fix(ai): chaine de repli web_search (SearXNG -> DuckDuckGo -> Bing, BUG-051)

tests/test_web_tools.py

@@ -33,6 +33,40 @@ def _ctx() -> ToolContext:
     return ToolContext(user={"username": "tester", "vaults": []}, mode=ToolMode.IN_APP)
 
 
+def _html_get(routes: dict[str, str], searxng: Any = None):
+    """Dispatch httpx.get by URL marker: searxng JSON, HTML providers, else error."""
+
+    def fake_get(url, params=None, **kw):
+        url = str(url)
+        if "search.dracodev.net" in url:
+            if searxng is not None:
+                return searxng
+            raise web.httpx.ConnectError("searxng down")
+        for marker, body in routes.items():
+            if marker in url:
+                return FakeResponse(content=body.encode("utf-8"), url=url)
+        raise web.httpx.ConnectError(f"no route: {url}")
+
+    return fake_get
+
+
+DDG_HTML = (
+    '<div class="result">'
+    '<a rel="nofollow" class="result__a" '
+    'href="//duckduckgo.com/l/?uddg=https%3A%2F%2Fexample.com%2Fpage&amp;rut=1">'
+    "Example <b>Page</b></a>"
+    '<a class="result__snippet" href="#">A useful snippet</a>'
+    "</div>"
+)
+
+BING_HTML = (
+    '<h2 class=""><a target="_blank" '
+    'href="https://www.bing.com/ck/a?u=a1aHR0cHM6Ly9leGFtcGxlLmNvbS9iaW5n&amp;ntb=1">'
+    "Bing <strong>Result</strong></a></h2>"
+    '<p class="b_lineclamp2">Bing snippet here</p>'
+)
+
+
 class TestRegistration:
     def test_tools_registered_read_only_in_app(self):
         for name in ("web_search", "fetch_url"):
@@ -58,7 +92,7 @@ class TestWebSearch:
 
         monkeypatch.setattr(web.httpx, "get", fake_get)
         out = web.web_search(_ctx(), web.WebSearchInput(query="pizza", max_results=3))
-        assert out["engine"] == "searxng"
+        assert out["provider"] == "searxng"
         assert out["count"] == 3
         assert len(out["results"][0]["snippet"]) <= 600
         assert captured["params"]["q"] == "pizza"
@@ -70,11 +104,11 @@ class TestWebSearch:
     def test_empty_result_set_warns_about_blocked_engines(self, monkeypatch):
         """An all-blocked instance answers 200 with no results: the model must
         be told instead of retrying the same search until the quota burns."""
-        monkeypatch.setattr(web.httpx, "get", lambda *a, **kw: FakeResponse(json_data={
+        monkeypatch.setattr(web.httpx, "get", _html_get({}, searxng=FakeResponse(json_data={
             "results": [],
             "number_of_results": 0,
             "unresponsive_engines": [["duckduckgo", "CAPTCHA"], ["google", "access denied"]],
-        }))
+        })))
         out = web.web_search(_ctx(), web.WebSearchInput(query="meteo montreal"))
         assert out["count"] == 0
         assert out["unresponsive_engines"] == ["duckduckgo", "google"]
@@ -100,6 +134,41 @@ class TestWebSearch:
             web.web_search(_ctx(), web.WebSearchInput(query="x"))
         assert ei.value.code == "web_search_unavailable"
 
+    def test_falls_back_to_duckduckgo_when_searxng_empty(self, monkeypatch):
+        monkeypatch.setattr(web.httpx, "get", _html_get(
+            {"html.duckduckgo.com": DDG_HTML},
+            searxng=FakeResponse(json_data={"results": []}),
+        ))
+        out = web.web_search(_ctx(), web.WebSearchInput(query="python release"))
+        assert out["provider"] == "duckduckgo"
+        assert out["count"] == 1
+        assert out["results"][0]["title"] == "Example Page"
+        assert out["results"][0]["url"] == "https://example.com/page"
+        assert out["results"][0]["snippet"] == "A useful snippet"
+
+    def test_falls_back_to_bing_when_searxng_unreachable(self, monkeypatch):
+        monkeypatch.setattr(web.httpx, "get", _html_get(
+            {"bing.com": BING_HTML},
+            searxng=None,
+        ))
+        out = web.web_search(_ctx(), web.WebSearchInput(query="python release"))
+        assert out["provider"] == "bing"
+        assert out["count"] == 1
+        assert out["results"][0]["title"] == "Bing Result"
+        assert out["results"][0]["url"] == "https://example.com/bing"
+        assert out["results"][0]["snippet"] == "Bing snippet here"
+
+    def test_fallback_can_be_disabled(self, monkeypatch):
+        monkeypatch.setattr(web, "WEB_FALLBACK_ENABLED", False)
+        monkeypatch.setattr(web.httpx, "get", _html_get(
+            {"html.duckduckgo.com": DDG_HTML, "bing.com": BING_HTML},
+            searxng=FakeResponse(json_data={"results": []}),
+        ))
+        out = web.web_search(_ctx(), web.WebSearchInput(query="python release"))
+        assert out["count"] == 0
+        assert out["provider"] == "searxng"
+        assert "warning" in out
+
 
 class TestFetchUrl:
     def test_html_converted_to_text(self, monkeypatch):