retriever: split keyword hit selection

atlasbot/engine/answerer.py

@@ -1082,6 +1082,26 @@ async def _score_chunk_group(
     return scored
 
 
+def _keyword_hits(
+    ranked: list[dict[str, Any]],
+    head: dict[str, Any],
+    keywords: list[str] | None,
+) -> list[dict[str, Any]]:
+    if not keywords:
+        return []
+    lowered = [kw.lower() for kw in keywords if isinstance(kw, str) and kw.strip()]
+    if not lowered:
+        return []
+    hits: list[dict[str, Any]] = []
+    for item in ranked:
+        if item is head:
+            continue
+        text = str(item.get("text") or "").lower()
+        if any(kw in text for kw in lowered):
+            hits.append(item)
+    return hits
+
+
 def _select_chunks(
     chunks: list[dict[str, Any]],
     scores: dict[str, float],
@@ -1095,29 +1115,17 @@ def _select_chunks(
     head = chunks[0]
     selected.append(head)
 
-    keyword_hits: list[dict[str, Any]] = []
-    if keywords:
-        lowered = [kw.lower() for kw in keywords if isinstance(kw, str) and kw.strip()]
-        for item in ranked:
-            if item is head:
-                continue
-            text = str(item.get("text") or "").lower()
-            if any(kw in text for kw in lowered):
-                keyword_hits.append(item)
-
-    for item in keyword_hits:
+    for item in _keyword_hits(ranked, head, keywords):
         if len(selected) >= plan.chunk_top:
             return selected
-        if item in selected:
-            continue
-        selected.append(item)
+        if item not in selected:
+            selected.append(item)
 
     for item in ranked:
         if len(selected) >= plan.chunk_top:
             break
-        if item in selected:
-            continue
-        selected.append(item)
+        if item not in selected:
+            selected.append(item)
     return selected