auf_savoirs_en_partage_django/savoirs/recherche.py

   1 # -*- encoding: utf-8 -*-
   2 import urllib, httplib, time, simplejson, pprint, math, re
   3 from django.conf import settings
   4 from auf_savoirs_en_partage_backend.sep.io import SEP
   5 from savoirs import configuration
   6
   7 def google_search (page, q, data):
   8     params = {'q': q,
   9               'rsz': 'large',
  10               'v': '1.0',
  11               'start': page * configuration['resultats_par_page'],
  12               }
  13     if not settings.DEBUG:
  14         #TODO: corriger ts
  15         params['cref'] = "http://savoirsenpartage.auf.org/recherche.xml?%s" \
  16                 % int(time.time())
  17
  18     url = "/ajax/services/search/web?" + \
  19             urllib.urlencode (params)
  20     handle = httplib.HTTPConnection ('ajax.googleapis.com')
  21     handle.request ("GET", url)
  22     r = handle.getresponse ()
  23     response = simplejson.loads(r.read ())
  24     #print pprint.pformat (params)
  25     #print pprint.pformat (response)
  26     handle.close ()
  27
  28     for i in response['responseData']['cursor']['pages']:
  29         p = int (i['label']) - 1
  30         if p > data['last_page']:
  31             data['last_page'] = p
  32
  33     for r in response['responseData']['results']:
  34         data['results'].append( {'uri': r['url'],
  35                     'content': r['content'],
  36                     'title': r['title']} )
  37
  38     data['more_link'] = response['responseData']['cursor']['moreResultsUrl']
  39
  40
  41 def sep_build_content (regexp, description):
  42     maxlen = 200
  43     content = description
  44     if len (description) > maxlen:
  45         loc = regexp.search (description)
  46         if loc:
  47             f = loc.start () - (maxlen / 2)
  48             t = 0
  49             if f < 0:
  50                 t = -f
  51                 f = 0
  52             t += loc.start () + (maxlen / 2)
  53             if f > 0:
  54                 while description[f] != '.' and f > 0:
  55                     f -= 1
  56                 if f > 0:
  57                     f += 1
  58             if t < len (description):
  59                 while t < len (description) and description[t] != '.':
  60                     t += 1
  61                 t += 1
  62             content = description[f:t]
  63             if f > 0:
  64                 content = "(...) " + content
  65             if t < (len (description) - 1):
  66                 content = content + " (...)"
  67     content = regexp.sub (r'<b>\1</b>', content)
  68     return content
  69
  70
  71 def sep_search (page, q, data):
  72     f = page * configuration['resultats_par_page']
  73     t = f + 8
  74     s = SEP ()
  75     matches = s.search ({'q': q.encode ('utf-8')})
  76     data['last_page'] = math.ceil (float(len (matches)) / \
  77             float(configuration['resultats_par_page'])) - 1
  78     set = s.get (matches[f:t])
  79     print set
  80     regexp = re.compile (r'(%s)' % q, re.IGNORECASE)
  81     for r in set:
  82         uri = r.get ("source", "")
  83         if len (uri) == 0:
  84             uri = r.get ("uri")
  85         title = regexp.sub (r'<b>\1</b>', r.get ("title", ""))
  86         content = sep_build_content (regexp, r.get ("description", ""))
  87
  88         data['results'].append ({'uri': uri, 'id': r.get("uri"), 'title': title, 'content': content})
  89
  90
  91 def cherche (page, q):
  92     rc = {'results': [], 'last_page': 0, 'more_link': ''}
  93
  94     engin = configuration['engin_recherche']
  95
  96     if engin == 'google':
  97         google_search (page, q, rc)
  98
  99     elif engin == 'sep':
 100         sep_search (page, q, rc)
 101
 102     return rc