From 8cbc9f2d5823eb984e99e15c963e306610007fa1 Mon Sep 17 00:00:00 2001
From: Alexandre Flament <alex@al-f.net>
Date: Thu, 24 Dec 2020 09:28:16 +0100
Subject: [enh] add checker

---
 searx/search/processors/abstract.py          | 12 ++++++++
 searx/search/processors/online.py            | 44 ++++++++++++++++++++++++++++
 searx/search/processors/online_currency.py   | 10 +++++++
 searx/search/processors/online_dictionary.py | 18 ++++++++++++
 4 files changed, 84 insertions(+)

(limited to 'searx/search/processors')

diff --git a/searx/search/processors/abstract.py b/searx/search/processors/abstract.py
index cf3fd7236..eb8d296ec 100644
--- a/searx/search/processors/abstract.py
+++ b/searx/search/processors/abstract.py
@@ -37,3 +37,15 @@ class EngineProcessor:
     @abstractmethod
     def search(self, query, params, result_container, start_time, timeout_limit):
         pass
+
+    def get_tests(self):
+        tests = getattr(self.engine, 'tests', None)
+        if tests is None:
+            tests = getattr(self.engine, 'additional_tests', {})
+            tests.update(self.get_default_tests())
+            return tests
+        else:
+            return tests
+
+    def get_default_tests(self):
+        return {}
diff --git a/searx/search/processors/online.py b/searx/search/processors/online.py
index b62f8059e..54d63b4c9 100644
--- a/searx/search/processors/online.py
+++ b/searx/search/processors/online.py
@@ -211,3 +211,47 @@ class OnlineProcessor(EngineProcessor):
                 # reset the suspend variables
                 self.engine.continuous_errors = 0
                 self.engine.suspend_end_time = 0
+
+    def get_default_tests(self):
+        tests = {}
+
+        tests['simple'] = {
+            'matrix': {'query': ('time', 'time')},
+            'result_container': ['not_empty'],
+        }
+
+        if getattr(self.engine, 'paging', False):
+            # [1, 2, 3] --> isinstance(l, (list, tuple)) ??
+            tests['paging'] = {
+                'matrix': {'query': 'time',
+                           'pageno': (1, 2, 3)},
+                'result_container': ['not_empty'],
+                'test': ['unique_results']
+            }
+
+        if getattr(self.engine, 'time_range', False):
+            tests['time_range'] = {
+                'matrix': {'query': 'time',
+                           'time_range': (None, 'day')},
+                'result_container': ['not_empty'],
+                'test': ['unique_results']
+            }
+
+        if getattr(self.engine, 'lang', False):
+            tests['lang_fr'] = {
+                'matrix': {'query': 'paris', 'lang': 'fr'},
+                'result_container': ['not_empty', ('has_lang', 'fr')],
+            }
+            tests['lang_en'] = {
+                'matrix': {'query': 'paris', 'lang': 'en'},
+                'result_container': ['not_empty', ('has_lang', 'en')],
+            }
+
+        if getattr(self.engine, 'safesearch', False):
+            tests['safesearch'] = {
+                'matrix': {'query': 'porn',
+                           'safesearch': (0, 2)},
+                'test': ['unique_results']
+            }
+
+        return tests
diff --git a/searx/search/processors/online_currency.py b/searx/search/processors/online_currency.py
index f0e919c03..132c10594 100644
--- a/searx/search/processors/online_currency.py
+++ b/searx/search/processors/online_currency.py
@@ -55,3 +55,13 @@ class OnlineCurrencyProcessor(OnlineProcessor):
         params['from_name'] = iso4217_to_name(from_currency, 'en')
         params['to_name'] = iso4217_to_name(to_currency, 'en')
         return params
+
+    def get_default_tests(self):
+        tests = {}
+
+        tests['currency'] = {
+            'matrix': {'query': '1337 usd in rmb'},
+            'result_container': ['has_answer'],
+        }
+
+        return tests
diff --git a/searx/search/processors/online_dictionary.py b/searx/search/processors/online_dictionary.py
index 8e9ef1620..987c710a1 100644
--- a/searx/search/processors/online_dictionary.py
+++ b/searx/search/processors/online_dictionary.py
@@ -35,3 +35,21 @@ class OnlineDictionaryProcessor(OnlineProcessor):
         params['query'] = query
 
         return params
+
+    def get_default_tests(self):
+        tests = {}
+
+        if getattr(self.engine, 'paging', False):
+            tests['translation_paging'] = {
+                'matrix': {'query': 'en-es house',
+                           'pageno': (1, 2, 3)},
+                'result_container': ['not_empty', ('one_title_contains', 'house')],
+                'test': ['unique_results']
+            }
+        else:
+            tests['translation'] = {
+                'matrix': {'query': 'en-es house'},
+                'result_container': ['not_empty', ('one_title_contains', 'house')],
+            }
+
+        return tests
-- 
cgit v1.2.3


From f3e1bd308f8abb62b3ce0070973e0a494d15b122 Mon Sep 17 00:00:00 2001
From: Alexandre Flament <alex@al-f.net>
Date: Fri, 8 Jan 2021 19:05:56 +0100
Subject: [mod] checker: minor adjustements on the default tests

the query "time" is convinient because most of the search engine will return some results,
but some engines in the general category will return documentation about the HTML tags <time> or <input type="time">
---
 searx/search/processors/online.py | 14 ++++++++------
 1 file changed, 8 insertions(+), 6 deletions(-)

(limited to 'searx/search/processors')

diff --git a/searx/search/processors/online.py b/searx/search/processors/online.py
index 54d63b4c9..0ceb0adf2 100644
--- a/searx/search/processors/online.py
+++ b/searx/search/processors/online.py
@@ -179,15 +179,15 @@ class OnlineProcessor(EngineProcessor):
                 requests_exception = True
             elif (issubclass(e.__class__, SearxEngineCaptchaException)):
                 result_container.add_unresponsive_engine(self.engine_name, 'CAPTCHA required')
-                logger.exception('engine {0} : CAPTCHA')
+                logger.exception('engine {0} : CAPTCHA'.format(self.engine_name))
                 suspended_time = e.suspended_time  # pylint: disable=no-member
             elif (issubclass(e.__class__, SearxEngineTooManyRequestsException)):
                 result_container.add_unresponsive_engine(self.engine_name, 'too many requests')
-                logger.exception('engine {0} : Too many requests')
+                logger.exception('engine {0} : Too many requests'.format(self.engine_name))
                 suspended_time = e.suspended_time  # pylint: disable=no-member
             elif (issubclass(e.__class__, SearxEngineAccessDeniedException)):
                 result_container.add_unresponsive_engine(self.engine_name, 'blocked')
-                logger.exception('engine {0} : Searx is blocked')
+                logger.exception('engine {0} : Searx is blocked'.format(self.engine_name))
                 suspended_time = e.suspended_time  # pylint: disable=no-member
             else:
                 result_container.add_unresponsive_engine(self.engine_name, 'unexpected crash')
@@ -216,22 +216,24 @@ class OnlineProcessor(EngineProcessor):
         tests = {}
 
         tests['simple'] = {
-            'matrix': {'query': ('time', 'time')},
+            'matrix': {'query': ('life', 'computer')},
             'result_container': ['not_empty'],
         }
 
         if getattr(self.engine, 'paging', False):
-            # [1, 2, 3] --> isinstance(l, (list, tuple)) ??
             tests['paging'] = {
                 'matrix': {'query': 'time',
                            'pageno': (1, 2, 3)},
                 'result_container': ['not_empty'],
                 'test': ['unique_results']
             }
+            if 'general' in self.engine.categories:
+                # avoid documentation about HTML tags (<time> and <input type="time">)
+                tests['paging']['matrix']['query'] = 'news'
 
         if getattr(self.engine, 'time_range', False):
             tests['time_range'] = {
-                'matrix': {'query': 'time',
+                'matrix': {'query': 'news',
                            'time_range': (None, 'day')},
                 'result_container': ['not_empty'],
                 'test': ['unique_results']
-- 
cgit v1.2.3