[feat] engines: add support for iseek.com (general)

2026-06-22 17:48:33 +02:00 · 2026-06-11 11:00:08 +02:00
parent 9375c0a6b6
commit cce0957f54
2 changed files with 95 additions and 0 deletions
@@ -0,0 +1,88 @@
+# SPDX-License-Identifier: AGPL-3.0-or-later
+"""iseek_ is a search engine by the AI company Vantage Labs LLC,
+that focuses on medical and educational applicances.
+Although it's an AI company, it doesn't include any AI stuff in its results.
+
+.. _iseek : https://www.iseek.ai/
+"""
+
+import base64
+from hashlib import sha256
+import typing as t
+from urllib.parse import urlencode
+
+from searx.result_types import EngineResults
+
+if t.TYPE_CHECKING:
+    from searx.search.processors import OnlineParams
+    from searx.extended_types import SXNG_Response
+
+
+about = {
+    "website": 'https://www.iseek.com',
+    "wikidata_id": None,
+    "official_api_documentation": None,
+    "use_official_api": False,
+    "require_api_key": False,
+    "results": "JSON",
+}
+categories = ["general"]
+paging = True
+
+base_url = "https://api.iseek.com"
+page_size = 10
+
+
+def _get_new_token(query: str, pageno: int) -> str:
+    """Create a new ``qToken``. This reduced the time for fetching subsequent pages
+    from 4 seconds to 200ms when testing."""
+    # The website uses a random value as qToken for the first page. For our use case,
+    # it's easier if the qToken can be deterministically re-calculated based on the search query,
+    # so that we can the same result when calling _get_new_token for the second, third, ... page
+    #
+    # var qToken = Math.ceil(Math.random() * parseInt("ZZZZ", 36)).toString(36);
+    # while (qToken.length < 4) qToken = '0' + qToken;
+    # qToken = qToken + "_" + pageno
+    query_hash = sha256(query.encode()).digest()
+    hash_start = base64.b64encode(query_hash).decode()[0:4]
+    return f"{hash_start}_{pageno}"
+
+
+def request(query: str, params: "OnlineParams"):
+    offset = (params["pageno"] - 1) * page_size
+
+    # always seems to find 20 results max
+    if offset >= 20:
+        params["url"] = None
+        return
+
+    args = {
+        "q": query,
+        "key": "core-web",
+        "num": str(page_size),
+        "off": offset,
+        "rSort": "__metasearch_score_d:desc",
+        # it supports many more fields, but none of them are really relevant
+        "names": "title_t,content_txt,url_s",
+        "qNames": "title_t",
+        "qToken": _get_new_token(query, params["pageno"]),
+    }
+    params["url"] = f"{base_url}/search?{urlencode(args)}"
+
+
+def response(resp: "SXNG_Response") -> EngineResults:
+    res = EngineResults()
+
+    for group in resp.json()["data"]:
+        group: dict[str, t.Any]
+        for result in group["doclist"]["docs"]:
+            result: dict[str, str]
+            res.add(
+                res.types.MainResult(
+                    url=result["url_s"],
+                    title=result["title_t"],
+                    content="".join(result["content_txt"]),
+                )
+            )
+
+    return res