Fix ./manage.sh robot_tests (#1633)

doi_resolvers / default_doi_resolver were missing in the settings_robots.yml file, so the test server was not able to start (crash). Since the output wasn't displayed, it was not obvious why the Selenium couldn't connect to searx.

manage.sh

@@ -35,7 +35,7 @@ install_geckodriver() {
     if [ -z "$NOTFOUND" ]; then
         return
     fi
-    GECKODRIVER_VERSION="v0.19.1"
+    GECKODRIVER_VERSION="v0.24.0"
     PLATFORM="`python -c "import six; import platform; six.print_(platform.system().lower(), platform.architecture()[0])"`"
     case "$PLATFORM" in
         "linux 32bit" | "linux2 32bit") ARCH="linux32";;

requirements-dev.txt

@@ -4,8 +4,8 @@ nose2[coverage_plugin]
 cov-core==1.15.0
 pep8==1.7.0
 plone.testing==5.0.0
-splinter==0.7.5
+splinter==0.11.0
 transifex-client==0.12.2
 unittest2==1.1.0
 zope.testrunner==4.5.1
-selenium==3.5.0
+selenium==3.141.0

searx/settings_robot.yml

@@ -39,3 +39,11 @@ engines:
 locales:
     en : English
     hu : Magyar
+
+doi_resolvers :
+  oadoi.org : 'https://oadoi.org/'
+  doi.org : 'https://doi.org/'
+  doai.io  : 'http://doai.io/'
+  sci-hub.tw : 'http://sci-hub.tw/'
+
+default_doi_resolver : 'oadoi.org'

searx/testing.py

@@ -6,7 +6,6 @@ import os
 import subprocess
 import traceback
 
-
 from os.path import dirname, join, abspath
 
 from splinter import Browser
@@ -49,6 +48,7 @@ class SearxRobotLayer():
         exe = 'python'
 
         # set robot settings path
+        os.environ['SEARX_DEBUG'] = '1'
         os.environ['SEARX_SETTINGS_PATH'] = abspath(
             dirname(__file__) + '/settings_robot.yml')
 
@@ -58,6 +58,8 @@ class SearxRobotLayer():
             stdout=subprocess.PIPE,
             stderr=subprocess.STDOUT
         )
+        if hasattr(self.server.stdout, 'read1'):
+            print(self.server.stdout.read1(1024).decode('utf-8'))
 
     def tearDown(self):
         os.kill(self.server.pid, 9)