in ('win', 'lin', 'mac'): # this check prevents possible SQL injections raise BadArgumentError('Invalid OS to order on') order_by = '%s_count' % abbreviated_os where.append("%s > 0" % order_by) where = ' AND '.join(where) table_to_use = 'tcbs' date_range_field = 'report_date' if dbParams['date_range_type'] == 'build': table_to_use = 'tcbs_build' date_range_field = 'build_date' sql = """ WITH tcbs_r as ( SELECT tcbs.signature_id, signature, pv.product_name, version_string, sum(report_count) as report_count, sum(win_count) as win_count, sum(lin_count) as lin_count, sum(mac_count) as mac_count, sum(hang_count) as hang_count, plugin_count(process_type,report_count) as plugin_count, content_count(process_type,report_count) as content_count, first_report, version_list, sum(startup_count) as startup_count, sum(is_gc_count) as is_gc_count FROM %s tcbs JOIN signatures USING (signature_id) JOIN product_versions AS pv USING (product_version_id) JOIN signature_products_rollup AS spr ON spr.signature_id = tcbs.signature_id AND spr.product_name = pv.product_name WHERE pv.product_name = %%s AND version_string = %%s AND tcbs.%s BETWEEN %%s AND %%s %s GROUP BY tcbs.signature_id, signature, pv.product_name, version_string, first_report, spr.version_list ), tcbs_window AS ( SELECT tcbs_r.*, sum(report_count) over () as total_crashes, dense_rank() over (order by report_count desc) as ranking FROM tcbs_r ) SELECT signature, report_count, win_count, lin_count, mac_count, hang_count, plugin_count, content_count, first_report, version_list, %s / total_crashes::float as percent_of_total, startup_count / %s::float as startup_percent, is_gc_count, total_crashes::int FROM tcbs_window ORDER BY %s DESC LIMIT %s """ % (table_to_use, date_range_field, where, order_by, order_by, order_by, dbParams["limit"]) params = ( dbParams['product'], dbParams['version'], dbParams['startDate'], dbParams['to_date'], ) try: cursor = connection.cursor() return db.execute(cursor, sql, params) except Exception: connection.rollback() raise else: connection.commit()