webapp.py 10 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295296297298299300301302303304305306307308309310311312313314315316317
  1. #!/usr/bin/env python
  2. '''
  3. searx is free software: you can redistribute it and/or modify
  4. it under the terms of the GNU Affero General Public License as published by
  5. the Free Software Foundation, either version 3 of the License, or
  6. (at your option) any later version.
  7. searx is distributed in the hope that it will be useful,
  8. but WITHOUT ANY WARRANTY; without even the implied warranty of
  9. MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  10. GNU Affero General Public License for more details.
  11. You should have received a copy of the GNU Affero General Public License
  12. along with searx. If not, see < http://www.gnu.org/licenses/ >.
  13. (C) 2013- by Adam Tauber, <asciimoo@gmail.com>
  14. '''
  15. import json
  16. import cStringIO
  17. import os
  18. from flask import (
  19. Flask, request, render_template, url_for, Response, make_response,
  20. redirect, send_from_directory
  21. )
  22. from flask.ext.babel import Babel
  23. from searx import settings, searx_dir
  24. from searx.engines import (
  25. search as do_search, categories, engines, get_engines_stats,
  26. engine_shortcuts
  27. )
  28. from searx.utils import UnicodeWriter, highlight_content, html_to_text
  29. from searx.languages import language_codes
  30. from searx.search import Search
  31. app = Flask(
  32. __name__,
  33. static_folder=os.path.join(searx_dir, 'static'),
  34. template_folder=os.path.join(searx_dir, 'templates')
  35. )
  36. app.secret_key = settings['server']['secret_key']
  37. babel = Babel(app)
  38. #TODO configurable via settings.yml
  39. favicons = ['wikipedia', 'youtube', 'vimeo', 'soundcloud',
  40. 'twitter', 'stackoverflow', 'github']
  41. @babel.localeselector
  42. def get_locale():
  43. locale = request.accept_languages.best_match(settings['locales'].keys())
  44. if request.cookies.get('locale', '') in settings['locales']:
  45. locale = request.cookies.get('locale', '')
  46. if 'locale' in request.args\
  47. and request.args['locale'] in settings['locales']:
  48. locale = request.args['locale']
  49. if 'locale' in request.form\
  50. and request.form['locale'] in settings['locales']:
  51. locale = request.form['locale']
  52. return locale
  53. def get_base_url():
  54. if settings['server']['base_url']:
  55. hostname = settings['server']['base_url']
  56. else:
  57. scheme = 'http'
  58. if request.is_secure:
  59. scheme = 'https'
  60. hostname = url_for('index', _external=True, _scheme=scheme)
  61. return hostname
  62. def render(template_name, **kwargs):
  63. global categories
  64. kwargs['categories'] = ['general']
  65. kwargs['categories'].extend(x for x in
  66. sorted(categories.keys()) if x != 'general')
  67. if not 'selected_categories' in kwargs:
  68. kwargs['selected_categories'] = []
  69. cookie_categories = request.cookies.get('categories', '').split(',')
  70. for ccateg in cookie_categories:
  71. if ccateg in categories:
  72. kwargs['selected_categories'].append(ccateg)
  73. if not len(kwargs['selected_categories']):
  74. kwargs['selected_categories'] = ['general']
  75. return render_template(template_name, **kwargs)
  76. @app.route('/', methods=['GET', 'POST'])
  77. def index():
  78. """Render index page.
  79. Supported outputs: html, json, csv, rss.
  80. """
  81. if not request.args and not request.form:
  82. return render('index.html')
  83. try:
  84. search = Search(request)
  85. except:
  86. return render('index.html')
  87. # TODO moar refactor - do_search integration into Search class
  88. search.results, search.suggestions = do_search(search.query,
  89. request,
  90. search.engines,
  91. search.pageno,
  92. search.lang)
  93. for result in search.results:
  94. if not search.paging and engines[result['engine']].paging:
  95. search.paging = True
  96. if search.request_data.get('format', 'html') == 'html':
  97. if 'content' in result:
  98. result['content'] = highlight_content(result['content'],
  99. search.query.encode('utf-8')) # noqa
  100. result['title'] = highlight_content(result['title'],
  101. search.query.encode('utf-8'))
  102. else:
  103. if 'content' in result:
  104. result['content'] = html_to_text(result['content']).strip()
  105. # removing html content and whitespace duplications
  106. result['title'] = ' '.join(html_to_text(result['title'])
  107. .strip().split())
  108. if len(result['url']) > 74:
  109. url_parts = result['url'][:35], result['url'][-35:]
  110. result['pretty_url'] = '{0}[...]{1}'.format(*url_parts)
  111. else:
  112. result['pretty_url'] = result['url']
  113. for engine in result['engines']:
  114. if engine in favicons:
  115. result['favicon'] = engine
  116. if search.request_data.get('format') == 'json':
  117. return Response(json.dumps({'query': search.query,
  118. 'results': search.results}),
  119. mimetype='application/json')
  120. elif search.request_data.get('format') == 'csv':
  121. csv = UnicodeWriter(cStringIO.StringIO())
  122. keys = ('title', 'url', 'content', 'host', 'engine', 'score')
  123. if len(search.results):
  124. csv.writerow(keys)
  125. for row in search.results:
  126. row['host'] = row['parsed_url'].netloc
  127. csv.writerow([row.get(key, '') for key in keys])
  128. csv.stream.seek(0)
  129. response = Response(csv.stream.read(), mimetype='application/csv')
  130. cont_disp = 'attachment;Filename=searx_-_{0}.csv'.format(search.query)
  131. response.headers.add('Content-Disposition', cont_disp)
  132. return response
  133. elif search.request_data.get('format') == 'rss':
  134. response_rss = render(
  135. 'opensearch_response_rss.xml',
  136. results=search.results,
  137. q=search.request_data['q'],
  138. number_of_results=len(search.results),
  139. base_url=get_base_url()
  140. )
  141. return Response(response_rss, mimetype='text/xml')
  142. return render(
  143. 'results.html',
  144. results=search.results,
  145. q=search.request_data['q'],
  146. selected_categories=search.categories,
  147. paging=search.paging,
  148. pageno=search.pageno,
  149. suggestions=search.suggestions
  150. )
  151. @app.route('/about', methods=['GET'])
  152. def about():
  153. """Render about page"""
  154. return render('about.html')
  155. @app.route('/preferences', methods=['GET', 'POST'])
  156. def preferences():
  157. """Render preferences page.
  158. Settings that are going to be saved as cookies."""
  159. lang = None
  160. if request.cookies.get('language')\
  161. and request.cookies['language'] in (x[0] for x in language_codes):
  162. lang = request.cookies['language']
  163. blocked_engines = []
  164. if request.method == 'GET':
  165. blocked_engines = request.cookies.get('blocked_engines', '').split(',')
  166. else:
  167. selected_categories = []
  168. locale = None
  169. for pd_name, pd in request.form.items():
  170. if pd_name.startswith('category_'):
  171. category = pd_name[9:]
  172. if not category in categories:
  173. continue
  174. selected_categories.append(category)
  175. elif pd_name == 'locale' and pd in settings['locales']:
  176. locale = pd
  177. elif pd_name == 'language' and (pd == 'all' or
  178. pd in (x[0] for
  179. x in language_codes)):
  180. lang = pd
  181. elif pd_name.startswith('engine_'):
  182. engine_name = pd_name.replace('engine_', '', 1)
  183. if engine_name in engines:
  184. blocked_engines.append(engine_name)
  185. resp = make_response(redirect('/'))
  186. user_blocked_engines = request.cookies.get('blocked_engines', '').split(',') # noqa
  187. if sorted(blocked_engines) != sorted(user_blocked_engines):
  188. # cookie max age: 4 weeks
  189. resp.set_cookie(
  190. 'blocked_engines', ','.join(blocked_engines),
  191. max_age=60 * 60 * 24 * 7 * 4
  192. )
  193. if locale:
  194. # cookie max age: 4 weeks
  195. resp.set_cookie(
  196. 'locale', locale,
  197. max_age=60 * 60 * 24 * 7 * 4
  198. )
  199. if lang:
  200. # cookie max age: 4 weeks
  201. resp.set_cookie(
  202. 'language', lang,
  203. max_age=60 * 60 * 24 * 7 * 4
  204. )
  205. if selected_categories:
  206. # cookie max age: 4 weeks
  207. resp.set_cookie(
  208. 'categories', ','.join(selected_categories),
  209. max_age=60 * 60 * 24 * 7 * 4
  210. )
  211. return resp
  212. return render('preferences.html',
  213. locales=settings['locales'],
  214. current_locale=get_locale(),
  215. current_language=lang or 'all',
  216. language_codes=language_codes,
  217. categs=categories.items(),
  218. blocked_engines=blocked_engines,
  219. shortcuts={y: x for x, y in engine_shortcuts.items()})
  220. @app.route('/stats', methods=['GET'])
  221. def stats():
  222. """Render engine statistics page."""
  223. global categories
  224. stats = get_engines_stats()
  225. return render('stats.html', stats=stats)
  226. @app.route('/robots.txt', methods=['GET'])
  227. def robots():
  228. return Response("""User-agent: *
  229. Allow: /
  230. Allow: /about
  231. Disallow: /stats
  232. Disallow: /engines
  233. """, mimetype='text/plain')
  234. @app.route('/opensearch.xml', methods=['GET'])
  235. def opensearch():
  236. method = 'post'
  237. # chrome/chromium only supports HTTP GET....
  238. if request.headers.get('User-Agent', '').lower().find('webkit') >= 0:
  239. method = 'get'
  240. ret = render('opensearch.xml', method=method, host=get_base_url())
  241. resp = Response(response=ret,
  242. status=200,
  243. mimetype="application/xml")
  244. return resp
  245. @app.route('/favicon.ico')
  246. def favicon():
  247. return send_from_directory(os.path.join(app.root_path, 'static/img'),
  248. 'favicon.png',
  249. mimetype='image/vnd.microsoft.icon')
  250. def run():
  251. from gevent import monkey
  252. monkey.patch_all()
  253. app.run(
  254. debug=settings['server']['debug'],
  255. use_debugger=settings['server']['debug'],
  256. port=settings['server']['port']
  257. )
  258. if __name__ == "__main__":
  259. run()