network.py 14 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295296297298299300301302303304305306307308309310311312313314315316317318319320321322323324325326327328329330331332333334335336337338339340341342343344345346347348349350351352353354355356357358359360361362363364365366367368369370371372373374375376377378379380381
  1. # SPDX-License-Identifier: AGPL-3.0-or-later
  2. # lint: pylint
  3. # pylint: disable=global-statement
  4. # pylint: disable=missing-module-docstring, missing-class-docstring
  5. import atexit
  6. import asyncio
  7. import ipaddress
  8. from itertools import cycle
  9. import httpx
  10. from searx import logger, searx_debug
  11. from .client import new_client, get_loop, AsyncHTTPTransportNoHttp
  12. logger = logger.getChild('network')
  13. DEFAULT_NAME = '__DEFAULT__'
  14. NETWORKS = {}
  15. # requests compatibility when reading proxy settings from settings.yml
  16. PROXY_PATTERN_MAPPING = {
  17. 'http': 'http://',
  18. 'https': 'https://',
  19. 'socks4': 'socks4://',
  20. 'socks5': 'socks5://',
  21. 'socks5h': 'socks5h://',
  22. 'http:': 'http://',
  23. 'https:': 'https://',
  24. 'socks4:': 'socks4://',
  25. 'socks5:': 'socks5://',
  26. 'socks5h:': 'socks5h://',
  27. }
  28. ADDRESS_MAPPING = {
  29. 'ipv4': '0.0.0.0',
  30. 'ipv6': '::'
  31. }
  32. class Network:
  33. __slots__ = (
  34. 'enable_http', 'verify', 'enable_http2',
  35. 'max_connections', 'max_keepalive_connections', 'keepalive_expiry',
  36. 'local_addresses', 'proxies', 'using_tor_proxy', 'max_redirects', 'retries', 'retry_on_http_error',
  37. '_local_addresses_cycle', '_proxies_cycle', '_clients', '_logger'
  38. )
  39. _TOR_CHECK_RESULT = {}
  40. def __init__(
  41. # pylint: disable=too-many-arguments
  42. self,
  43. enable_http=True,
  44. verify=True,
  45. enable_http2=False,
  46. max_connections=None,
  47. max_keepalive_connections=None,
  48. keepalive_expiry=None,
  49. proxies=None,
  50. using_tor_proxy=False,
  51. local_addresses=None,
  52. retries=0,
  53. retry_on_http_error=None,
  54. max_redirects=30,
  55. logger_name=None):
  56. self.enable_http = enable_http
  57. self.verify = verify
  58. self.enable_http2 = enable_http2
  59. self.max_connections = max_connections
  60. self.max_keepalive_connections = max_keepalive_connections
  61. self.keepalive_expiry = keepalive_expiry
  62. self.proxies = proxies
  63. self.using_tor_proxy = using_tor_proxy
  64. self.local_addresses = local_addresses
  65. self.retries = retries
  66. self.retry_on_http_error = retry_on_http_error
  67. self.max_redirects = max_redirects
  68. self._local_addresses_cycle = self.get_ipaddress_cycle()
  69. self._proxies_cycle = self.get_proxy_cycles()
  70. self._clients = {}
  71. self._logger = logger.getChild(logger_name) if logger_name else logger
  72. self.check_parameters()
  73. def check_parameters(self):
  74. for address in self.iter_ipaddresses():
  75. if '/' in address:
  76. ipaddress.ip_network(address, False)
  77. else:
  78. ipaddress.ip_address(address)
  79. if self.proxies is not None and not isinstance(self.proxies, (str, dict)):
  80. raise ValueError('proxies type has to be str, dict or None')
  81. def iter_ipaddresses(self):
  82. local_addresses = self.local_addresses
  83. if not local_addresses:
  84. return
  85. if isinstance(local_addresses, str):
  86. local_addresses = [local_addresses]
  87. for address in local_addresses:
  88. yield address
  89. def get_ipaddress_cycle(self):
  90. while True:
  91. count = 0
  92. for address in self.iter_ipaddresses():
  93. if '/' in address:
  94. for a in ipaddress.ip_network(address, False).hosts():
  95. yield str(a)
  96. count += 1
  97. else:
  98. a = ipaddress.ip_address(address)
  99. yield str(a)
  100. count += 1
  101. if count == 0:
  102. yield None
  103. def iter_proxies(self):
  104. if not self.proxies:
  105. return
  106. # https://www.python-httpx.org/compatibility/#proxy-keys
  107. if isinstance(self.proxies, str):
  108. yield 'all://', [self.proxies]
  109. else:
  110. for pattern, proxy_url in self.proxies.items():
  111. pattern = PROXY_PATTERN_MAPPING.get(pattern, pattern)
  112. if isinstance(proxy_url, str):
  113. proxy_url = [proxy_url]
  114. yield pattern, proxy_url
  115. def get_proxy_cycles(self):
  116. proxy_settings = {}
  117. for pattern, proxy_urls in self.iter_proxies():
  118. proxy_settings[pattern] = cycle(proxy_urls)
  119. while True:
  120. # pylint: disable=stop-iteration-return
  121. yield tuple((pattern, next(proxy_url_cycle)) for pattern, proxy_url_cycle in proxy_settings.items())
  122. async def log_response(self, response: httpx.Response):
  123. request = response.request
  124. status = f"{response.status_code} {response.reason_phrase}"
  125. response_line = f"{response.http_version} {status}"
  126. content_type = response.headers.get("Content-Type")
  127. content_type = f' ({content_type})' if content_type else ''
  128. self._logger.debug(
  129. f'HTTP Request: {request.method} {request.url} "{response_line}"{content_type}'
  130. )
  131. @staticmethod
  132. async def check_tor_proxy(client: httpx.AsyncClient, proxies) -> bool:
  133. if proxies in Network._TOR_CHECK_RESULT:
  134. return Network._TOR_CHECK_RESULT[proxies]
  135. result = True
  136. # ignore client._transport because it is not used with all://
  137. for transport in client._mounts.values(): # pylint: disable=protected-access
  138. if isinstance(transport, AsyncHTTPTransportNoHttp):
  139. continue
  140. if not getattr(transport, '_rdns', False):
  141. result = False
  142. break
  143. else:
  144. response = await client.get('https://check.torproject.org/api/ip')
  145. if not response.json()['IsTor']:
  146. result = False
  147. Network._TOR_CHECK_RESULT[proxies] = result
  148. return result
  149. async def get_client(self, verify=None, max_redirects=None):
  150. verify = self.verify if verify is None else verify
  151. max_redirects = self.max_redirects if max_redirects is None else max_redirects
  152. local_address = next(self._local_addresses_cycle)
  153. proxies = next(self._proxies_cycle) # is a tuple so it can be part of the key
  154. key = (verify, max_redirects, local_address, proxies)
  155. hook_log_response = self.log_response if searx_debug else None
  156. if key not in self._clients or self._clients[key].is_closed:
  157. client = new_client(
  158. self.enable_http,
  159. verify,
  160. self.enable_http2,
  161. self.max_connections,
  162. self.max_keepalive_connections,
  163. self.keepalive_expiry,
  164. dict(proxies),
  165. local_address,
  166. 0,
  167. max_redirects,
  168. hook_log_response
  169. )
  170. if self.using_tor_proxy and not await self.check_tor_proxy(client, proxies):
  171. await client.aclose()
  172. raise httpx.ProxyError('Network configuration problem: not using Tor')
  173. self._clients[key] = client
  174. return self._clients[key]
  175. async def aclose(self):
  176. async def close_client(client):
  177. try:
  178. await client.aclose()
  179. except httpx.HTTPError:
  180. pass
  181. await asyncio.gather(*[close_client(client) for client in self._clients.values()], return_exceptions=False)
  182. @staticmethod
  183. def get_kwargs_clients(kwargs):
  184. kwargs_clients = {}
  185. if 'verify' in kwargs:
  186. kwargs_clients['verify'] = kwargs.pop('verify')
  187. if 'max_redirects' in kwargs:
  188. kwargs_clients['max_redirects'] = kwargs.pop('max_redirects')
  189. return kwargs_clients
  190. def is_valid_respones(self, response):
  191. # pylint: disable=too-many-boolean-expressions
  192. if ((self.retry_on_http_error is True and 400 <= response.status_code <= 599)
  193. or (isinstance(self.retry_on_http_error, list) and response.status_code in self.retry_on_http_error)
  194. or (isinstance(self.retry_on_http_error, int) and response.status_code == self.retry_on_http_error)
  195. ):
  196. return False
  197. return True
  198. async def request(self, method, url, **kwargs):
  199. retries = self.retries
  200. while retries >= 0: # pragma: no cover
  201. kwargs_clients = Network.get_kwargs_clients(kwargs)
  202. client = await self.get_client(**kwargs_clients)
  203. try:
  204. response = await client.request(method, url, **kwargs)
  205. if self.is_valid_respones(response) or retries <= 0:
  206. return response
  207. except (httpx.RequestError, httpx.HTTPStatusError) as e:
  208. if retries <= 0:
  209. raise e
  210. retries -= 1
  211. async def stream(self, method, url, **kwargs):
  212. retries = self.retries
  213. while retries >= 0: # pragma: no cover
  214. kwargs_clients = Network.get_kwargs_clients(kwargs)
  215. client = await self.get_client(**kwargs_clients)
  216. try:
  217. response = client.stream(method, url, **kwargs)
  218. if self.is_valid_respones(response) or retries <= 0:
  219. return response
  220. except (httpx.RequestError, httpx.HTTPStatusError) as e:
  221. if retries <= 0:
  222. raise e
  223. retries -= 1
  224. @classmethod
  225. async def aclose_all(cls):
  226. await asyncio.gather(*[network.aclose() for network in NETWORKS.values()], return_exceptions=False)
  227. def get_network(name=None):
  228. return NETWORKS.get(name or DEFAULT_NAME)
  229. def check_network_configuration():
  230. async def check():
  231. exception_count = 0
  232. for network in NETWORKS.values():
  233. if network.using_tor_proxy:
  234. try:
  235. await network.get_client()
  236. except Exception: # pylint: disable=broad-except
  237. network._logger.exception('Error') # pylint: disable=protected-access
  238. exception_count += 1
  239. return exception_count
  240. future = asyncio.run_coroutine_threadsafe(check(), get_loop())
  241. exception_count = future.result()
  242. if exception_count > 0:
  243. raise RuntimeError("Invalid network configuration")
  244. def initialize(settings_engines=None, settings_outgoing=None):
  245. # pylint: disable=import-outside-toplevel)
  246. from searx.engines import engines
  247. from searx import settings
  248. # pylint: enable=import-outside-toplevel)
  249. settings_engines = settings_engines or settings['engines']
  250. settings_outgoing = settings_outgoing or settings['outgoing']
  251. # default parameters for AsyncHTTPTransport
  252. # see https://github.com/encode/httpx/blob/e05a5372eb6172287458b37447c30f650047e1b8/httpx/_transports/default.py#L108-L121 # pylint: disable=line-too-long
  253. default_params = {
  254. 'enable_http': False,
  255. 'verify': True,
  256. 'enable_http2': settings_outgoing['enable_http2'],
  257. 'max_connections': settings_outgoing['pool_connections'],
  258. 'max_keepalive_connections': settings_outgoing['pool_maxsize'],
  259. 'keepalive_expiry': settings_outgoing['keepalive_expiry'],
  260. 'local_addresses': settings_outgoing['source_ips'],
  261. 'using_tor_proxy': settings_outgoing['using_tor_proxy'],
  262. 'proxies': settings_outgoing['proxies'],
  263. 'max_redirects': settings_outgoing['max_redirects'],
  264. 'retries': settings_outgoing['retries'],
  265. 'retry_on_http_error': None,
  266. }
  267. def new_network(params, logger_name=None):
  268. nonlocal default_params
  269. result = {}
  270. result.update(default_params)
  271. result.update(params)
  272. if logger_name:
  273. result['logger_name'] = logger_name
  274. return Network(**result)
  275. def iter_networks():
  276. nonlocal settings_engines
  277. for engine_spec in settings_engines:
  278. engine_name = engine_spec['name']
  279. engine = engines.get(engine_name)
  280. if engine is None:
  281. continue
  282. network = getattr(engine, 'network', None)
  283. yield engine_name, engine, network
  284. if NETWORKS:
  285. done()
  286. NETWORKS.clear()
  287. NETWORKS[DEFAULT_NAME] = new_network({}, logger_name='default')
  288. NETWORKS['ipv4'] = new_network({'local_addresses': '0.0.0.0'}, logger_name='ipv4')
  289. NETWORKS['ipv6'] = new_network({'local_addresses': '::'}, logger_name='ipv6')
  290. # define networks from outgoing.networks
  291. for network_name, network in settings_outgoing['networks'].items():
  292. NETWORKS[network_name] = new_network(network, logger_name=network_name)
  293. # define networks from engines.[i].network (except references)
  294. for engine_name, engine, network in iter_networks():
  295. if network is None:
  296. network = {}
  297. for attribute_name, attribute_value in default_params.items():
  298. if hasattr(engine, attribute_name):
  299. network[attribute_name] = getattr(engine, attribute_name)
  300. else:
  301. network[attribute_name] = attribute_value
  302. NETWORKS[engine_name] = new_network(network, logger_name=engine_name)
  303. elif isinstance(network, dict):
  304. NETWORKS[engine_name] = new_network(network, logger_name=engine_name)
  305. # define networks from engines.[i].network (references)
  306. for engine_name, engine, network in iter_networks():
  307. if isinstance(network, str):
  308. NETWORKS[engine_name] = NETWORKS[network]
  309. # the /image_proxy endpoint has a dedicated network.
  310. # same parameters than the default network, but HTTP/2 is disabled.
  311. # It decreases the CPU load average, and the total time is more or less the same
  312. if 'image_proxy' not in NETWORKS:
  313. image_proxy_params = default_params.copy()
  314. image_proxy_params['enable_http2'] = False
  315. NETWORKS['image_proxy'] = new_network(image_proxy_params, logger_name='image_proxy')
  316. @atexit.register
  317. def done():
  318. """Close all HTTP client
  319. Avoid a warning at exit
  320. see https://github.com/encode/httpx/blob/1a6e254f72d9fd5694a1c10a28927e193ab4f76b/httpx/_client.py#L1785
  321. Note: since Network.aclose has to be async, it is not possible to call this method on Network.__del__
  322. So Network.aclose is called here using atexit.register
  323. """
  324. try:
  325. loop = get_loop()
  326. if loop:
  327. future = asyncio.run_coroutine_threadsafe(Network.aclose_all(), loop)
  328. # wait 3 seconds to close the HTTP clients
  329. future.result(3)
  330. finally:
  331. NETWORKS.clear()
  332. NETWORKS[DEFAULT_NAME] = Network()