__init__.py 6.3 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169
  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. import sys
  15. import threading
  16. from os.path import realpath, dirname
  17. from babel.localedata import locale_identifiers
  18. from urllib.parse import urlparse
  19. from operator import itemgetter
  20. from searx import settings
  21. from searx import logger
  22. from searx.data import ENGINES_LANGUAGES
  23. from searx.exceptions import SearxEngineResponseException
  24. from searx.network import get, initialize as initialize_network, set_context_network_name
  25. from searx.utils import load_module, match_language, get_engine_from_settings, gen_useragent
  26. logger = logger.getChild('engines')
  27. engine_dir = dirname(realpath(__file__))
  28. engines = {}
  29. categories = {'general': []}
  30. babel_langs = [lang_parts[0] + '-' + lang_parts[-1] if len(lang_parts) > 1 else lang_parts[0]
  31. for lang_parts in (lang_code.split('_') for lang_code in locale_identifiers())]
  32. engine_shortcuts = {}
  33. engine_default_args = {'paging': False,
  34. 'categories': ['general'],
  35. 'supported_languages': [],
  36. 'safesearch': False,
  37. 'timeout': settings['outgoing']['request_timeout'],
  38. 'shortcut': '-',
  39. 'disabled': False,
  40. 'enable_http': False,
  41. 'time_range_support': False,
  42. 'engine_type': 'online',
  43. 'display_error_messages': True,
  44. 'tokens': []}
  45. def load_engine(engine_data):
  46. engine_name = engine_data['name']
  47. if '_' in engine_name:
  48. logger.error('Engine name contains underscore: "{}"'.format(engine_name))
  49. sys.exit(1)
  50. if engine_name.lower() != engine_name:
  51. logger.warn('Engine name is not lowercase: "{}", converting to lowercase'.format(engine_name))
  52. engine_name = engine_name.lower()
  53. engine_data['name'] = engine_name
  54. engine_module = engine_data['engine']
  55. try:
  56. engine = load_module(engine_module + '.py', engine_dir)
  57. except (SyntaxError, KeyboardInterrupt, SystemExit, SystemError, ImportError, RuntimeError):
  58. logger.exception('Fatal exception in engine "{}"'.format(engine_module))
  59. sys.exit(1)
  60. except:
  61. logger.exception('Cannot load engine "{}"'.format(engine_module))
  62. return None
  63. for param_name, param_value in engine_data.items():
  64. if param_name == 'engine':
  65. pass
  66. elif param_name == 'categories':
  67. if param_value == 'none':
  68. engine.categories = []
  69. else:
  70. engine.categories = list(map(str.strip, param_value.split(',')))
  71. else:
  72. setattr(engine, param_name, param_value)
  73. for arg_name, arg_value in engine_default_args.items():
  74. if not hasattr(engine, arg_name):
  75. setattr(engine, arg_name, arg_value)
  76. # checking required variables
  77. for engine_attr in dir(engine):
  78. if engine_attr.startswith('_'):
  79. continue
  80. if engine_attr == 'inactive' and getattr(engine, engine_attr) is True:
  81. return None
  82. if getattr(engine, engine_attr) is None:
  83. logger.error('Missing engine config attribute: "{0}.{1}"'
  84. .format(engine.name, engine_attr))
  85. sys.exit(1)
  86. # assign supported languages from json file
  87. if engine_data['name'] in ENGINES_LANGUAGES:
  88. setattr(engine, 'supported_languages', ENGINES_LANGUAGES[engine_data['name']])
  89. # find custom aliases for non standard language codes
  90. if hasattr(engine, 'supported_languages'):
  91. if hasattr(engine, 'language_aliases'):
  92. language_aliases = getattr(engine, 'language_aliases')
  93. else:
  94. language_aliases = {}
  95. for engine_lang in getattr(engine, 'supported_languages'):
  96. iso_lang = match_language(engine_lang, babel_langs, fallback=None)
  97. if iso_lang and iso_lang != engine_lang and not engine_lang.startswith(iso_lang) and \
  98. iso_lang not in getattr(engine, 'supported_languages'):
  99. language_aliases[iso_lang] = engine_lang
  100. setattr(engine, 'language_aliases', language_aliases)
  101. # language_support
  102. setattr(engine, 'language_support', len(getattr(engine, 'supported_languages', [])) > 0)
  103. # assign language fetching method if auxiliary method exists
  104. if hasattr(engine, '_fetch_supported_languages'):
  105. headers = {
  106. 'User-Agent': gen_useragent(),
  107. 'Accept-Language': 'ja-JP,ja;q=0.8,en-US;q=0.5,en;q=0.3', # bing needs a non-English language
  108. }
  109. setattr(engine, 'fetch_supported_languages',
  110. lambda: engine._fetch_supported_languages(get(engine.supported_languages_url, headers=headers)))
  111. # tor related settings
  112. if settings['outgoing'].get('using_tor_proxy'):
  113. # use onion url if using tor.
  114. if hasattr(engine, 'onion_url'):
  115. engine.search_url = engine.onion_url + getattr(engine, 'search_path', '')
  116. elif 'onions' in engine.categories:
  117. # exclude onion engines if not using tor.
  118. return None
  119. engine.timeout += settings['outgoing'].get('extra_proxy_timeout', 0)
  120. for category_name in engine.categories:
  121. categories.setdefault(category_name, []).append(engine)
  122. if engine.shortcut in engine_shortcuts:
  123. logger.error('Engine config error: ambigious shortcut: {0}'.format(engine.shortcut))
  124. sys.exit(1)
  125. engine_shortcuts[engine.shortcut] = engine.name
  126. return engine
  127. def load_engines(engine_list):
  128. global engines, engine_shortcuts
  129. engines.clear()
  130. engine_shortcuts.clear()
  131. for engine_data in engine_list:
  132. engine = load_engine(engine_data)
  133. if engine is not None:
  134. engines[engine.name] = engine
  135. return engines