summaryrefslogtreecommitdiff
path: root/searx/engines/xpath.py
diff options
context:
space:
mode:
authorAlexandre FLAMENT <alexandre.flament@hesge.ch>2022-09-02 07:33:20 +0000
committerMarkus Heiser <markus.heiser@darmarit.de>2022-09-04 09:07:28 +0200
commitdd0887be186d208846cdc7c3df13dde020dfa957 (patch)
tree2d2518e194b905407c396d70b2f8f40864869ec8 /searx/engines/xpath.py
parenta15dfa5ee11228881f7a846f89196cbdb69021fb (diff)
downloadsearxng-dd0887be186d208846cdc7c3df13dde020dfa957.tar.gz
searxng-dd0887be186d208846cdc7c3df13dde020dfa957.zip
xpath engine: change raise_for_httperror to no_result_for_http_status
no_result_for_http_status contains a list of HTTP status. These HTTP status are seen an empty result list. In other cases an exception is thrown as usual. Previously raise_for_httperror were ignoring all HTTP error, which make defective engines invisible in the stats.
Diffstat (limited to 'searx/engines/xpath.py')
-rw-r--r--searx/engines/xpath.py23
1 files changed, 18 insertions, 5 deletions
diff --git a/searx/engines/xpath.py b/searx/engines/xpath.py
index 97656705a..f9528e92d 100644
--- a/searx/engines/xpath.py
+++ b/searx/engines/xpath.py
@@ -22,6 +22,7 @@ from urllib.parse import urlencode
from lxml import html
from searx.utils import extract_text, extract_url, eval_xpath, eval_xpath_list
+from searx.network import raise_for_httperror
search_url = None
"""
@@ -60,9 +61,14 @@ lang_all = 'en'
'''Replacement ``{lang}`` in :py:obj:`search_url` if language ``all`` is
selected.
'''
-raise_for_httperror = True
-'''True by default: raise an exception if the HTTP code of response is ``>=
-300``'''
+
+no_result_for_http_status = []
+'''Return empty result for these HTTP status codes instead of throwing an error.
+
+.. code:: yaml
+
+ no_result_for_http_status: []
+'''
soft_max_redirects = 0
'''Maximum redirects, soft limit. Record an error but don't stop the engine'''
@@ -179,12 +185,19 @@ def request(query, params):
params['url'] = search_url.format(**fargs)
params['soft_max_redirects'] = soft_max_redirects
- params['raise_for_httperror'] = raise_for_httperror
+
+ params['raise_for_httperror'] = False
+
return params
-def response(resp):
+def response(resp): # pylint: disable=too-many-branches
'''Scrap *results* from the response (see :ref:`engine results`).'''
+ if no_result_for_http_status and resp.status_code in no_result_for_http_status:
+ return []
+
+ raise_for_httperror(resp)
+
results = []
dom = html.fromstring(resp.text)
is_onion = 'onions' in categories