diff options
| author | Markus Heiser <markus.heiser@darmarit.de> | 2019-12-10 17:35:23 +0100 |
|---|---|---|
| committer | Markus Heiser <markus.heiser@darmarit.de> | 2019-12-10 17:35:23 +0100 |
| commit | 9988cdcbc514f799c5a2d2a111c82b8a123bbae3 (patch) | |
| tree | 4f6701f3b443c617e5646b2a9dff3ab55c3084bc /searx/engines/google.py | |
| parent | 56b7e05721394c0490ff43960cb7fa5626e83e3d (diff) | |
| parent | 50e1fcaa5e7369a64829552fc96cd24c787706c9 (diff) | |
Merge branch 'master' of https://github.com/asciimoo/searx into dos2unix
Signed-off-by: Markus Heiser <markus.heiser@darmarit.de>
Diffstat (limited to 'searx/engines/google.py')
| -rw-r--r-- | searx/engines/google.py | 21 |
1 files changed, 7 insertions, 14 deletions
diff --git a/searx/engines/google.py b/searx/engines/google.py index 19bde710d..eed3a044e 100644 --- a/searx/engines/google.py +++ b/searx/engines/google.py @@ -107,13 +107,12 @@ images_path = '/images' supported_languages_url = 'https://www.google.com/preferences?#languages' # specific xpath variables -results_xpath = '//div[@class="g"]' -url_xpath = './/h3/a/@href' -title_xpath = './/h3' -content_xpath = './/span[@class="st"]' -content_misc_xpath = './/div[@class="f slp"]' -suggestion_xpath = '//p[@class="_Bmc"]' -spelling_suggestion_xpath = '//a[@class="spell"]' +results_xpath = '//div[contains(@class, "ZINbbc")]' +url_xpath = './/div[@class="kCrYT"][1]/a/@href' +title_xpath = './/div[@class="kCrYT"][1]/a/div[1]' +content_xpath = './/div[@class="kCrYT"][2]//div[contains(@class, "BNeawe")]//div[contains(@class, "BNeawe")]' +suggestion_xpath = '//div[contains(@class, "ZINbbc")][last()]//div[@class="rVLSBd"]/a//div[contains(@class, "BNeawe")]' +spelling_suggestion_xpath = '//div[@id="scc"]//a' # map : detail location map_address_xpath = './/div[@class="s"]//table//td[2]/span/text()' @@ -199,10 +198,6 @@ def request(query, params): params['headers']['Accept-Language'] = language + ',' + language + '-' + country params['headers']['Accept'] = 'text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8' - # Force Safari 3.1 on Mac OS X (Leopard) user agent to avoid loading the new UI that Searx can't parse - params['headers']['User-Agent'] = ("Mozilla/5.0 (Macintosh; U; Intel Mac OS X 10_5_4)" - "AppleWebKit/525.18 (KHTML, like Gecko) Version/3.1.2 Safari/525.20.1") - params['google_hostname'] = google_hostname return params @@ -274,9 +269,7 @@ def response(resp): content = extract_text_from_dom(result, content_xpath) if content is None: continue - content_misc = extract_text_from_dom(result, content_misc_xpath) - if content_misc is not None: - content = content_misc + "<br />" + content + # append result results.append({'url': url, 'title': title, |