# -*- coding: utf-8 -*-
# Copyright 2009 Martin Borho <martin@borho.net>
# GPL - see License.txt for details
from urllib import urlencode, quote_plus
from baas.core.plugins import Plugin
from baas.core.helpers import strip_tags, xmlify, htmlentities_decode, load_url
try:
# appengine
from django.utils import simplejson
except:
import simplejson
class Gsearch (Plugin):
def __init__(self, config, format='xmpp'):
super(Gsearch,self).__init__(config, format)
self.result_limit = 8
def get_map(self):
"""
returns the command map for the plugin
"""
cmd_map = [
('gnews',self.news),
('gweb', self.web),
('metacritic', self.metacritic),
('imdb', self.imdb),
('wikipedia', self.wikipedia),
('wiktionary', self.wiktionary),
('amazon', self.amazon)]
return cmd_map
def get_limits(self):
"""
returns the limit map for the plugin commands
for e.g. several result pages
"""
limit_map = [
('gnews',self.result_limit),
('gweb', self.result_limit),
('metacritic', self.result_limit),
('imdb', self.result_limit),
('wikipedia', self.result_limit),
('wiktionary', self.result_limit),
('amazon', self.result_limit)]
return limit_map
def get_help(self):
"""
returns the help text for the plugin
"""
additional = '''gnews and gweb can be also combined with lang-codes, like #de, #en, #es etc:
gnews:hamburg #de
gweb:xmpp #de
IMDb search can be narrowed down with #en, #de, #es, #pt, #it or #fr
Amazon search with #com, #co.uk, #co.jp, #cn, #de, #ca, #at or #fr
The different wikipedia or wiktionary versions can be selected via #en, #de, #fr etc.'''
return {
'commands': [
'gnews:word - google news search',
'gweb:word - google web search',
'metacritic:title - search for reviews on metacritc.com',
'imdb:title - search for movie on IMDb',
'wikipedia:thing - search on wikipedia',
'wiktionary:word - search on wiktionary',
'amazon: product - search for products on amazon'],
'additional': [additional],
}
def _api_request(self, mode, params):
url_params = urlencode( params )
api_url = 'http://ajax.googleapis.com/ajax/services/search/'+mode+'?%s' % (url_params)
response = load_url(api_url)
api_response = simplejson.loads(response)
if api_response.get('responseStatus') == 200:
return api_response
else:
return None
def _extract_hits(self, result):
if result:
hits = result.get('responseData',{}).get('results')
# handle single result
if type(hits) == dict:
hits = [hits]
return hits
else:
return None
def _build_query_term(self, term):
return 'intitle:' + ' intitle:'.join(term.split())+' '
def web(self, term):
'''
searches web by yahoo
'''
term = term.strip()
lang = 'en'
page = 1
if term:
(term, page) = self.extract_page_param(term)
if term.find('#')+1:
term, lang = term.split('#',1)
term = term.strip()
params = {
'v':'1.0',
'q':term.encode('utf-8').lower(),
'rsz':'large',
'start':(page-1)*self.result_limit
}
if lang.startswith('lang_'):
# specific lang code
params['lr'] = lang
else:
# short lang code
params['hl'] = lang
params['gl'] = lang
response = self._api_request('web', params)
hits = self._extract_hits(response)
title = 'Web search for %s\n' % term
return self.render(data=hits, title=title)
def news(self, term):
'''
searches news
'''
term = term.strip()
lang = None
if term:
(term, page) = self.extract_page_param(term)
if term.find('#')+1:
term, lang = term.split('#',1)
term = term.strip()
params = {
'v':'1.0',
'q':term.encode('utf-8').lower(),
'rsz':'large',
'start':(page-1)*self.result_limit
#'scoring':'d',
}
if lang:
params['ned'] = lang
response = self._api_request('news', params)
hits = self._extract_hits(response)
title = 'Google news search for %s\n' % term
return self.render(data=hits, title=title)
def metacritic(self, term):
'''
searches metacritic
'''
term = term.strip()
if term:
(term, page) = self.extract_page_param(term)
query = 'site:metacritic.com intitle:reviews %s' % self._build_query_term(term)
params = {
'v':'1.0',
'q':query.encode('utf-8').lower(),
'rsz':'large',
'start':(page-1)*self.result_limit
}
response = self._api_request('web', params)
hits = self._extract_hits(response)
title = 'Reviews for "%s"\n' % term
return self.render(data=hits, title=title)
def imdb(self, term):
'''
searches metacritic
'''
term = term.strip()
lang = None
if term:
(term, page) = self.extract_page_param(term)
if term.find('#')+1:
term, lang = term.split('#',1)
term = term.strip()
tld = lang if lang and lang != 'en' else 'com'
query = 'site:imdb.%s inurl:"/title/" %s' % (tld, self._build_query_term(term))
params = {
'v':'1.0',
'q':query.encode('utf-8').lower(),
'rsz':'large',
'start':(page-1)*self.result_limit
}
response = self._api_request('web', params)
hits = self._extract_hits(response)
title = 'Results on IMDb for "%s"\n' % term
return self.render(data=hits, title=title)
def wikipedia(self, term):
'''
searches metacritic
'''
term = term.strip()
lang = 'en'
if term:
(term, page) = self.extract_page_param(term)
if term.find('#')+1:
term, lang = term.split('#',1)
term = term.strip()
query = 'site:%s.wikipedia.org inurl:"/wiki/" %s' % (lang, term)
params = {
'v':'1.0',
'q':query.encode('utf-8').lower(),
'rsz':'large',
'start':(page-1)*self.result_limit
}
response = self._api_request('web', params)
hits = self._extract_hits(response)
title = 'Wikipedia entries for "%s"\n' % term
return self.render(data=hits, title=title)
def wiktionary(self, term):
'''
searches metacritic
'''
term = term.strip()
lang = 'en'
if term:
(term, page) = self.extract_page_param(term)
if term.find('#')+1:
term, lang = term.split('#',1)
term = term.strip()
query = 'site:%s.wiktionary.org inurl:"/wiki/" %s' % (lang, term)
params = {
'v':'1.0',
'q':query.encode('utf-8').lower(),
'rsz':'large',
'start':(page-1)*self.result_limit,
'hl':lang,
'gl':lang
}
response = self._api_request('web', params)
hits = self._extract_hits(response)
title = 'Wiktionary entries for "%s"\n' % term
return self.render(data=hits, title=title)
def amazon(self, term):
'''
searches metacritic
'''
term = term.strip()
lang = None
if term:
(term, page) = self.extract_page_param(term)
if term.find('#')+1:
term, lang = term.split('#',1)
term = term.strip()
tld = lang if lang else 'com'
query = 'site:www.amazon.%s " %s' % (tld, self._build_query_term(term))
params = {
'v':'1.0',
'q':query.encode('utf-8').lower(),
'rsz':'large',
'start':(page-1)*self.result_limit
}
tld_lang_map = { 'co.uk':'lang_en','com':'lang_en', 'fr':'lang_fr',
'de':'lang_de', 'co.jp':'lang_ja'}
params['lr'] = tld_lang_map.get(lang,'')
response = self._api_request('web', params)
hits = self._extract_hits(response)
title = 'Products on Amazon for "%s"\n' % term
return self.render(data=hits, title=title)
def render_xmpp(self, hits, title):
'''
renders the result for xmpp responses
'''
result = title+"\n"
if hits:
for row in hits:
title = row['titleNoFormatting']
if row.get('publisher'):
title = "%s: %s" % (row.get('publisher'), title)
result += "%s - %s\n" % (htmlentities_decode(title), row['unescapedUrl'])
else:
result += 'No hits found!'
return strip_tags(result)
def render_wave(self, hits, title):
'''
renders the result for wave responses
'''
result = " <br/><br/><b>%s</b><br/>" % xmlify(title)
if hits:
for row in hits:
title = row['titleNoFormatting']
if row.get('publisher'):
title = "%s: %s" % (row.get('publisher'), title)
title = xmlify(htmlentities_decode(title))
result += '<a href="%s">%s</a><br/><br/>' % (xmlify(row['unescapedUrl']), title)
else:
result += 'No hits found!'
return result
