def test_prettify_url(self):
data = (('https://searx.me/', 'https://searx.me/'),
(u'https://searx.me/ű', u'https://searx.me/ű'),
('https://searx.me/' + (100 * 'a'), 'https://searx.me/[...]aaaaaaaaaaaaaaaaa'),
(u'https://searx.me/' + (100 * u'ű'), u'https://searx.me/[...]űűűűűűűűűűűűűűűűű'))
for test_url, expected in data:
self.assertEqual(utils.prettify_url(test_url, max_length=32), expected)
def test_prettify_url(self):
data = (('https://searx.me/', 'https://searx.me/'),
('https://searx.me/ű', 'https://searx.me/ű'),
('https://searx.me/' + (100 * 'a'), 'https://searx.me/[...]aaaaaaaaaaaaaaaaa'),
('https://searx.me/' + (100 * 'ű'), 'https://searx.me/[...]űűűűűűűűűűűűűűűűű'))
for test_url, expected in data:
self.assertEqual(utils.prettify_url(test_url, max_length=32), expected)
def index():
"""Render index page.
Supported outputs: html, json, csv, rss.
"""
if not request.args and not request.form:
return render(
'index.html',
)
try:
search = Search(request)
except:
return render(
'index.html',
)
if plugins.call('pre_search', request, locals()):
search.search(request)
plugins.call('post_search', request, locals())
for result in search.result_container.get_ordered_results():
plugins.call('on_result', request, locals())
if not search.paging and engines[result['engine']].paging:
search.paging = True
if search.request_data.get('format', 'html') == 'html':
if 'content' in result:
result['content'] = highlight_content(result['content'],
search.query.encode('utf-8')) # noqa
result['title'] = highlight_content(result['title'],
search.query.encode('utf-8'))
else:
if result.get('content'):
result['content'] = html_to_text(result['content']).strip()
# removing html content and whitespace duplications
result['title'] = ' '.join(html_to_text(result['title']).strip().split())
result['pretty_url'] = prettify_url(result['url'])
# TODO, check if timezone is calculated right
if 'publishedDate' in result:
result['pubdate'] = result['publishedDate'].strftime('%Y-%m-%d %H:%M:%S%z')
if result['publishedDate'].replace(tzinfo=None) >= datetime.now() - timedelta(days=1):
timedifference = datetime.now() - result['publishedDate'].replace(tzinfo=None)
minutes = int((timedifference.seconds / 60) % 60)
hours = int(timedifference.seconds / 60 / 60)
if hours == 0:
result['publishedDate'] = gettext(u'{minutes} minute(s) ago').format(minutes=minutes)
else:
result['publishedDate'] = gettext(u'{hours} hour(s), {minutes} minute(s) ago').format(hours=hours, minutes=minutes) # noqa
else:
result['publishedDate'] = format_date(result['publishedDate'])
if search.request_data.get('format') == 'json':
return Response(json.dumps({'query': search.query,
'results': search.result_container.get_ordered_results()}),
mimetype='application/json')
elif search.request_data.get('format') == 'csv':
csv = UnicodeWriter(cStringIO.StringIO())
keys = ('title', 'url', 'content', 'host', 'engine', 'score')
csv.writerow(keys)
for row in search.result_container.get_ordered_results():
row['host'] = row['parsed_url'].netloc
csv.writerow([row.get(key, '') for key in keys])
csv.stream.seek(0)
response = Response(csv.stream.read(), mimetype='application/csv')
cont_disp = 'attachment;Filename=searx_-_{0}.csv'.format(search.query)
response.headers.add('Content-Disposition', cont_disp)
return response
elif search.request_data.get('format') == 'rss':
response_rss = render(
'opensearch_response_rss.xml',
results=search.result_container.get_ordered_results(),
q=search.request_data['q'],
number_of_results=search.result_container.results_length(),
base_url=get_base_url()
)
return Response(response_rss, mimetype='text/xml')
return render(
'results.html',
results=search.result_container.get_ordered_results(),
q=search.request_data['q'],
selected_categories=search.categories,
paging=search.paging,
pageno=search.pageno,
base_url=get_base_url(),
suggestions=search.result_container.suggestions,
answers=search.result_container.answers,
infoboxes=search.result_container.infoboxes,
theme=get_current_theme_name(),
favicons=global_favicons[themes.index(get_current_theme_name())]
)
def index():
"""Render index page.
Supported outputs: html, json, csv, rss.
"""
# output_format
output_format = request.form.get('format', 'html')
if output_format not in ['html', 'csv', 'json', 'rss']:
output_format = 'html'
# check if there is query
if request.form.get('q') is None:
if output_format == 'html':
return render(
'index.html',
)
else:
return index_error(output_format, 'No query'), 400
# search
search_query = None
raw_text_query = None
result_container = None
try:
search_query, raw_text_query = get_search_query_from_webapp(request.preferences, request.form)
# search = Search(search_query) # without plugins
search = SearchWithPlugins(search_query, request.user_plugins, request)
result_container = search.search()
except Exception as e:
# log exception
logger.exception('search error')
# is it an invalid input parameter or something else ?
if (issubclass(e.__class__, SearxParameterException)):
return index_error(output_format, e.message), 400
else:
return index_error(output_format, gettext('search error')), 500
# results
results = result_container.get_ordered_results()
number_of_results = result_container.results_number()
if number_of_results < result_container.results_length():
number_of_results = 0
# UI
advanced_search = request.form.get('advanced_search', None)
# Server-Timing header
request.timings = result_container.get_timings()
# output
for result in results:
if output_format == 'html':
if 'content' in result and result['content']:
result['content'] = highlight_content(escape(result['content'][:1024]), search_query.query)
if 'title' in result and result['title']:
result['title'] = highlight_content(escape(result['title'] or u''), search_query.query)
else:
if result.get('content'):
result['content'] = html_to_text(result['content']).strip()
# removing html content and whitespace duplications
result['title'] = ' '.join(html_to_text(result['title']).strip().split())
if 'url' in result:
result['pretty_url'] = prettify_url(result['url'])
# TODO, check if timezone is calculated right
if 'publishedDate' in result:
try: # test if publishedDate >= 1900 (datetime module bug)
result['pubdate'] = result['publishedDate'].strftime('%Y-%m-%d %H:%M:%S%z')
except ValueError:
result['publishedDate'] = None
else:
if result['publishedDate'].replace(tzinfo=None) >= datetime.now() - timedelta(days=1):
timedifference = datetime.now() - result['publishedDate'].replace(tzinfo=None)
minutes = int((timedifference.seconds / 60) % 60)
hours = int(timedifference.seconds / 60 / 60)
if hours == 0:
result['publishedDate'] = gettext(u'{minutes} minute(s) ago').format(minutes=minutes)
else:
result['publishedDate'] = gettext(u'{hours} hour(s), {minutes} minute(s) ago').format(hours=hours, minutes=minutes) # noqa
else:
result['publishedDate'] = format_date(result['publishedDate'])
if output_format == 'json':
return Response(json.dumps({'query': search_query.query.decode('utf-8'),
'number_of_results': number_of_results,
'results': results,
'answers': list(result_container.answers),
'corrections': list(result_container.corrections),
'infoboxes': result_container.infoboxes,
'suggestions': list(result_container.suggestions),
'unresponsive_engines': __get_translated_errors(result_container.unresponsive_engines)}, # noqa
default=lambda item: list(item) if isinstance(item, set) else item),
mimetype='application/json')
elif output_format == 'csv':
csv = UnicodeWriter(StringIO())
keys = ('title', 'url', 'content', 'host', 'engine', 'score', 'type')
csv.writerow(keys)
for row in results:
row['host'] = row['parsed_url'].netloc
row['type'] = 'result'
csv.writerow([row.get(key, '') for key in keys])
for a in result_container.answers:
row = {'title': a, 'type': 'answer'}
csv.writerow([row.get(key, '') for key in keys])
for a in result_container.suggestions:
row = {'title': a, 'type': 'suggestion'}
csv.writerow([row.get(key, '') for key in keys])
for a in result_container.corrections:
row = {'title': a, 'type': 'correction'}
csv.writerow([row.get(key, '') for key in keys])
csv.stream.seek(0)
response = Response(csv.stream.read(), mimetype='application/csv')
cont_disp = 'attachment;Filename=searx_-_{0}.csv'.format(search_query.query.decode('utf-8'))
response.headers.add('Content-Disposition', cont_disp)
return response
elif output_format == 'rss':
response_rss = render(
'opensearch_response_rss.xml',
results=results,
answers=result_container.answers,
corrections=result_container.corrections,
suggestions=result_container.suggestions,
q=request.form['q'],
number_of_results=number_of_results,
base_url=get_base_url(),
override_theme='__common__',
)
return Response(response_rss, mimetype='text/xml')
# HTML output format
# suggestions: use RawTextQuery to get the suggestion URLs with the same bang
suggestion_urls = list(map(lambda suggestion: {
'url': raw_text_query.changeSearchQuery(suggestion).getFullQuery(),
'title': suggestion
},
result_container.suggestions))
correction_urls = list(map(lambda correction: {
'url': raw_text_query.changeSearchQuery(correction).getFullQuery(),
'title': correction
},
result_container.corrections))
#
return render(
'results.html',
results=results,
q=request.form['q'],
selected_categories=search_query.categories,
pageno=search_query.pageno,
time_range=search_query.time_range,
number_of_results=format_decimal(number_of_results),
advanced_search=advanced_search,
suggestions=suggestion_urls,
answers=result_container.answers,
corrections=correction_urls,
infoboxes=result_container.infoboxes,
paging=result_container.paging,
unresponsive_engines=__get_translated_errors(result_container.unresponsive_engines),
current_language=match_language(search_query.lang,
LANGUAGE_CODES,
fallback=request.preferences.get_value("language")),
base_url=get_base_url(),
theme=get_current_theme_name(),
favicons=global_favicons[themes.index(get_current_theme_name())],
timeout_limit=request.form.get('timeout_limit', None)
)
def index():
"""Render index page.
Supported outputs: html, json, csv, rss.
"""
if not request.args and not request.form:
return render('index.html', )
try:
search = Search(request)
except:
return render('index.html', )
if plugins.call('pre_search', request, locals()):
search.search(request)
plugins.call('post_search', request, locals())
for result in search.result_container.get_ordered_results():
plugins.call('on_result', request, locals())
if not search.paging and engines[result['engine']].paging:
search.paging = True
if search.request_data.get('format', 'html') == 'html':
if 'content' in result:
result['content'] = highlight_content(
result['content'], search.query.encode('utf-8')) # noqa
result['title'] = highlight_content(result['title'],
search.query.encode('utf-8'))
else:
if result.get('content'):
result['content'] = html_to_text(result['content']).strip()
# removing html content and whitespace duplications
result['title'] = ' '.join(
html_to_text(result['title']).strip().split())
result['pretty_url'] = prettify_url(result['url'])
# TODO, check if timezone is calculated right
if 'publishedDate' in result:
result['pubdate'] = result['publishedDate'].strftime(
'%Y-%m-%d %H:%M:%S%z')
if result['publishedDate'].replace(
tzinfo=None) >= datetime.now() - timedelta(days=1):
timedifference = datetime.now(
) - result['publishedDate'].replace(tzinfo=None)
minutes = int((timedifference.seconds / 60) % 60)
hours = int(timedifference.seconds / 60 / 60)
if hours == 0:
result['publishedDate'] = gettext(
u'{minutes} minute(s) ago').format(minutes=minutes)
else:
result['publishedDate'] = gettext(
u'{hours} hour(s), {minutes} minute(s) ago').format(
hours=hours, minutes=minutes) # noqa
else:
result['publishedDate'] = format_date(result['publishedDate'])
if search.request_data.get('format') == 'json':
return Response(json.dumps({
'query':
search.query,
'results':
search.result_container.get_ordered_results()
}),
mimetype='application/json')
elif search.request_data.get('format') == 'csv':
csv = UnicodeWriter(cStringIO.StringIO())
keys = ('title', 'url', 'content', 'host', 'engine', 'score')
csv.writerow(keys)
for row in search.result_container.get_ordered_results():
row['host'] = row['parsed_url'].netloc
csv.writerow([row.get(key, '') for key in keys])
csv.stream.seek(0)
response = Response(csv.stream.read(), mimetype='application/csv')
cont_disp = 'attachment;Filename=searx_-_{0}.csv'.format(search.query)
response.headers.add('Content-Disposition', cont_disp)
return response
elif search.request_data.get('format') == 'rss':
response_rss = render(
'opensearch_response_rss.xml',
results=search.result_container.get_ordered_results(),
q=search.request_data['q'],
number_of_results=search.result_container.results_length(),
base_url=get_base_url())
return Response(response_rss, mimetype='text/xml')
return render('results.html',
results=search.result_container.get_ordered_results(),
q=search.request_data['q'],
selected_categories=search.categories,
paging=search.paging,
pageno=search.pageno,
base_url=get_base_url(),
suggestions=search.result_container.suggestions,
answers=search.result_container.answers,
infoboxes=search.result_container.infoboxes,
theme=get_current_theme_name(),
favicons=global_favicons[themes.index(
get_current_theme_name())])
def index():
"""Render index page.
Supported outputs: html, json, csv, rss.
"""
# output_format
output_format = request.form.get('format', 'html')
if output_format not in ['html', 'csv', 'json', 'rss']:
output_format = 'html'
# check if there is query
if request.form.get('q') is None:
if output_format == 'html':
return render(
'index.html',
)
else:
return index_error(output_format, 'No query'), 400
# search
search_query = None
result_container = None
try:
search_query = get_search_query_from_webapp(request.preferences, request.form)
# search = Search(search_query) # without plugins
search = SearchWithPlugins(search_query, request.user_plugins, request)
result_container = search.search()
except Exception as e:
# log exception
logger.exception('search error')
# is it an invalid input parameter or something else ?
if (issubclass(e.__class__, SearxParameterException)):
return index_error(output_format, e.message), 400
else:
return index_error(output_format, gettext('search error')), 500
# results
results = result_container.get_ordered_results()
number_of_results = result_container.results_number()
if number_of_results < result_container.results_length():
number_of_results = 0
# UI
advanced_search = request.form.get('advanced_search', None)
# output
for result in results:
if output_format == 'html':
if 'content' in result and result['content']:
result['content'] = highlight_content(escape(result['content'][:1024]), search_query.query)
result['title'] = highlight_content(escape(result['title'] or u''), search_query.query)
else:
if result.get('content'):
result['content'] = html_to_text(result['content']).strip()
# removing html content and whitespace duplications
result['title'] = ' '.join(html_to_text(result['title']).strip().split())
result['pretty_url'] = prettify_url(result['url'])
# TODO, check if timezone is calculated right
if 'publishedDate' in result:
try: # test if publishedDate >= 1900 (datetime module bug)
result['pubdate'] = result['publishedDate'].strftime('%Y-%m-%d %H:%M:%S%z')
except ValueError:
result['publishedDate'] = None
else:
if result['publishedDate'].replace(tzinfo=None) >= datetime.now() - timedelta(days=1):
timedifference = datetime.now() - result['publishedDate'].replace(tzinfo=None)
minutes = int((timedifference.seconds / 60) % 60)
hours = int(timedifference.seconds / 60 / 60)
if hours == 0:
result['publishedDate'] = gettext(u'{minutes} minute(s) ago').format(minutes=minutes)
else:
result['publishedDate'] = gettext(u'{hours} hour(s), {minutes} minute(s) ago').format(hours=hours, minutes=minutes) # noqa
else:
result['publishedDate'] = format_date(result['publishedDate'])
if output_format == 'json':
return Response(json.dumps({'query': search_query.query.decode('utf-8'),
'number_of_results': number_of_results,
'results': results,
'answers': list(result_container.answers),
'corrections': list(result_container.corrections),
'infoboxes': result_container.infoboxes,
'suggestions': list(result_container.suggestions),
'unresponsive_engines': list(result_container.unresponsive_engines)},
default=lambda item: list(item) if isinstance(item, set) else item),
mimetype='application/json')
elif output_format == 'csv':
csv = UnicodeWriter(StringIO())
keys = ('title', 'url', 'content', 'host', 'engine', 'score')
csv.writerow(keys)
for row in results:
row['host'] = row['parsed_url'].netloc
csv.writerow([row.get(key, '') for key in keys])
csv.stream.seek(0)
response = Response(csv.stream.read(), mimetype='application/csv')
cont_disp = 'attachment;Filename=searx_-_{0}.csv'.format(search_query.query)
response.headers.add('Content-Disposition', cont_disp)
return response
elif output_format == 'rss':
response_rss = render(
'opensearch_response_rss.xml',
results=results,
q=request.form['q'],
number_of_results=number_of_results,
base_url=get_base_url(),
override_theme='__common__',
)
return Response(response_rss, mimetype='text/xml')
return render(
'results.html',
results=results,
q=request.form['q'],
selected_categories=search_query.categories,
pageno=search_query.pageno,
time_range=search_query.time_range,
number_of_results=format_decimal(number_of_results),
advanced_search=advanced_search,
suggestions=result_container.suggestions,
answers=result_container.answers,
corrections=result_container.corrections,
infoboxes=result_container.infoboxes,
paging=result_container.paging,
unresponsive_engines=result_container.unresponsive_engines,
current_language=match_language(search_query.lang,
LANGUAGE_CODES,
fallback=settings['search']['language']),
base_url=get_base_url(),
theme=get_current_theme_name(),
favicons=global_favicons[themes.index(get_current_theme_name())]
)
def index():
"""Render index page.
Supported outputs: html, json, csv, rss.
"""
if request.form.get('q') is None:
return render(
'index.html',
)
# search
search_query = None
result_container = None
try:
search_query = get_search_query_from_webapp(request.preferences, request.form)
# search = Search(search_query) # without plugins
search = SearchWithPlugins(search_query, request)
result_container = search.search()
except:
request.errors.append(gettext('search error'))
logger.exception('search error')
return render(
'index.html',
)
results = result_container.get_ordered_results()
# UI
advanced_search = request.form.get('advanced_search', None)
output_format = request.form.get('format', 'html')
if output_format not in ['html', 'csv', 'json', 'rss']:
output_format = 'html'
# output
for result in results:
if output_format == 'html':
if 'content' in result and result['content']:
result['content'] = highlight_content(escape(result['content'][:1024]),
search_query.query.encode('utf-8'))
result['title'] = highlight_content(escape(result['title'] or u''),
search_query.query.encode('utf-8'))
else:
if result.get('content'):
result['content'] = html_to_text(result['content']).strip()
# removing html content and whitespace duplications
result['title'] = ' '.join(html_to_text(result['title']).strip().split())
result['pretty_url'] = prettify_url(result['url'])
# TODO, check if timezone is calculated right
if 'publishedDate' in result:
try: # test if publishedDate >= 1900 (datetime module bug)
result['pubdate'] = result['publishedDate'].strftime('%Y-%m-%d %H:%M:%S%z')
except ValueError:
result['publishedDate'] = None
else:
if result['publishedDate'].replace(tzinfo=None) >= datetime.now() - timedelta(days=1):
timedifference = datetime.now() - result['publishedDate'].replace(tzinfo=None)
minutes = int((timedifference.seconds / 60) % 60)
hours = int(timedifference.seconds / 60 / 60)
if hours == 0:
result['publishedDate'] = gettext(u'{minutes} minute(s) ago').format(minutes=minutes)
else:
result['publishedDate'] = gettext(u'{hours} hour(s), {minutes} minute(s) ago').format(hours=hours, minutes=minutes) # noqa
else:
result['publishedDate'] = format_date(result['publishedDate'])
number_of_results = result_container.results_number()
if number_of_results < result_container.results_length():
number_of_results = 0
if output_format == 'json':
return Response(json.dumps({'query': search_query.query,
'number_of_results': number_of_results,
'results': results,
'answers': list(result_container.answers),
'infoboxes': result_container.infoboxes,
'suggestions': list(result_container.suggestions)}),
mimetype='application/json')
elif output_format == 'csv':
csv = UnicodeWriter(cStringIO.StringIO())
keys = ('title', 'url', 'content', 'host', 'engine', 'score')
csv.writerow(keys)
for row in results:
row['host'] = row['parsed_url'].netloc
csv.writerow([row.get(key, '') for key in keys])
csv.stream.seek(0)
response = Response(csv.stream.read(), mimetype='application/csv')
cont_disp = 'attachment;Filename=searx_-_{0}.csv'.format(search_query.query.encode('utf-8'))
response.headers.add('Content-Disposition', cont_disp)
return response
elif output_format == 'rss':
response_rss = render(
'opensearch_response_rss.xml',
results=results,
q=request.form['q'],
number_of_results=number_of_results,
base_url=get_base_url()
)
return Response(response_rss, mimetype='text/xml')
return render(
'results.html',
results=results,
q=request.form['q'],
selected_categories=search_query.categories,
pageno=search_query.pageno,
time_range=search_query.time_range,
number_of_results=format_decimal(number_of_results),
advanced_search=advanced_search,
suggestions=result_container.suggestions,
answers=result_container.answers,
infoboxes=result_container.infoboxes,
paging=result_container.paging,
base_url=get_base_url(),
theme=get_current_theme_name(),
favicons=global_favicons[themes.index(get_current_theme_name())]
)
def search(self, task):
global number_of_searches
# init vars
requests = []
results_queue = Queue()
results = {}
# increase number of searches
number_of_searches += 1
# set default useragent
# user_agent = request.headers.get('User-Agent', '')
user_agent = gen_useragent()
# start search-reqest for all selected engines
for selected_engine in self.engines:
if selected_engine['name'] not in engines:
continue
engine = engines[selected_engine['name']]
# if paging is not supported, skip
if self.pageno > 1 and not engine.paging:
continue
# if search-language is set and engine does not
# provide language-support, skip
if self.lang != 'all' and not engine.language_support:
continue
# set default request parameters
request_params = default_request_params()
request_params['headers']['User-Agent'] = user_agent
request_params['category'] = selected_engine['category']
request_params['started'] = time()
request_params['pageno'] = self.pageno
if hasattr(engine, 'language') and engine.language:
request_params['language'] = engine.language
else:
request_params['language'] = self.lang
# try:
# 0 = None, 1 = Moderate, 2 = Strict
# request_params['safesearch'] = int(request.cookies.get('safesearch'))
# except Exception:
request_params['safesearch'] = settings['search']['safe_search']
# update request parameters dependent on
# search-engine (contained in engines folder)
engine.request(task['query'].encode('utf-8'), request_params)
# update request parameters dependent on
# search-engine (contained in engines folder)
if request_params['url'] is None:
# TODO add support of offline engines
pass
# create a callback wrapper for the search engine results
callback = make_callback(
selected_engine['name'],
results_queue,
engine.response,
request_params)
# create dictionary which contain all
# informations about the request
request_args = dict(
headers=request_params['headers'],
hooks=dict(response=callback),
cookies=request_params['cookies'],
timeout=engine.timeout,
verify=request_params['verify']
)
# specific type of request (GET or POST)
if request_params['method'] == 'GET':
req = requests_lib.get
else:
req = requests_lib.post
request_args['data'] = request_params['data']
# ignoring empty urls
if not request_params['url']:
continue
# append request to list
requests.append((req, request_params['url'],
request_args,
selected_engine['name']))
if not requests:
return self
# send all search-request
threaded_requests(requests)
while not results_queue.empty():
engine_name, engine_results = results_queue.get_nowait()
# TODO type checks
[self.suggestions.append(x['suggestion'])
for x in list(engine_results)
if 'suggestion' in x
and engine_results.remove(x) is None]
[self.answers.append(x['answer'])
for x in list(engine_results)
if 'answer' in x
and engine_results.remove(x) is None]
self.infoboxes.extend(x for x in list(engine_results)
if 'infobox' in x
and engine_results.remove(x) is None)
results[engine_name] = engine_results
# update engine-specific stats
for engine_name, engine_results in results.items():
engines[engine_name].stats['search_count'] += 1
engines[engine_name].stats['result_count'] += len(engine_results)
# score results and remove duplications
self.results = score_results(results)
# merge infoboxes according to their ids
self.infoboxes = merge_infoboxes(self.infoboxes)
# update engine stats, using calculated score
for result in self.results:
plugins.callAPI('on_result', self.plugins, locals())
for res_engine in result['engines']:
engines[result['engine']] \
.stats['score_count'] += result['score']
result['pretty_url'] = prettify_url(result['url'])
# TODO, check if timezone is calculated right
if 'publishedDate' in result:
result['pubdate'] = result['publishedDate'].strftime('%Y-%m-%d %H:%M:%S%z')
if not self.paging and engines[result['engine']].paging:
self.paging = True
if 'content' in result:
result['content_html'] = highlight_content(result['content'],
self.query.encode('utf-8')) # noqa
result['title_html'] = highlight_content(result['title'],
self.query.encode('utf-8'))
if result.get('content'):
result['content'] = html_to_text(result['content']).strip()
# removing html content and whitespace duplications
result['title'] = ' '.join(html_to_text(result['title']).strip().split())
# return results, suggestions, answers and infoboxes
return self
def index():
"""Render index page.
Supported outputs: html, json, csv, rss.
"""
# output_format
output_format = request.form.get('format', 'html')
if output_format not in ['html', 'csv', 'json', 'rss']:
output_format = 'html'
# check if there is query
if request.form.get('q') is None or len(request.form.get('q')) == 0:
if output_format == 'html':
return render(
'index.html',
)
else:
return index_error(output_format, 'No query'), 400
# search
search_query = None
result_container = None
try:
search_query = get_search_query_from_webapp(request.preferences, request.form)
# search = Search(search_query) # without plugins
search = SearchWithPlugins(search_query, request.user_plugins, request)
result_container = search.search()
except Exception as e:
# log exception
logger.exception('search error')
# is it an invalid input parameter or something else ?
if (issubclass(e.__class__, SearxParameterException)):
return index_error(output_format, e.message), 400
else:
return index_error(output_format, gettext('search error')), 500
# results
results = result_container.get_ordered_results()
number_of_results = result_container.results_number()
if number_of_results < result_container.results_length():
number_of_results = 0
# UI
advanced_search = request.form.get('advanced_search', None)
# output
for result in results:
if output_format == 'html':
if 'content' in result and result['content']:
result['content'] = highlight_content(prettify_content(
escape(result['content'][:1024])), search_query.query)
result['title'] = highlight_content(escape(result['title'] or u''), search_query.query)
else:
if result.get('content'):
result['content'] = html_to_text(result['content']).strip()
# removing html content and whitespace duplications
result['title'] = ' '.join(html_to_text(result['title']).strip().split())
if result.__contains__('showurl'):
result['pretty_url'] = prettify_url(result['showurl'])
else:
result['pretty_url'] = prettify_url(result['url'])
# TODO, check if timezone is calculated right
if 'publishedDate' in result:
try: # test if publishedDate >= 1900 (datetime module bug)
result['pubdate'] = result['publishedDate'].strftime('%Y-%m-%d %H:%M:%S%z')
except ValueError:
result['publishedDate'] = None
else:
if result['publishedDate'].replace(tzinfo=None) >= datetime.now() - timedelta(days=1):
timedifference = datetime.now() - result['publishedDate'].replace(tzinfo=None)
minutes = int((timedifference.seconds / 60) % 60)
hours = int(timedifference.seconds / 60 / 60)
if hours == 0:
result['publishedDate'] = gettext(u'{minutes} minute(s) ago').format(minutes=minutes)
else:
result['publishedDate'] = gettext(u'{hours} hour(s), {minutes} minute(s) ago').format(hours=hours, minutes=minutes) # noqa
else:
result['publishedDate'] = format_date(result['publishedDate'])
return render(
'results.html',
results=results,
q=request.form['q'],
selected_categories=search_query.categories,
pageno=search_query.pageno,
time_range=search_query.time_range if search_query.time_range else "",
number_of_results=format_decimal(round(number_of_results)),
advanced_search=advanced_search,
suggestions=result_container.suggestions,
answers=result_container.answers,
corrections=result_container.corrections,
infoboxes=result_container.infoboxes,
paging=result_container.paging,
unresponsive_engines=result_container.unresponsive_engines,
current_language=match_language(search_query.lang,
LANGUAGE_CODES,
fallback=settings['search']['language']),
base_url=get_base_url(),
theme=get_current_theme_name(),
favicons=global_favicons[themes.index(get_current_theme_name())]
)
def index():
"""Render index page.
Supported outputs: html, json, csv, rss.
"""
if not request.args and not request.form:
return render("index.html")
try:
search = Search(request)
except:
return render("index.html")
if plugins.call("pre_search", request, locals()):
search.search(request)
plugins.call("post_search", request, locals())
for result in search.result_container.get_ordered_results():
plugins.call("on_result", request, locals())
if not search.paging and engines[result["engine"]].paging:
search.paging = True
if search.request_data.get("format", "html") == "html":
if "content" in result:
result["content"] = highlight_content(result["content"], search.query.encode("utf-8")) # noqa
result["title"] = highlight_content(result["title"], search.query.encode("utf-8"))
else:
if result.get("content"):
result["content"] = html_to_text(result["content"]).strip()
# removing html content and whitespace duplications
result["title"] = " ".join(html_to_text(result["title"]).strip().split())
result["pretty_url"] = prettify_url(result["url"])
# TODO, check if timezone is calculated right
if "publishedDate" in result:
try: # test if publishedDate >= 1900 (datetime module bug)
result["pubdate"] = result["publishedDate"].strftime("%Y-%m-%d %H:%M:%S%z")
except ValueError:
result["publishedDate"] = None
else:
if result["publishedDate"].replace(tzinfo=None) >= datetime.now() - timedelta(days=1):
timedifference = datetime.now() - result["publishedDate"].replace(tzinfo=None)
minutes = int((timedifference.seconds / 60) % 60)
hours = int(timedifference.seconds / 60 / 60)
if hours == 0:
result["publishedDate"] = gettext(u"{minutes} minute(s) ago").format(minutes=minutes)
else:
result["publishedDate"] = gettext(u"{hours} hour(s), {minutes} minute(s) ago").format(
hours=hours, minutes=minutes
) # noqa
else:
result["publishedDate"] = format_date(result["publishedDate"])
if search.request_data.get("format") == "json":
return Response(
json.dumps({"query": search.query, "results": search.result_container.get_ordered_results()}),
mimetype="application/json",
)
elif search.request_data.get("format") == "csv":
csv = UnicodeWriter(cStringIO.StringIO())
keys = ("title", "url", "content", "host", "engine", "score")
csv.writerow(keys)
for row in search.result_container.get_ordered_results():
row["host"] = row["parsed_url"].netloc
csv.writerow([row.get(key, "") for key in keys])
csv.stream.seek(0)
response = Response(csv.stream.read(), mimetype="application/csv")
cont_disp = "attachment;Filename=searx_-_{0}.csv".format(search.query.encode("utf-8"))
response.headers.add("Content-Disposition", cont_disp)
return response
elif search.request_data.get("format") == "rss":
response_rss = render(
"opensearch_response_rss.xml",
results=search.result_container.get_ordered_results(),
q=search.request_data["q"],
number_of_results=search.result_container.results_length(),
base_url=get_base_url(),
)
return Response(response_rss, mimetype="text/xml")
return render(
"results.html",
results=search.result_container.get_ordered_results(),
q=search.request_data["q"],
selected_categories=search.categories,
paging=search.paging,
pageno=search.pageno,
base_url=get_base_url(),
suggestions=search.result_container.suggestions,
answers=search.result_container.answers,
infoboxes=search.result_container.infoboxes,
theme=get_current_theme_name(),
favicons=global_favicons[themes.index(get_current_theme_name())],
)
def index():
"""Render index page.
Supported outputs: html, json, csv, rss.
"""
if request.form.get("q") is None:
return render("index.html")
# search
search_query = None
result_container = None
try:
search_query = get_search_query_from_webapp(request.preferences, request.form)
# search = Search(search_query) # without plugins
search = SearchWithPlugins(search_query, request)
result_container = search.search()
except:
request.errors.append(gettext("search error"))
logger.exception("search error")
return render("index.html")
results = result_container.get_ordered_results()
# UI
advanced_search = request.form.get("advanced_search", None)
output_format = request.form.get("format", "html")
if output_format not in ["html", "csv", "json", "rss"]:
output_format = "html"
# output
for result in results:
if output_format == "html":
if "content" in result and result["content"]:
result["content"] = highlight_content(
escape(result["content"][:1024]), search_query.query.encode("utf-8")
)
result["title"] = highlight_content(escape(result["title"] or u""), search_query.query.encode("utf-8"))
else:
if result.get("content"):
result["content"] = html_to_text(result["content"]).strip()
# removing html content and whitespace duplications
result["title"] = " ".join(html_to_text(result["title"]).strip().split())
result["pretty_url"] = prettify_url(result["url"])
# TODO, check if timezone is calculated right
if "publishedDate" in result:
try: # test if publishedDate >= 1900 (datetime module bug)
result["pubdate"] = result["publishedDate"].strftime("%Y-%m-%d %H:%M:%S%z")
except ValueError:
result["publishedDate"] = None
else:
if result["publishedDate"].replace(tzinfo=None) >= datetime.now() - timedelta(days=1):
timedifference = datetime.now() - result["publishedDate"].replace(tzinfo=None)
minutes = int((timedifference.seconds / 60) % 60)
hours = int(timedifference.seconds / 60 / 60)
if hours == 0:
result["publishedDate"] = gettext(u"{minutes} minute(s) ago").format(minutes=minutes)
else:
result["publishedDate"] = gettext(u"{hours} hour(s), {minutes} minute(s) ago").format(
hours=hours, minutes=minutes
) # noqa
else:
result["publishedDate"] = format_date(result["publishedDate"])
number_of_results = result_container.results_number()
if number_of_results < result_container.results_length():
number_of_results = 0
if output_format == "json":
return Response(
json.dumps(
{
"query": search_query.query,
"number_of_results": number_of_results,
"results": results,
"answers": list(result_container.answers),
"infoboxes": result_container.infoboxes,
"suggestions": list(result_container.suggestions),
}
),
mimetype="application/json",
)
elif output_format == "csv":
csv = UnicodeWriter(cStringIO.StringIO())
keys = ("title", "url", "content", "host", "engine", "score")
csv.writerow(keys)
for row in results:
row["host"] = row["parsed_url"].netloc
csv.writerow([row.get(key, "") for key in keys])
csv.stream.seek(0)
response = Response(csv.stream.read(), mimetype="application/csv")
cont_disp = "attachment;Filename=searx_-_{0}.csv".format(search_query.query.encode("utf-8"))
response.headers.add("Content-Disposition", cont_disp)
return response
elif output_format == "rss":
response_rss = render(
"opensearch_response_rss.xml",
results=results,
q=request.form["q"],
number_of_results=number_of_results,
base_url=get_base_url(),
)
return Response(response_rss, mimetype="text/xml")
return render(
"results.html",
results=results,
q=request.form["q"],
selected_categories=search_query.categories,
pageno=search_query.pageno,
time_range=search_query.time_range,
number_of_results=format_decimal(number_of_results),
advanced_search=advanced_search,
suggestions=result_container.suggestions,
answers=result_container.answers,
infoboxes=result_container.infoboxes,
paging=result_container.paging,
base_url=get_base_url(),
theme=get_current_theme_name(),
favicons=global_favicons[themes.index(get_current_theme_name())],
)
def index():
"""Render index page.
Supported outputs: html, json, csv, rss.
"""
if request.form.get('q') is None:
return render(
'index.html',
)
# search
search_query = None
result_container = None
try:
search_query = get_search_query_from_webapp(request.preferences, request.form)
# search = Search(search_query) # without plugins
search = SearchWithPlugins(search_query, request)
result_container = search.search()
except:
request.errors.append(gettext('search error'))
logger.exception('search error')
return render(
'index.html',
)
results = result_container.get_ordered_results()
# UI
advanced_search = request.form.get('advanced_search', None)
output_format = request.form.get('format', 'html')
if output_format not in ['html', 'csv', 'json', 'rss']:
output_format = 'html'
# output
for result in results:
if output_format == 'html':
if 'content' in result and result['content']:
result['content'] = highlight_content(result['content'][:1024], search_query.query.encode('utf-8'))
result['title'] = highlight_content(result['title'], search_query.query.encode('utf-8'))
else:
if result.get('content'):
result['content'] = html_to_text(result['content']).strip()
# removing html content and whitespace duplications
result['title'] = ' '.join(html_to_text(result['title']).strip().split())
result['pretty_url'] = prettify_url(result['url'])
# TODO, check if timezone is calculated right
if 'publishedDate' in result:
try: # test if publishedDate >= 1900 (datetime module bug)
result['pubdate'] = result['publishedDate'].strftime('%Y-%m-%d %H:%M:%S%z')
except ValueError:
result['publishedDate'] = None
else:
if result['publishedDate'].replace(tzinfo=None) >= datetime.now() - timedelta(days=1):
timedifference = datetime.now() - result['publishedDate'].replace(tzinfo=None)
minutes = int((timedifference.seconds / 60) % 60)
hours = int(timedifference.seconds / 60 / 60)
if hours == 0:
result['publishedDate'] = gettext(u'{minutes} minute(s) ago').format(minutes=minutes)
else:
result['publishedDate'] = gettext(u'{hours} hour(s), {minutes} minute(s) ago').format(hours=hours, minutes=minutes) # noqa
else:
result['publishedDate'] = format_date(result['publishedDate'])
number_of_results = result_container.results_number()
if number_of_results < result_container.results_length():
number_of_results = 0
if output_format == 'json':
return Response(json.dumps({'query': search_query.query,
'number_of_results': number_of_results,
'results': results}),
mimetype='application/json')
elif output_format == 'csv':
csv = UnicodeWriter(cStringIO.StringIO())
keys = ('title', 'url', 'content', 'host', 'engine', 'score')
csv.writerow(keys)
for row in results:
row['host'] = row['parsed_url'].netloc
csv.writerow([row.get(key, '') for key in keys])
csv.stream.seek(0)
response = Response(csv.stream.read(), mimetype='application/csv')
cont_disp = 'attachment;Filename=searx_-_{0}.csv'.format(search_query.query.encode('utf-8'))
response.headers.add('Content-Disposition', cont_disp)
return response
elif output_format == 'rss':
response_rss = render(
'opensearch_response_rss.xml',
results=results,
q=request.form['q'],
number_of_results=number_of_results,
base_url=get_base_url()
)
return Response(response_rss, mimetype='text/xml')
return render(
'results.html',
results=results,
q=request.form['q'],
selected_categories=search_query.categories,
pageno=search_query.pageno,
time_range=search_query.time_range,
number_of_results=format_decimal(number_of_results),
advanced_search=advanced_search,
suggestions=result_container.suggestions,
answers=result_container.answers,
infoboxes=result_container.infoboxes,
paging=result_container.paging,
base_url=get_base_url(),
theme=get_current_theme_name(),
favicons=global_favicons[themes.index(get_current_theme_name())]
)