fix: Fix MetaFieldGroupingRanker to handle unhashable subgroup_by values like list (#9791)

* Fixes

* Add reno

---------

Co-authored-by: David S. Batista <dsbatista@gmail.com>

Author: sjrl

haystack/components/rankers/meta_field_grouping_ranker.py

@@ -94,14 +94,17 @@ def run(self, documents: list[Document]) -> dict[str, Any]:
         for doc in documents:
             group_value = str(doc.meta.get(self.group_by, ""))
 
-            if group_value:
-                subgroup_value = "no_subgroup"
-                if self.subgroup_by and self.subgroup_by in doc.meta:
-                    subgroup_value = doc.meta[self.subgroup_by]
-
-                document_groups[group_value][subgroup_value].append(doc)
-            else:
+            # If no group value, add to no_group_docs and continue
+            if not group_value:
                 no_group_docs.append(doc)
+                continue
+
+            # Get subgroup value or use a default if not specified
+            subgroup_value = "no_subgroup"
+            if self.subgroup_by and self.subgroup_by in doc.meta:
+                subgroup_value = str(doc.meta[self.subgroup_by])
+
+            document_groups[group_value][subgroup_value].append(doc)
 
         ordered_docs = []
         for subgroups in document_groups.values():

releasenotes/notes/fix-metafieldgroupingranker-bc33f5c886067e00.yaml

@@ -0,0 +1,4 @@
+---
+fixes:
+  - |
+    Fix `MetaFieldGroupingRanker` to still work when `subgroup_by` values are unhashable types like list. We handle this by stringfying the contents of `doc.meta[subgroup_by]` in the same we do this for values of `doc.meta[group_by]`.

test/components/rankers/test_meta_field_grouping_ranker.py

@@ -97,7 +97,7 @@ def test_run_with_lists(self) -> None:
         """
         Test if the MetaFieldGroupingRanker component can handle list values in the metadata.
         """
-        ranker = MetaFieldGroupingRanker(group_by="value_list", subgroup_by="subvaluelist", sort_docs_by="split_id")
+        ranker = MetaFieldGroupingRanker(group_by="value_list", subgroup_by="sub_value_list", sort_docs_by="split_id")
         result = ranker.run(documents=DOC_LIST)
         assert "documents" in result
         assert len(DOC_LIST) == len(result["documents"])