def test_highlight_content(self):
self.assertEqual(webutils.highlight_content(0, None), None)
self.assertEqual(webutils.highlight_content(None, None), None)
self.assertEqual(webutils.highlight_content('', None), None)
self.assertEqual(webutils.highlight_content(False, None), None)
contents = [
'<html></html>'
'not<'
]
for content in contents:
self.assertEqual(webutils.highlight_content(content, None), content)
content = 'a'
query = 'test'
self.assertEqual(webutils.highlight_content(content, query), content)
query = 'a test'
self.assertEqual(webutils.highlight_content(content, query), content)
def test_highlight_content(self):
self.assertEqual(webutils.highlight_content(0, None), None)
self.assertEqual(webutils.highlight_content(None, None), None)
self.assertEqual(webutils.highlight_content('', None), None)
self.assertEqual(webutils.highlight_content(False, None), None)
contents = ['<html></html>' 'not<']
for content in contents:
self.assertEqual(webutils.highlight_content(content, None),
content)
content = 'a'
query = 'test'
self.assertEqual(webutils.highlight_content(content, query), content)
query = 'a test'
self.assertEqual(webutils.highlight_content(content, query), content)
data = ((
'" test "', 'a test string',
'a <span class="highlight">test</span> string'
), (
'"a"', 'this is a test string',
'this is<span class="highlight"> a </span>test string'
), (
'a test', 'this is a test string that matches entire query',
'this is <span class="highlight">a test</span> string that matches entire query'
), ('this a test', 'this is a string to test.', (
'<span class="highlight">this</span> is<span class="highlight"> a </span>'
'string to <span class="highlight">test</span>.'
)), (
'match this "exact phrase"',
'this string contains the exact phrase we want to match',
('<span class="highlight">this</span> string contains the <span class="highlight">exact</span>'
' <span class="highlight">phrase</span> we want to <span class="highlight">match</span>'
)))
for query, content, expected in data:
self.assertEqual(webutils.highlight_content(content, query),
expected)
def search():
"""Search query in q and return results.
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 (not None and not an empty string)
if not request.form.get('q'):
if output_format == 'html':
return render(
'index.html',
advanced_search=request.preferences.get_value('advanced_search'),
selected_categories=get_selected_categories(request.preferences, request.form),
)
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 SearxParameterException as e:
logger.exception('search error: SearxParameterException')
return index_error(output_format, e.message), 400
except Exception as e:
logger.exception('search error')
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
# checkin for a external bang
if result_container.redirect_url:
return redirect(result_container.redirect_url)
# 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 ''), 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('{minutes} minute(s) ago').format(minutes=minutes)
else:
result['publishedDate'] = gettext('{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,
'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)
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.changeQuery(suggestion).getFullQuery(),
'title': suggestion
},
result_container.suggestions))
correction_urls = list(map(lambda correction: {
'url': raw_text_query.changeQuery(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),
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)
)
