From 6ca9d3784c576dee44edc510a945e7889399766f Mon Sep 17 00:00:00 2001
From: Bnyro <bnyro@tutanota.com>
Date: Mon, 8 Jun 2026 07:09:06 +0200
Subject: [PATCH] [feat] engines: add seek-ninja general engine (#6217)

Add support for https://seek.ninja (general)

It's very slow because the engine uses Server-side events, that incrementally
send data in their HTTP response [1].

I.e. we wait for the end of the response (7+ seconds), even though the results
data arrives within a few seconds -> it's very slow, because SearXNG wants to
get the full response body before it calls the `response(resp)` method

We could use httpx-sse [2], but I'm not sure how to integrate this into SearXNG
and if it's worth it

[1] https://developer.mozilla.org/en-US/docs/Web/API/Server-sent_events/
[2] https://github.com/florimondmanca/httpx-sse
---
 searx/engines/seekninja.py | 113 +++++++++++++++++++++++++++++++++++++
 searx/settings.yml         |   8 +++
 2 files changed, 121 insertions(+)
 create mode 100644 searx/engines/seekninja.py

diff --git a/searx/engines/seekninja.py b/searx/engines/seekninja.py
new file mode 100644
index 000000000..882274712
--- /dev/null
+++ b/searx/engines/seekninja.py
@@ -0,0 +1,113 @@
+# SPDX-License-Identifier: AGPL-3.0-or-later
+"""Seek ninja (general)"""
+
+from json import loads
+from hashlib import sha256
+from urllib.parse import urlencode, quote_plus
+
+import typing as t
+
+from searx.extended_types import SXNG_Response
+from searx.network import get
+from searx.result_types import EngineResults
+from searx.utils import extr, html_to_text
+
+if t.TYPE_CHECKING:
+    from searx.search.processors import OnlineParams
+
+about = {
+    "website": "https://seek.ninja",
+    "official_api_documentation": None,
+    "use_official_api": False,
+    "require_api_key": False,
+    "results": "JSON",
+}
+
+safesearch = True
+
+base_url = "https://seek.ninja"
+categories = ["general"]
+
+safe_search_map = {0: "off", 1: "moderate", 2: "strict"}
+
+PowChallenge = dict[str, t.Any]
+
+
+def _get_challenge(query: str) -> PowChallenge:
+    """Extract the challenge parameters (i.e. nonce, difficulty, ...) from the
+    search website."""
+
+    resp = get(f"{base_url}/s?q={quote_plus(query)}")
+    challenge_raw_json = "{" + extr(resp.text, "pow: {", "},") + "}"
+    return loads(challenge_raw_json)
+
+
+def _solve_pow(challenge: PowChallenge) -> list[int]:
+    """Solves a Proof of Work SHA256 challenges. This is a 1:1 port of the
+    site's JS code.
+
+    On a high-level, it tries to ``k`` amount of solutions, where its sha256
+    hash begins with: ``leading`` 0s, i.e.
+
+    .. code: js
+
+       sha256(nonce || solution).startswith("0" * leading)
+    """
+    nonce = challenge["nonce"]
+    k = int(challenge["k"])
+    indifficulty = float(challenge["indifficulty"])
+
+    leading = int(indifficulty)
+    frac = indifficulty - leading
+    prefix = "".join("0" for _ in range(0, leading))
+
+    maxNib = 15 - int(frac * 16) if frac else 15
+
+    solutions: list[int] = []
+    ans = 0
+    while len(solutions) < k:
+        h = sha256(f"{nonce}{ans}".encode()).hexdigest()
+        if h.startswith(prefix) and (not frac or int(h[leading], base=16) <= maxNib):
+            solutions.append(ans)
+        ans += 1
+    return solutions
+
+
+def request(query: str, params: 'OnlineParams') -> None:
+    challenge = _get_challenge(query)
+    solution = _solve_pow(challenge)
+    args = {
+        "q": query,
+        "panswers": ",".join(str(s) for s in solution),
+        "pid": challenge["challengeId"],
+        "adult": safe_search_map[params["safesearch"]],
+    }
+    params["url"] = f"{base_url}/search-sse?{urlencode(args)}"
+
+
+def response(resp: 'SXNG_Response') -> EngineResults:
+    res = EngineResults()
+    # The response is a stream of server-side events,
+    # so it is split into `event: <type>` and `data: {"results": ...}`
+    # see https://developer.mozilla.org/en-US/docs/Web/API/Server-sent_events/
+    events = resp.text.split("\n\n")
+    for event in events:
+        event_parts = event.split("\n", maxsplit=2)
+        if len(event_parts) != 2:
+            continue
+
+        event_name, data = event_parts
+        if not event_name.endswith("resultsUpdate"):
+            continue
+
+        json_data = loads(data.removeprefix("data: "))
+        for result in json_data["results"]:
+            res.add(
+                res.types.MainResult(
+                    url=result["url"],
+                    title=result["title"],
+                    content=html_to_text(result["blurb"]),
+                )
+            )
+
+    return res
diff --git a/searx/settings.yml b/searx/settings.yml
index da5d73a37..db0aebc19 100644
--- a/searx/settings.yml
+++ b/searx/settings.yml
@@ -1976,6 +1976,14 @@ engines:
   #       - ...
   #   disabled: true
 
+  - name: seekninja
+    engine: seekninja
+    shortcut: sen
+    # very slow due to its server-side events architecture
+    timeout: 10
+    disabled: true
+    inactive: true
+
   - name: semantic scholar
     engine: semantic_scholar
     shortcut: se