| 123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210 | 
'''searx is free software: you can redistribute it and/or modifyit under the terms of the GNU Affero General Public License as published bythe Free Software Foundation, either version 3 of the License, or(at your option) any later version.searx is distributed in the hope that it will be useful,but WITHOUT ANY WARRANTY; without even the implied warranty ofMERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See theGNU Affero General Public License for more details.You should have received a copy of the GNU Affero General Public Licensealong with searx. If not, see < http://www.gnu.org/licenses/ >.(C) 2013- by Adam Tauber, <asciimoo@gmail.com>'''from os.path import realpath, dirname, splitext, joinimport sysfrom imp import load_sourcefrom flask.ext.babel import gettextfrom operator import itemgetterfrom searx import settingsfrom searx import loggerlogger = logger.getChild('engines')engine_dir = dirname(realpath(__file__))engines = {}categories = {'general': []}engine_shortcuts = {}def load_module(filename):    modname = splitext(filename)[0]    if modname in sys.modules:        del sys.modules[modname]    filepath = join(engine_dir, filename)    module = load_source(modname, filepath)    module.name = modname    return moduledef load_engine(engine_data):    engine_name = engine_data['engine']    engine = load_module(engine_name + '.py')    for param_name in engine_data:        if param_name == 'engine':            continue        if param_name == 'categories':            if engine_data['categories'] == 'none':                engine.categories = []            else:                engine.categories = map(                    str.strip, engine_data['categories'].split(','))            continue        setattr(engine, param_name, engine_data[param_name])    if not hasattr(engine, 'paging'):        engine.paging = False    if not hasattr(engine, 'categories'):        engine.categories = ['general']    if not hasattr(engine, 'language_support'):        engine.language_support = True    if not hasattr(engine, 'timeout'):        engine.timeout = settings['server']['request_timeout']    if not hasattr(engine, 'shortcut'):        engine.shortcut = ''    if not hasattr(engine, 'disabled'):        engine.disabled = False    # checking required variables    for engine_attr in dir(engine):        if engine_attr.startswith('_'):            continue        if getattr(engine, engine_attr) is None:            logger.error('Missing engine config attribute: "{0}.{1}"'                  .format(engine.name, engine_attr))            sys.exit(1)    engine.stats = {        'result_count': 0,        'search_count': 0,        'page_load_time': 0,        'score_count': 0,        'errors': 0    }    if hasattr(engine, 'categories'):        for category_name in engine.categories:            categories.setdefault(category_name, []).append(engine)    else:        categories['general'].append(engine)    if engine.shortcut:        if engine.shortcut in engine_shortcuts:            logger.error('Engine config error: ambigious shortcut: {0}'                  .format(engine.shortcut))            sys.exit(1)        engine_shortcuts[engine.shortcut] = engine.name    return enginedef get_engines_stats():    # TODO refactor    pageloads = []    results = []    scores = []    errors = []    scores_per_result = []    max_pageload = max_results = max_score = max_errors = max_score_per_result = 0  # noqa    for engine in engines.values():        if engine.stats['search_count'] == 0:            continue        results_num = \            engine.stats['result_count'] / float(engine.stats['search_count'])        load_times = engine.stats['page_load_time'] / float(engine.stats['search_count'])  # noqa        if results_num:            score = engine.stats['score_count'] / float(engine.stats['search_count'])  # noqa            score_per_result = score / results_num        else:            score = score_per_result = 0.0        max_results = max(results_num, max_results)        max_pageload = max(load_times, max_pageload)        max_score = max(score, max_score)        max_score_per_result = max(score_per_result, max_score_per_result)        max_errors = max(max_errors, engine.stats['errors'])        pageloads.append({'avg': load_times, 'name': engine.name})        results.append({'avg': results_num, 'name': engine.name})        scores.append({'avg': score, 'name': engine.name})        errors.append({'avg': engine.stats['errors'], 'name': engine.name})        scores_per_result.append({            'avg': score_per_result,            'name': engine.name        })    for engine in pageloads:        if max_pageload:            engine['percentage'] = int(engine['avg'] / max_pageload * 100)        else:            engine['percentage'] = 0    for engine in results:        if max_results:            engine['percentage'] = int(engine['avg'] / max_results * 100)        else:            engine['percentage'] = 0    for engine in scores:        if max_score:            engine['percentage'] = int(engine['avg'] / max_score * 100)        else:            engine['percentage'] = 0    for engine in scores_per_result:        if max_score_per_result:            engine['percentage'] = int(engine['avg']                                       / max_score_per_result * 100)        else:            engine['percentage'] = 0    for engine in errors:        if max_errors:            engine['percentage'] = int(float(engine['avg']) / max_errors * 100)        else:            engine['percentage'] = 0    return [        (            gettext('Page loads (sec)'),            sorted(pageloads, key=itemgetter('avg'))        ),        (            gettext('Number of results'),            sorted(results, key=itemgetter('avg'), reverse=True)        ),        (            gettext('Scores'),            sorted(scores, key=itemgetter('avg'), reverse=True)        ),        (            gettext('Scores per result'),            sorted(scores_per_result, key=itemgetter('avg'), reverse=True)        ),        (            gettext('Errors'),            sorted(errors, key=itemgetter('avg'), reverse=True)        ),    ]if 'engines' not in settings or not settings['engines']:    logger.error('No engines found. Edit your settings.yml')    exit(2)for engine_data in settings['engines']:    engine = load_engine(engine_data)    engines[engine.name] = engine
 |