Fix regex quoting for DuckDuckGo search

app.py

@@ -148,15 +148,14 @@ API_RESOURCES = {
     },
     "web_search": {
         "name": "Web search",
-        "description": "Search the web using a self-hosted SearXNG metasearch engine.",
+        "description": "Search the web via DuckDuckGo. No API key required.",
         "endpoint": "/respite/search",
         "method": "GET",
         "input": {
             "q": "string (search query)",
-            "categories": "string (general, images, news, etc.)",
             "max_results": "integer (1-50, default 10)",
         },
-        "output": "JSON with title, url, content, engine, score per result",
+        "output": "JSON with title, url, content per result",
     }
 }
 
@@ -206,67 +205,46 @@ def _get_runtime_specs():
 
 
 # ---------------------------------------------------------------------------
-# SearXNG search (uses public instances with JSON API)
+# Web search (DuckDuckGo HTML scraping, no API key required)
 # ---------------------------------------------------------------------------
 
-SEARXNG_INSTANCES = [
-    "https://search.ononoki.org",
-    "https://searx.be",
-    "https://search.sapti.me",
-    "https://searxng.site",
-]
-_searxng_active_instance = None
+import re as _re
 
 
-def _probe_searxng_instances():
-    """Find a working SearXNG instance."""
-    global _searxng_active_instance
-    for instance in SEARXNG_INSTANCES:
-        try:
-            url = f"{instance}/search?q=test&format=json"
-            req = urllib.request.Request(url, headers={"User-Agent": "RespiteAPI/1.0"})
-            resp = urllib.request.urlopen(req, timeout=5)
-            _json.loads(resp.read())
-            _searxng_active_instance = instance
-            _log(f"SearXNG instance ready: {instance}")
-            return
-        except Exception:
-            continue
-    _log("Warning: No working SearXNG instance found")
-
-
-# Probe instances in background so it doesn't block startup
-threading.Thread(target=_probe_searxng_instances, daemon=True).start()
-
-
-def _search_web(query, categories="general", max_results=10):
-    """Query a public SearXNG instance."""
-    if not _searxng_active_instance:
-        raise RuntimeError("No SearXNG instance available")
-    params = {
-        "q": query,
-        "format": "json",
-        "categories": categories,
-        "language": "en",
-    }
-    url = f"{_searxng_active_instance}/search?{urllib.parse.urlencode(params)}"
-    req = urllib.request.Request(url, headers={"User-Agent": "RespiteAPI/1.0"})
+def _search_web(query, max_results=10):
+    """Search the web via DuckDuckGo HTML endpoint."""
+    url = f"https://html.duckduckgo.com/html/?{urllib.parse.urlencode({'q': query})}"
+    req = urllib.request.Request(url, headers={
+        "User-Agent": "Mozilla/5.0 (compatible; RespiteAPI/1.0)",
+    })
     resp = urllib.request.urlopen(req, timeout=10)
-    data = _json.loads(resp.read())
+    html = resp.read().decode("utf-8", errors="replace")
+
     results = []
-    for r in data.get("results", [])[:max_results]:
+    # Extract search results from DuckDuckGo HTML
+    for match in _re.finditer(
+        r'<a[^>]+class="result__a"[^>]*href="([^"]+)"[^>]*>(.+?)</a>.*?'
+        r'<a[^>]+class="result__snippet"[^>]*>(.*?)</a>',
+        html, _re.DOTALL,
+    ):
+        if len(results) >= max_results:
+            break
+        raw_url = match.group(1)
+        title = _re.sub(r'<[^>]+>', '', match.group(2)).strip()
+        snippet = _re.sub(r'<[^>]+>', '', match.group(3)).strip()
+        # DuckDuckGo wraps URLs in a redirect; extract the real URL
+        ud_match = _re.search(r'uddg=([^"&]+)', raw_url)
+        real_url = urllib.parse.unquote(ud_match.group(1)) if ud_match else raw_url
         results.append({
-            "title": r.get("title", ""),
-            "url": r.get("url", ""),
-            "content": r.get("content", ""),
-            "engine": r.get("engine", ""),
-            "score": r.get("score", 0),
+            "title": title,
+            "url": real_url,
+            "content": snippet,
         })
+
     return {
         "query": query,
-        "instance": _searxng_active_instance,
         "results": results,
-        "total_results": len(data.get("results", [])),
+        "total_results": len(results),
     }