diff options
Diffstat (limited to 'searx/engines/google_images.py')
| -rw-r--r-- | searx/engines/google_images.py | 46 |
1 files changed, 24 insertions, 22 deletions
diff --git a/searx/engines/google_images.py b/searx/engines/google_images.py index a380170ca..636913114 100644 --- a/searx/engines/google_images.py +++ b/searx/engines/google_images.py @@ -15,7 +15,6 @@ from json import loads from lxml import html from searx.url_utils import urlencode - # engine dependent config categories = ['images'] paging = True @@ -25,8 +24,6 @@ number_of_results = 100 search_url = 'https://www.google.com/search'\ '?{query}'\ - '&asearch=ichunk'\ - '&async=_id:rg_s,_pms:s'\ '&tbm=isch'\ '&yv=2'\ '&{search_options}' @@ -66,30 +63,35 @@ def request(query, params): def response(resp): results = [] - g_result = loads(resp.text) - - dom = html.fromstring(g_result[1][1]) + dom = html.fromstring(resp.text) # parse results - for result in dom.xpath('//div[@data-ved]'): + for result in dom.xpath('//div[contains(@class, "rg_meta")]/text()'): try: - metadata = loads(''.join(result.xpath('./div[contains(@class, "rg_meta")]/text()'))) + metadata = loads(result) + + img_format = metadata.get('ity', '') + img_width = metadata.get('ow', '') + img_height = metadata.get('oh', '') + if img_width and img_height: + img_format += " {0}x{1}".format(img_width, img_height) + + source = metadata.get('st', '') + source_url = metadata.get('isu', '') + if source_url: + source += " ({0})".format(source_url) + + results.append({'url': metadata['ru'], + 'title': metadata['pt'], + 'content': metadata.get('s', ''), + 'source': source, + 'img_format': img_format, + 'thumbnail_src': metadata['tu'], + 'img_src': metadata['ou'], + 'template': 'images.html'}) + except: continue - thumbnail_src = metadata['tu'] - - # http to https - thumbnail_src = thumbnail_src.replace("http://", "https://") - - # append result - results.append({'url': metadata['ru'], - 'title': metadata['pt'], - 'content': metadata['s'], - 'thumbnail_src': thumbnail_src, - 'img_src': metadata['ou'], - 'template': 'images.html'}) - - # return results return results |