From 1be19f8b5820d1c7b369f80cc48c6014a6d41085 Mon Sep 17 00:00:00 2001
From: Bnyro <bnyro@tutanota.com>
Date: Tue, 4 Nov 2025 18:00:02 +0100
Subject: [feat] sourcehut engine: implement as custom module, fix user agent

SourceHut uses a foss bot protection tool called `go-away` (which I can
recommend BTW).  It blocks common crawler user agents, such as the standard
Firefox user agent.  Hence, we're now using our custom SearXNG user agent to
clarify we're not a crawler.

Closes: https://github.com/searxng/searxng/issues/5270
Co-authored-by: Markus Heiser <markus.heiser@darmarit.de>
---
 searx/engines/sourcehut.py | 90 ++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 90 insertions(+)
 create mode 100644 searx/engines/sourcehut.py

(limited to 'searx/engines')

diff --git a/searx/engines/sourcehut.py b/searx/engines/sourcehut.py
new file mode 100644
index 000000000..fa3df8b23
--- /dev/null
+++ b/searx/engines/sourcehut.py
@@ -0,0 +1,90 @@
+# SPDX-License-Identifier: AGPL-3.0-or-later
+"""Engine to search in the collaborative software platform SourceHut_.
+
+.. _SourceHut: https://sourcehut.org/
+
+Configuration
+=============
+
+You can configure the following setting:
+
+- :py:obj:`sourcehut_sort_order`
+
+.. code:: yaml
+
+  - name: sourcehut
+    shortcut: srht
+    engine: sourcehut
+    # sourcehut_sort_order: longest-active
+
+Implementations
+===============
+
+"""
+
+import typing as t
+
+from urllib.parse import urlencode
+from lxml import html
+
+from searx.utils import eval_xpath, eval_xpath_list, extract_text, searxng_useragent
+from searx.result_types import EngineResults
+
+if t.TYPE_CHECKING:
+    from searx.extended_types import SXNG_Response
+    from searx.search.processors import OnlineParams
+
+
+about = {
+    "website": "https://sourcehut.org",
+    "wikidata_id": "Q78514485",
+    "official_api_documentation": "https://man.sr.ht/",
+    "use_official_api": False,
+    "require_api_key": False,
+    "results": "HTML",
+}
+
+categories = ["it", "repos"]
+paging = True
+
+base_url: str = "https://sr.ht/projects"
+"""Browse public projects."""
+
+
+sourcehut_sort_order: str = "recently-updated"
+"""The sort order of the results.  Possible values:
+
+- ``recently-updated``
+- ``longest-active``
+"""
+
+
+def request(query: str, params: "OnlineParams") -> None:
+
+    args = {"search": query, "page": params["pageno"], "sort": sourcehut_sort_order}
+    params["url"] = f"{base_url}?{urlencode(args)}"
+
+    # standard user agents are blocked by 'go-away', a foss bot detection tool
+    params["headers"]["User-Agent"] = searxng_useragent()
+
+
+def response(resp: "SXNG_Response") -> EngineResults:
+
+    res = EngineResults()
+    doc = html.fromstring(resp.text)
+
+    for item in eval_xpath_list(doc, "(//div[@class='event-list'])[1]/div[contains(@class, 'event')]"):
+        res.add(
+            res.types.LegacyResult(
+                template="packages.html",
+                url=base_url + (extract_text(eval_xpath(item, "./h4/a[2]/@href")) or ""),
+                title=extract_text(eval_xpath(item, "./h4")),
+                package_name=extract_text(eval_xpath(item, "./h4/a[2]")),
+                content=extract_text(eval_xpath(item, "./p")),
+                maintainer=(extract_text(eval_xpath(item, "./h4/a[1]")) or "").removeprefix("~"),
+                tags=[
+                    tag.removeprefix("#") for tag in eval_xpath_list(item, "./div[contains(@class, 'tags')]/a/text()")
+                ],
+            )
+        )
+    return res
-- 
cgit v1.2.3