__init__.py 6.5 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207
  1. '''
  2. searx is free software: you can redistribute it and/or modify
  3. it under the terms of the GNU Affero General Public License as published by
  4. the Free Software Foundation, either version 3 of the License, or
  5. (at your option) any later version.
  6. searx is distributed in the hope that it will be useful,
  7. but WITHOUT ANY WARRANTY; without even the implied warranty of
  8. MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  9. GNU Affero General Public License for more details.
  10. You should have received a copy of the GNU Affero General Public License
  11. along with searx. If not, see < http://www.gnu.org/licenses/ >.
  12. (C) 2013- by Adam Tauber, <asciimoo@gmail.com>
  13. '''
  14. from os.path import realpath, dirname, splitext, join
  15. import sys
  16. from imp import load_source
  17. from flask_babel import gettext
  18. from operator import itemgetter
  19. from searx import settings
  20. from searx import logger
  21. logger = logger.getChild('engines')
  22. engine_dir = dirname(realpath(__file__))
  23. engines = {}
  24. categories = {'general': []}
  25. engine_shortcuts = {}
  26. engine_default_args = {'paging': False,
  27. 'categories': ['general'],
  28. 'language_support': True,
  29. 'safesearch': False,
  30. 'timeout': settings['outgoing']['request_timeout'],
  31. 'shortcut': '-',
  32. 'disabled': False,
  33. 'suspend_end_time': 0,
  34. 'continuous_errors': 0}
  35. def load_module(filename):
  36. modname = splitext(filename)[0]
  37. if modname in sys.modules:
  38. del sys.modules[modname]
  39. filepath = join(engine_dir, filename)
  40. module = load_source(modname, filepath)
  41. module.name = modname
  42. return module
  43. def load_engine(engine_data):
  44. engine_name = engine_data['engine']
  45. try:
  46. engine = load_module(engine_name + '.py')
  47. except:
  48. logger.exception('Cannot load engine "{}"'.format(engine_name))
  49. return None
  50. for param_name in engine_data:
  51. if param_name == 'engine':
  52. continue
  53. if param_name == 'categories':
  54. if engine_data['categories'] == 'none':
  55. engine.categories = []
  56. else:
  57. engine.categories = map(
  58. str.strip, engine_data['categories'].split(','))
  59. continue
  60. setattr(engine, param_name, engine_data[param_name])
  61. for arg_name, arg_value in engine_default_args.iteritems():
  62. if not hasattr(engine, arg_name):
  63. setattr(engine, arg_name, arg_value)
  64. # checking required variables
  65. for engine_attr in dir(engine):
  66. if engine_attr.startswith('_'):
  67. continue
  68. if getattr(engine, engine_attr) is None:
  69. logger.error('Missing engine config attribute: "{0}.{1}"'
  70. .format(engine.name, engine_attr))
  71. sys.exit(1)
  72. engine.stats = {
  73. 'result_count': 0,
  74. 'search_count': 0,
  75. 'page_load_time': 0,
  76. 'score_count': 0,
  77. 'errors': 0
  78. }
  79. for category_name in engine.categories:
  80. categories.setdefault(category_name, []).append(engine)
  81. if engine.shortcut in engine_shortcuts:
  82. logger.error('Engine config error: ambigious shortcut: {0}'.format(engine.shortcut))
  83. sys.exit(1)
  84. engine_shortcuts[engine.shortcut] = engine.name
  85. return engine
  86. def get_engines_stats():
  87. # TODO refactor
  88. pageloads = []
  89. results = []
  90. scores = []
  91. errors = []
  92. scores_per_result = []
  93. max_pageload = max_results = max_score = max_errors = max_score_per_result = 0 # noqa
  94. for engine in engines.values():
  95. if engine.stats['search_count'] == 0:
  96. continue
  97. results_num = \
  98. engine.stats['result_count'] / float(engine.stats['search_count'])
  99. load_times = engine.stats['page_load_time'] / float(engine.stats['search_count']) # noqa
  100. if results_num:
  101. score = engine.stats['score_count'] / float(engine.stats['search_count']) # noqa
  102. score_per_result = score / results_num
  103. else:
  104. score = score_per_result = 0.0
  105. max_results = max(results_num, max_results)
  106. max_pageload = max(load_times, max_pageload)
  107. max_score = max(score, max_score)
  108. max_score_per_result = max(score_per_result, max_score_per_result)
  109. max_errors = max(max_errors, engine.stats['errors'])
  110. pageloads.append({'avg': load_times, 'name': engine.name})
  111. results.append({'avg': results_num, 'name': engine.name})
  112. scores.append({'avg': score, 'name': engine.name})
  113. errors.append({'avg': engine.stats['errors'], 'name': engine.name})
  114. scores_per_result.append({
  115. 'avg': score_per_result,
  116. 'name': engine.name
  117. })
  118. for engine in pageloads:
  119. if max_pageload:
  120. engine['percentage'] = int(engine['avg'] / max_pageload * 100)
  121. else:
  122. engine['percentage'] = 0
  123. for engine in results:
  124. if max_results:
  125. engine['percentage'] = int(engine['avg'] / max_results * 100)
  126. else:
  127. engine['percentage'] = 0
  128. for engine in scores:
  129. if max_score:
  130. engine['percentage'] = int(engine['avg'] / max_score * 100)
  131. else:
  132. engine['percentage'] = 0
  133. for engine in scores_per_result:
  134. if max_score_per_result:
  135. engine['percentage'] = int(engine['avg']
  136. / max_score_per_result * 100)
  137. else:
  138. engine['percentage'] = 0
  139. for engine in errors:
  140. if max_errors:
  141. engine['percentage'] = int(float(engine['avg']) / max_errors * 100)
  142. else:
  143. engine['percentage'] = 0
  144. return [
  145. (
  146. gettext('Page loads (sec)'),
  147. sorted(pageloads, key=itemgetter('avg'))
  148. ),
  149. (
  150. gettext('Number of results'),
  151. sorted(results, key=itemgetter('avg'), reverse=True)
  152. ),
  153. (
  154. gettext('Scores'),
  155. sorted(scores, key=itemgetter('avg'), reverse=True)
  156. ),
  157. (
  158. gettext('Scores per result'),
  159. sorted(scores_per_result, key=itemgetter('avg'), reverse=True)
  160. ),
  161. (
  162. gettext('Errors'),
  163. sorted(errors, key=itemgetter('avg'), reverse=True)
  164. ),
  165. ]
  166. if 'engines' not in settings or not settings['engines']:
  167. logger.error('No engines found. Edit your settings.yml')
  168. exit(2)
  169. for engine_data in settings['engines']:
  170. engine = load_engine(engine_data)
  171. if engine is not None:
  172. engines[engine.name] = engine