Merge 5e11d697ab into 0f9694c90b

Engine was added in #2733 but the API does no longer exists. Related:

- https://github.com/searxng/searxng/issues/4038

Signed-off-by: Markus Heiser <markus.heiser@darmarit.de>

docs/index.rst

@@ -4,26 +4,31 @@ Welcome to SearXNG
 
   *Search without being tracked.*
 
-SearXNG is a free internet metasearch engine which aggregates results from more
+.. jinja:: searx
-than 70 search services.  Users are neither tracked nor profiled.  Additionally,
+
-SearXNG can be used over Tor for online anonymity.
+   SearXNG is a free internet metasearch engine which aggregates results from up
+   to {{engines | length}} :ref:`search services <configured engines>`.  Users
+   are neither tracked nor profiled.  Additionally, SearXNG can be used over Tor
+   for online anonymity.
 
 Get started with SearXNG by using one of the instances listed at searx.space_.
 If you don't trust anyone, you can set up your own, see :ref:`installation`.
 
-.. sidebar::  features
+.. jinja:: searx
 
-   - :ref:`self hosted <installation>`
+   .. sidebar::  features
-   - :ref:`no user tracking / no profiling <SearXNG protect privacy>`
+
-   - script & cookies are optional
+      - :ref:`self hosted <installation>`
-   - secure, encrypted connections
+      - :ref:`no user tracking / no profiling <SearXNG protect privacy>`
-   - :ref:`about 200 search engines <configured engines>`
+      - script & cookies are optional
-   - `about 60 translations <https://translate.codeberg.org/projects/searxng/searxng/>`_
+      - secure, encrypted connections
-   - about 100 `well maintained <https://uptime.searxng.org/>`__ instances on searx.space_
+      - :ref:`{{engines | length}} search engines <configured engines>`
-   - :ref:`easy integration of search engines <demo online engine>`
+      - `58 translations <https://translate.codeberg.org/projects/searxng/searxng/>`_
-   - professional development: `CI <https://github.com/searxng/searxng/actions>`_,
+      - about 70 `well maintained <https://uptime.searxng.org/>`__ instances on searx.space_
-     `quality assurance <https://dev.searxng.org/>`_ &
+      - :ref:`easy integration of search engines <demo online engine>`
-     `automated tested UI <https://dev.searxng.org/screenshots.html>`_
+      - professional development: `CI <https://github.com/searxng/searxng/actions>`_,
+	`quality assurance <https://dev.searxng.org/>`_ &
+	`automated tested UI <https://dev.searxng.org/screenshots.html>`_
 
 .. sidebar:: be a part
 

searx/engines/google.py

@@ -320,6 +320,51 @@ def _parse_data_images(dom):
     return data_image_map
 
 
+def _eval_answers(results, dom, xpath):
+    answer_list = eval_xpath(dom, xpath)
+    drop_elements = [
+        './/div[@class="nnFGuf"]',
+        './/script',  # Scripts like the calculator
+        './/table[@class="HOoTuc"]',  # The actual calculator controls
+        './/table[@class="ElumCf"]',  # The actual calculator buttons
+        './/div[@class="mDRaHd"]',  # Instructions with links
+        './/span[@class="W7GCoc CNbPnc"]',  # Feedback
+        './/*[@style="display:none"]',  # Hidden elements
+    ]
+    for item in answer_list:
+        for element in eval_xpath(item, ' | '.join(drop_elements)):
+            element.drop_tree()
+        extracted_table_html = None
+        is_safe = False
+        table_elements = eval_xpath(item, './/table')
+        if table_elements:
+            extracted_table = table_elements[0]
+            extracted_table.attrib.clear()
+            for element in extracted_table.xpath('.//*'):
+                element.attrib.clear()
+            extracted_table.set('cellpadding', '2')
+            extracted_table.set('cellspacing', '2')
+            extracted_table.set('border', '0')
+            extracted_table_html = html.tostring(extracted_table, pretty_print=True, encoding='unicode')
+            is_safe = True
+        for element in eval_xpath(item, './/table'):  # Drop all remaining tables
+            element.drop_tree()
+        answer_content = extract_text(item)
+        if extracted_table_html:
+            answer_content += '<p>' + extracted_table_html
+        url = (eval_xpath(item, '../..//a/@href') + [None])[0]
+        if url and url.startswith('/search?'):  # If the answer is a Google search link, don't use it
+            url = None
+        results.append(
+            {
+                'answer': answer_content,
+                'url': url,
+                'safe': is_safe,
+            }
+        )
+    return results
+
+
 def response(resp):
     """Get response from google's search request"""
     # pylint: disable=too-many-branches, too-many-statements
@@ -331,17 +376,21 @@ def response(resp):
     dom = html.fromstring(resp.text)
     data_image_map = _parse_data_images(dom)
 
-    # results --> answer
+    results = _eval_answers(results, dom, '//div[contains(@class, "card-section")]')  # Look for cards first
-    answer_list = eval_xpath(dom, '//div[contains(@class, "LGOjhe")]')
+    if not results:
-    for item in answer_list:
+        results = _eval_answers(results, dom, '//div[contains(@class, "LGOjhe")]')  # Look for rendered answers next
-        for bubble in eval_xpath(item, './/div[@class="nnFGuf"]'):
+    # Look for JS DOM encoded string answers last
-            bubble.drop_tree()
+    if not results:
-        results.append(
+        pattern = r"'\\x3c.*?\\x3e'"  # These are DOM encoded strings that Google will push into the DOM
-            {
+        matches = re.findall(pattern, resp.text)
-                'answer': extract_text(item),
+        for match in matches:
-                'url': (eval_xpath(item, '../..//a/@href') + [None])[0],
+            decoded_html = match.encode().decode('unicode_escape')
-            }
+            encoded_dom = html.fromstring(decoded_html)
-        )
+            sub_doms = eval_xpath(encoded_dom, '//div[contains(@class, "LGOjhe")]')
+            if sub_doms:
+                if '<span class="hgKElc"><b>' in decoded_html:  # Main answers start with a bold
+                    results = _eval_answers(results, encoded_dom, '//div[contains(@class, "LGOjhe")]')
+                break  # If it's a JS encoded answer, we only want the first one if it has bold above
 
     # parse results
 

searx/engines/internet_archive_scholar.py

@@ -1,71 +0,0 @@
-# SPDX-License-Identifier: AGPL-3.0-or-later
-"""Internet Archive scholar(science)
-"""
-
-from datetime import datetime
-from urllib.parse import urlencode
-from searx.utils import html_to_text
-
-about = {
-    "website": "https://scholar.archive.org/",
-    "wikidata_id": "Q115667709",
-    "official_api_documentation": "https://scholar.archive.org/api/redoc",
-    "use_official_api": True,
-    "require_api_key": False,
-    "results": "JSON",
-}
-categories = ['science', 'scientific publications']
-paging = True
-
-base_url = "https://scholar.archive.org"
-results_per_page = 15
-
-
-def request(query, params):
-    args = {
-        "q": query,
-        "limit": results_per_page,
-        "offset": (params["pageno"] - 1) * results_per_page,
-    }
-    params["url"] = f"{base_url}/search?{urlencode(args)}"
-    params["headers"]["Accept"] = "application/json"
-    return params
-
-
-def response(resp):
-    results = []
-
-    json = resp.json()
-
-    for result in json["results"]:
-        publishedDate, content, doi = None, '', None
-
-        if result['biblio'].get('release_date'):
-            publishedDate = datetime.strptime(result['biblio']['release_date'], "%Y-%m-%d")
-
-        if len(result['abstracts']) > 0:
-            content = result['abstracts'][0].get('body')
-        elif len(result['_highlights']) > 0:
-            content = result['_highlights'][0]
-
-        if len(result['releases']) > 0:
-            doi = result['releases'][0].get('doi')
-
-        results.append(
-            {
-                'template': 'paper.html',
-                'url': result['fulltext']['access_url'],
-                'title': result['biblio'].get('title') or result['biblio'].get('container_name'),
-                'content': html_to_text(content),
-                'publisher': result['biblio'].get('publisher'),
-                'doi': doi,
-                'journal': result['biblio'].get('container_name'),
-                'authors': result['biblio'].get('contrib_names'),
-                'tags': result['tags'],
-                'publishedDate': publishedDate,
-                'issns': result['biblio'].get('issns'),
-                'pdf_url': result['fulltext'].get('access_url'),
-            }
-        )
-
-    return results

searx/engines/loc.py

@@ -27,7 +27,7 @@ categories = ['images']
 paging = True
 
 endpoint = 'photos'
-base_url = 'https://loc.gov'
+base_url = 'https://www.loc.gov'
 search_string = "/{endpoint}/?sp={page}&{query}&fo=json"
 
 

searx/network/network.py

@@ -233,8 +233,7 @@ class Network:
             del kwargs['raise_for_httperror']
         return do_raise_for_httperror
 
-    @staticmethod
+    def patch_response(self, response, do_raise_for_httperror):
-    def patch_response(response, do_raise_for_httperror):
         if isinstance(response, httpx.Response):
             # requests compatibility (response is not streamed)
             # see also https://www.python-httpx.org/compatibility/#checking-for-4xx5xx-responses
@@ -242,8 +241,11 @@ class Network:
 
             # raise an exception
             if do_raise_for_httperror:
-                raise_for_httperror(response)
+                try:
-
+                    raise_for_httperror(response)
+                except:
+                    self._logger.warning(f"HTTP Request failed: {response.request.method} {response.request.url}")
+                    raise
         return response
 
     def is_valid_response(self, response):
@@ -269,7 +271,7 @@ class Network:
                 else:
                     response = await client.request(method, url, **kwargs)
                 if self.is_valid_response(response) or retries <= 0:
-                    return Network.patch_response(response, do_raise_for_httperror)
+                    return self.patch_response(response, do_raise_for_httperror)
             except httpx.RemoteProtocolError as e:
                 if not was_disconnected:
                     # the server has closed the connection:

searx/search/processors/online.py

@@ -137,9 +137,6 @@ class OnlineProcessor(EngineProcessor):
         self.engine.request(query, params)
 
         # ignoring empty urls
-        if params['url'] is None:
-            return None
-
         if not params['url']:
             return None
 

searx/settings.yml

@@ -1622,11 +1622,6 @@ engines:
     api_site: 'askubuntu'
     categories: [it, q&a]
 
-  - name: internetarchivescholar
-    engine: internet_archive_scholar
-    shortcut: ias
-    timeout: 15.0
-
   - name: superuser
     engine: stackexchange
     shortcut: su

searx/templates/simple/results.html

@@ -23,7 +23,11 @@
     <div id="answers" role="complementary" aria-labelledby="answers-title"><h4 class="title" id="answers-title">{{ _('Answers') }} : </h4>
         {%- for answer in answers.values() -%}
         <div class="answer">
-        <span>{{ answer.answer }}</span>
+          {%- if answer.safe -%}
+            <span>{{ answer.answer | safe }}</span>
+          {%- else -%}
+            <span>{{ answer.answer }}</span>
+          {%- endif -%}
           {%- if answer.url -%}
           <a href="{{ answer.url }}" class="answer-url"
              {%- if results_on_new_tab %} target="_blank" rel="noopener noreferrer"