search criteria

nuse_modules/google_search.py

@@ -6,14 +6,39 @@ import requests
 GOOGLE_API_KEY = os.getenv("GOOGLE_API_KEY")
 GOOGLE_CX_ID = os.getenv("GOOGLE_CX_ID")
 
+# Map of trusted domains per region
+TRUSTED_SOURCES_BY_REGION = {
+    "global": ["bbc.com", "reuters.com", "apnews.com", "nytimes.com"],
+    "india": ["thehindu.com", "indiatoday.in", "timesofindia.indiatimes.com", "scroll.in"],
+    "china": ["scmp.com", "chinadaily.com.cn"],
+    "middle_east": ["aljazeera.com", "arabnews.com"]
+}
+
+def infer_region_from_keywords(keywords: list[str]) -> str:
+    for kw in keywords:
+        k = kw.lower()
+        if k in {"india", "modi", "delhi", "supreme court"}:
+            return "india"
+        elif k in {"china", "beijing", "xi jinping"}:
+            return "china"
+        elif k in {"gaza", "israel", "palestine", "jerusalem"}:
+            return "middle_east"
+    return "global"
+
+def is_trusted_domain(url: str, region: str) -> bool:
+    trusted = TRUSTED_SOURCES_BY_REGION.get(region, TRUSTED_SOURCES_BY_REGION["global"])
+    return any(domain in url for domain in trusted)
+
 def search_google_news(keywords: list[str], num_results: int = 5):
     query = " ".join(keywords)
+    region = infer_region_from_keywords(keywords)
+
     url = (
         f"https://www.googleapis.com/customsearch/v1"
         f"?key={GOOGLE_API_KEY}&cx={GOOGLE_CX_ID}"
         f"&q={query}&num={num_results}"
     )
-    
+
     try:
         res = requests.get(url, timeout=10)
         res.raise_for_status()
@@ -21,11 +46,13 @@ def search_google_news(keywords: list[str], num_results: int = 5):
         results = []
 
         for item in data.get("items", []):
-            results.append({
-                "title": item.get("title"),
-                "link": item.get("link"),
-                "snippet": item.get("snippet"),
-            })
+            link = item.get("link")
+            if is_trusted_domain(link, region):
+                results.append({
+                    "title": item.get("title"),
+                    "link": link,
+                    "snippet": item.get("snippet"),
+                })
 
         return results
     except Exception as e: