Merge 1c9b28968d into abd9b271bc

during the revision in PR #3955 the query string was accidentally converted into
a list of words, further the query must be quoted before POSTed in the ``data``
field, see ``urllib.parse.quote_plus`` [1]

[1] https://docs.python.org/3/library/urllib.parse.html#urllib.parse.quote_plus

Closed: #4009
Co-Authored-by: @return42

docs/src/searx.botdetection.rst

@@ -53,6 +53,9 @@ Probe HTTP headers
 .. automodule:: searx.botdetection.http_user_agent
   :members:
 
+.. automodule:: searx.botdetection.sec_fetch
+  :members:
+
 .. _botdetection config:
 
 Config

searx/botdetection/_helpers.py

@@ -31,6 +31,9 @@ def dump_request(request: flask.Request):
         + " || Content-Length: %s" % request.headers.get('Content-Length')
         + " || Connection: %s" % request.headers.get('Connection')
         + " || User-Agent: %s" % request.headers.get('User-Agent')
+        + " || Sec-Fetch-Site: %s" % request.headers.get('Sec-Fetch-Site')
+        + " || Sec-Fetch-Mode: %s" % request.headers.get('Sec-Fetch-Mode')
+        + " || Sec-Fetch-Dest: %s" % request.headers.get('Sec-Fetch-Dest')
     )
 
 

searx/botdetection/http_sec_fetch.py

@@ -0,0 +1,59 @@
+# SPDX-License-Identifier: AGPL-3.0-or-later
+"""
+Method ``http_sec_fetch``
+-------------------------
+
+The ``http_sec_fetch`` method protect resources from web attacks with `Fetch
+Metadata`_.  A request is filtered out in case of:
+
+- http header Sec-Fetch-Mode_ is invalid
+- http header Sec-Fetch-Dest_ is invalid
+
+.. _Fetch Metadata:
+   https://developer.mozilla.org/en-US/docs/Glossary/Fetch_metadata_request_header
+
+.. Sec-Fetch-Dest:
+   https://developer.mozilla.org/en-US/docs/Web/API/Request/destination
+
+.. Sec-Fetch-Mode:
+   https://developer.mozilla.org/en-US/docs/Web/API/Request/mode
+
+
+"""
+# pylint: disable=unused-argument
+
+from __future__ import annotations
+from ipaddress import (
+    IPv4Network,
+    IPv6Network,
+)
+
+import flask
+import werkzeug
+
+from . import config
+from ._helpers import logger
+
+
+def filter_request(
+    network: IPv4Network | IPv6Network,
+    request: flask.Request,
+    cfg: config.Config,
+) -> werkzeug.Response | None:
+
+    val = request.headers.get("Sec-Fetch-Mode", "")
+    if val != "navigate":
+        logger.debug("invalid Sec-Fetch-Mode '%s'", val)
+        return flask.redirect(flask.url_for('index'), code=302)
+
+    val = request.headers.get("Sec-Fetch-Site", "")
+    if val not in ('same-origin', 'same-site', 'none'):
+        logger.debug("invalid Sec-Fetch-Site '%s'", val)
+        flask.redirect(flask.url_for('index'), code=302)
+
+    val = request.headers.get("Sec-Fetch-Dest", "")
+    if val != "document":
+        logger.debug("invalid Sec-Fetch-Dest '%s'", val)
+        flask.redirect(flask.url_for('index'), code=302)
+
+    return None

searx/engines/duckduckgo.py

@@ -6,7 +6,7 @@ DuckDuckGo Lite
 
 from typing import TYPE_CHECKING
 import re
-from urllib.parse import urlencode
+from urllib.parse import urlencode, quote_plus
 import json
 import babel
 import lxml.html
@@ -245,10 +245,12 @@ def request(query, params):
 
     # Advanced search syntax ends in CAPTCHA
     # https://duckduckgo.com/duckduckgo-help-pages/results/syntax/
-    query = [
-        x.removeprefix("site:").removeprefix("intitle:").removeprefix("inurl:").removeprefix("filetype:")
-        for x in query.split()
-    ]
+    query = " ".join(
+        [
+            x.removeprefix("site:").removeprefix("intitle:").removeprefix("inurl:").removeprefix("filetype:")
+            for x in query.split()
+        ]
+    )
     eng_region = traits.get_region(params['searxng_locale'], traits.all_locale)
     if eng_region == "wt-wt":
         # https://html.duckduckgo.com/html sets an empty value for "all".
@@ -261,7 +263,7 @@ def request(query, params):
 
     params['url'] = url
     params['method'] = 'POST'
-    params['data']['q'] = query
+    params['data']['q'] = quote_plus(query)
 
     # The API is not documented, so we do some reverse engineering and emulate
     # what https://html.duckduckgo.com/html does when you press "next Page" link

searx/limiter.py

@@ -111,6 +111,7 @@ from searx.botdetection import (
     http_accept_encoding,
     http_accept_language,
     http_user_agent,
+    http_sec_fetch,
     ip_limit,
     ip_lists,
     get_network,
@@ -178,16 +179,17 @@ def filter_request(request: flask.Request) -> werkzeug.Response | None:
         logger.error("BLOCK %s: matched BLOCKLIST - %s", network.compressed, msg)
         return flask.make_response(('IP is on BLOCKLIST - %s' % msg, 429))
 
-    # methods applied on /
+    # methods applied on all requests
 
     for func in [
         http_user_agent,
     ]:
         val = func.filter_request(network, request, cfg)
         if val is not None:
+            logger.debug(f"NOT OK ({func.__name__}): {network}: %s", dump_request(flask.request))
             return val
 
-    # methods applied on /search
+    # methods applied on /search requests
 
     if request.path == '/search':
 
@@ -196,12 +198,15 @@ def filter_request(request: flask.Request) -> werkzeug.Response | None:
             http_accept_encoding,
             http_accept_language,
             http_user_agent,
+            http_sec_fetch,
             ip_limit,
         ]:
             val = func.filter_request(network, request, cfg)
             if val is not None:
+                logger.debug(f"NOT OK ({func.__name__}): {network}: %s", dump_request(flask.request))
                 return val
-    logger.debug(f"OK {network}: %s", dump_request(flask.request))
+        logger.debug(f"OK: {network}: %s", dump_request(flask.request))
+
     return None