[fix] xpath.py: fix safesearch handling if safesearch set to disabled

searx/engines/sogou.py

@@ -1,9 +1,12 @@
 # SPDX-License-Identifier: AGPL-3.0-or-later
 """Sogou search engine for searxng"""
 
+import re
+from datetime import datetime
 from urllib.parse import urlencode
 from lxml import html
 
+from searx.exceptions import SearxEngineCaptchaException
 from searx.utils import extract_text
 
 # Metadata
@@ -21,7 +24,12 @@ categories = ["general"]
 paging = True
 time_range_support = True
 
-time_range_dict = {'day': 'inttime_day', 'week': 'inttime_week', 'month': 'inttime_month', 'year': 'inttime_year'}
+time_range_dict = {
+    "day": "inttime_day",
+    "week": "inttime_week",
+    "month": "inttime_month",
+    "year": "inttime_year",
+}
 
 # Base URL
 base_url = "https://www.sogou.com"
@@ -33,36 +41,99 @@ def request(query, params):
         "page": params["pageno"],
     }
 
-    if time_range_dict.get(params['time_range']):
+    if time_range_dict.get(params["time_range"]):
-        query_params["s_from"] = time_range_dict.get(params['time_range'])
+        query_params["s_from"] = time_range_dict.get(params["time_range"])
         query_params["tsn"] = 1
 
+    params["allow_redirects"] = False
     params["url"] = f"{base_url}/web?{urlencode(query_params)}"
     return params
 
 
 def response(resp):
+    if (
+        resp.status_code == 302
+        and resp.next_request is not None
+        and str(resp.next_request.url).startswith("http://www.sogou.com/antispider")
+    ):
+        raise SearxEngineCaptchaException()
+
     dom = html.fromstring(resp.text)
     results = []
 
-    for item in dom.xpath('//div[contains(@class, "vrwrap")]'):
+    # pylint: disable=line-too-long
-        title = extract_text(item.xpath('.//h3[contains(@class, "vr-title")]/a'))
+    for item in dom.xpath(
-        url = extract_text(item.xpath('.//h3[contains(@class, "vr-title")]/a/@href'))
+        '//div[contains(@class, "rb")] | //div[contains(@class, "vrwrap") and not(.//div[contains(@class, "special-wrap")])]'
+    ):
+        item_html = html.tostring(item, encoding="unicode")
 
-        if url.startswith("/link?url="):
+        if item.xpath('.//h3[@class="pt"]/a'):
-            url = f"{base_url}{url}"
+            result = _parse_results(item, item_html)
+        elif item.xpath('.//h3[contains(@class, "vr-title")]/a'):
+            result = _parse_results_with_image(item, item_html)
+        else:
+            continue
 
-        content = extract_text(item.xpath('.//div[contains(@class, "text-layout")]//p[contains(@class, "star-wiki")]'))
+        if result["title"] and result["url"]:
-        if not content:
+            results.append(result)
-            content = extract_text(item.xpath('.//div[contains(@class, "fz-mid space-txt")]'))
-
-        if title and url:
-            results.append(
-                {
-                    "title": title,
-                    "url": url,
-                    "content": content,
-                }
-            )
 
     return results
+
+
+def _extract_url(url, item_html):
+    if url and url.startswith("/link?url="):
+        match = re.search(r'data-url="([^"]+)"', item_html)
+        if match:
+            return match.group(1)
+        return f"{base_url}{url}"
+    return url
+
+
+def _parse_date(text):
+    if text:
+        text = text.strip().lstrip("-").strip()
+        date_match = re.search(r"(\d{4}-\d{1,2}-\d{1,2})", text)
+        if date_match:
+            try:
+                return datetime.strptime(date_match.group(1), "%Y-%m-%d")
+            except (ValueError, TypeError):
+                pass
+    return None
+
+
+def _parse_results(item, item_html):
+    title = extract_text(item.xpath('.//h3[@class="pt"]/a'))
+    content = extract_text(item.xpath('.//div[@class="ft"]'))
+    url = _extract_url(extract_text(item.xpath('.//h3[@class="pt"]/a/@href')), item_html)
+    publishedDate = _parse_date(extract_text(item.xpath(".//cite")))
+    return {
+        "title": title,
+        "url": url,
+        "content": content,
+        "publishedDate": publishedDate,
+    }
+
+
+def _parse_results_with_image(item, item_html):
+    title = extract_text(item.xpath('.//h3[contains(@class, "vr-title")]/a'))
+    content = extract_text(item.xpath('.//div[contains(@class, "attribute-centent")]'))
+    if not content:
+        content = extract_text(item.xpath('.//div[contains(@class, "fz-mid space-txt")]'))
+    url = _extract_url(extract_text(item.xpath('.//h3[contains(@class, "vr-title")]/a/@href')), item_html)
+    publishedDate = _parse_date(extract_text(item.xpath('.//span[@class="cite-date"]')))
+
+    thumbnail = None
+    try:
+        thumbnail_src = extract_text(item.xpath('.//div[contains(@class, "img-layout")]//img/@src'))
+        if thumbnail_src:
+            thumbnail = thumbnail_src.replace("http://", "https://")
+    except (ValueError, TypeError):
+        pass
+
+    return {
+        "title": title,
+        "url": url,
+        "content": content,
+        "publishedDate": publishedDate,
+        "thumbnail": thumbnail,
+    }

searx/engines/xpath.py

@@ -234,8 +234,9 @@ def request(query, params):
         time_range = time_range_url.format(time_range_val=time_range_val)
 
     safe_search = ''
-    if params['safesearch']:
+    safe_search_val = params.get('safesearch')
-        safe_search = safe_search_map[params['safesearch']]
+    if safe_search_val is not None:
+        safe_search = safe_search_map[safe_search_val]
 
     fargs = {
         'query': urlencode({'q': query})[2:],

searx/settings.yml

@@ -901,7 +901,12 @@ engines:
 
   - name: fynd
     engine: xpath
-    search_url: https://fynd.bot/?search={query}&offset={pageno}
+    search_url: https://fynd.bot/?search={query}&offset={pageno}{safe_search}
+    safesearch: true
+    safe_search_map:
+      0: '&safe=0'
+      1: '&safe=1'
+      2: '&safe=1'
     results_xpath: //div[contains(@class, "result-item")]
     url_xpath: .//a/@href
     title_xpath: .//div[contains(@class, "title-line")]