[json_engine] mirror xpath functionality

2025-01-11 02:25:40 +01:00 · 2025-01-11 02:25:40 +01:00 · f51008c414
commit f51008c414
parent e115aa048e
1 changed files with 151 additions and 11 deletions
--- a/searx/engines/json_engine.py
+++ b/searx/engines/json_engine.py
@ -8,6 +8,8 @@ Configuration
 Request:

 - :py:obj:`search_url`
+- :py:obj:`lang_all`
+- :py:obj:`soft_max_redirects`
 - :py:obj:`method`
 - :py:obj:`request_body`
 - :py:obj:`cookies`
@ -19,10 +21,22 @@ Paging:
 - :py:obj:`page_size`
 - :py:obj:`first_page_num`

+Time Range:
+
+- :py:obj:`time_range_support`
+- :py:obj:`time_range_url`
+- :py:obj:`time_range_map`
+
+Safe-Search:
+
+- :py:obj:`safe_search_support`
+- :py:obj:`safe_search_map`
+
 Response:

 - :py:obj:`title_html_to_text`
 - :py:obj:`content_html_to_text`
+- :py:obj:`no_result_for_http_status`

 JSON query:

@ -31,6 +45,8 @@ JSON query:
 - :py:obj:`url_prefix`
 - :py:obj:`title_query`
 - :py:obj:`content_query`
+- :py:obj:`thumbnail_query`
+- :py:obj:`thumbnail_prefix`
 - :py:obj:`suggestion_query`


@ -61,12 +77,13 @@ from collections.abc import Iterable
 from json import loads
 from urllib.parse import urlencode
 from searx.utils import to_string, html_to_text
+from searx.network import raise_for_httperror

 search_url = None
 """
 Search URL of the engine.  Example::

-    https://example.org/?search={query}&page={pageno}
+    https://example.org/?search={query}&page={pageno}{time_range}{safe_search}

 Replacements are:

@ -76,8 +93,41 @@ Replacements are:
 ``{pageno}``:
  Page number if engine supports paging :py:obj:`paging`

+``{lang}``:
+  ISO 639-1 language code (en, de, fr ..)
+
+``{time_range}``:
+  :py:obj:`URL parameter <time_range_url>` if engine :py:obj:`supports time
+  range <time_range_support>`.  The value for the parameter is taken from
+  :py:obj:`time_range_map`.
+
+``{safe_search}``:
+  Safe-search :py:obj:`URL parameter <safe_search_map>` if engine
+  :py:obj:`supports safe-search <safe_search_support>`.  The ``{safe_search}``
+  replacement is taken from the :py:obj:`safes_search_map`.  Filter results::
+
+      0: none, 1: moderate, 2:strict
+
+  If not supported, the URL parameter is an empty string.
+
 """

+lang_all = 'en'
+'''Replacement ``{lang}`` in :py:obj:`search_url` if language ``all`` is
+selected.
+'''
+
+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'''
+
 method = 'GET'
 '''Some engines might require to do POST requests for search.'''

@ -140,6 +190,12 @@ title_query = None
 content_query = None
 '''JSON query of result's ``content``. For the query string documentation see :py:obj:`results_query`'''

+thumbnail_query = False
+'''JSON query of result's ``thumbnail``. For the query string documentation see :py:obj:`results_query`'''
+
+thumbnail_prefix = ''
+'''String to prepend to the result's ``thumbnail``.'''
+
 suggestion_query = ''
 '''JSON query of result's ``suggestion``. For the query string documentation see :py:obj:`results_query`'''

@ -149,6 +205,53 @@ title_html_to_text = False
 content_html_to_text = False
 '''Extract text from a HTML content string'''

+time_range_support = False
+'''Engine supports search time range.'''
+
+time_range_url = '&hours={time_range_val}'
+'''Time range URL parameter in the in :py:obj:`search_url`.  If no time range is
+requested by the user, the URL parameter is an empty string.  The
+``{time_range_val}`` replacement is taken from the :py:obj:`time_range_map`.
+
+.. code:: yaml
+
+    time_range_url : '&days={time_range_val}'
+'''
+
+time_range_map = {
+    'day': 24,
+    'week': 24 * 7,
+    'month': 24 * 30,
+    'year': 24 * 365,
+}
+'''Maps time range value from user to ``{time_range_val}`` in
+:py:obj:`time_range_url`.
+
+.. code:: yaml
+
+    time_range_map:
+      day: 1
+      week: 7
+      month: 30
+      year: 365
+'''
+
+safe_search_support = False
+'''Engine supports safe-search.'''
+
+safe_search_map = {0: '&filter=none', 1: '&filter=moderate', 2: '&filter=strict'}
+'''Maps safe-search value to ``{safe_search}`` in :py:obj:`search_url`.
+
+.. code:: yaml
+
+    safesearch: true
+    safes_search_map:
+      0: '&filter=none'
+      1: '&filter=moderate'
+      2: '&filter=strict'
+
+'''
+

 def iterate(iterable):
    if isinstance(iterable, dict):
@ -207,10 +310,26 @@ def query(data, query_string):

 def request(query, params):  # pylint: disable=redefined-outer-name
    '''Build request parameters (see :ref:`engine request`).'''
-    fp = {'query': urlencode({'q': query})[2:]}  # pylint: disable=invalid-name
+    lang = lang_all
+    if params['language'] != 'all':
+        lang = params['language'][:2]

-    if paging and search_url.find('{pageno}') >= 0:
-        fp['pageno'] = (params['pageno'] - 1) * page_size + first_page_num
+    time_range = ''
+    if params.get('time_range'):
+        time_range_val = time_range_map.get(params.get('time_range'))
+        time_range = time_range_url.format(time_range_val=time_range_val)
+
+    safe_search = ''
+    if params['safesearch']:
+        safe_search = safe_search_map[params['safesearch']]
+
+    fp = {  # pylint: disable=invalid-name
+        'query': urlencode({'q': query})[2:],
+        'lang': lang,
+        'pageno': (params['pageno'] - 1) * page_size + first_page_num,
+        'time_range': time_range,
+        'safe_search': safe_search,
+    }

    params['cookies'].update(cookies)
    params['headers'].update(headers)
@ -223,6 +342,9 @@ def request(query, params):  # pylint: disable=redefined-outer-name
        fp['query'] = query
        params['data'] = request_body.format(**fp)

+    params['soft_max_redirects'] = soft_max_redirects
+    params['raise_for_httperror'] = False
+
    return params


@ -234,10 +356,16 @@ def response(resp):
    '''Scrap *results* from the response (see :ref:`engine results`).'''
    results = []

+    if no_result_for_http_status and resp.status_code in no_result_for_http_status:
+        return results
+
+    raise_for_httperror(resp)
+
    if not resp.text:
        return results

    json = loads(resp.text)
+    is_onion = 'onions' in categories

    title_filter = html_to_text if title_html_to_text else identity
    content_filter = html_to_text if content_html_to_text else identity
@ -256,13 +384,24 @@ def response(resp):
                content = query(result, content_query)[0]
            except:  # pylint: disable=bare-except
                content = ""
-            results.append(
-                {
+
+            tmp_result = {
                'url': url_prefix + to_string(url),
                'title': title_filter(to_string(title)),
                'content': content_filter(to_string(content)),
            }
-            )
+
+            if thumbnail_query:
+                try:
+                    thumbnail_query_result = query(result, thumbnail_query)[0]
+                    tmp_result['thumbnail'] = thumbnail_prefix + to_string(thumbnail_query_result)
+                except:  # pylint: disable=bare-except
+                    continue
+
+            if is_onion:
+                tmp_result['is_onion'] = True
+
+            results.append(tmp_result)
    else:
        for result in json:
            url = query(result, url_query)[0]
@ -274,6 +413,7 @@ def response(resp):
                    'url': url_prefix + to_string(url),
                    'title': title_filter(to_string(title)),
                    'content': content_filter(to_string(content)),
+                    'is_onion': is_onion,
                }
            )