Debugging and adding a failing test.

src/cpr_sdk/models/search.py

@@ -397,6 +397,8 @@ def from_vespa_response(cls, response_hit: dict) -> "Passage":
             document_content_type=fields.get("document_content_type"),
             document_cdn_object=fields.get("document_cdn_object"),
             document_source_url=fields.get("document_source_url"),
+            corpus_type_name=fields.get("corpus_type_name"),
+            corpus_import_id=fields.get("corpus_import_id"),
             text_block=fields["text_block"],
             text_block_id=fields["text_block_id"],
             text_block_type=fields["text_block_type"],

tests/local_vespa/test_app/schemas/document_passage.sd

@@ -54,13 +54,16 @@ schema document_passage {
     import field family_document_ref.family_publication_ts as family_publication_ts {}
     import field family_document_ref.family_publication_year as family_publication_year {}
     import field family_document_ref.family_geography as family_geography {}
+    import field family_document_ref.family_geographies as family_geographies {}
     import field family_document_ref.family_source as family_source {}
     import field family_document_ref.document_import_id as document_import_id {}
     import field family_document_ref.document_slug as document_slug {}
     import field family_document_ref.document_languages as document_languages {}
     import field family_document_ref.document_content_type as document_content_type {}
     import field family_document_ref.document_cdn_object as document_cdn_object {}
     import field family_document_ref.document_source_url as document_source_url {}
+    import field family_document_ref.corpus_import_id as corpus_import_id {}
+    import field family_document_ref.corpus_type_name as corpus_type_name {}
     import field search_weights_ref.passage_weight as passage_weight {}
 
     fieldset default {
@@ -75,13 +78,16 @@ schema document_passage {
         summary family_category {}
         summary family_publication_ts {}
         summary family_geography {}
+        summary family_geographies {}
         summary family_source {}
         summary document_import_id {}
         summary document_slug {}
         summary document_languages {}
         summary document_content_type {}
         summary document_cdn_object {}
         summary document_source_url {}
+        summary corpus_import_id {}
+        summary corpus_type_name {}
         summary text_block {}
         summary text_block_id {}
         summary text_block_type {}

tests/local_vespa/test_documents/family_document.json

@@ -309,7 +309,7 @@
       "family_slug": "environmental-strategy-for-2014-2023_9f8e",
       "document_source_url": "https://wedocs.unep.org/bitstream/handle/20.500.11822/9507/-Environmental_Strategy_for_the_years_2014-2023-2014Moldova_EnvironmentalStrategy_2014-202.pdf?sequence=3&isAllowed=y",
       "family_geography": "MDA",
-      "family_geographies": null,
+      "family_geographies": ["MDA"],
       "family_category": "Executive",
       "document_md5_sum": "bea7a05dae73fbbd629e687a71a15b95",
       "family_name_index": "Environmental Strategy for 2014-2023",

tests/test_search_adaptors.py

@@ -7,13 +7,16 @@
 from cpr_sdk.models.search import (
     Document,
     Filters,
+    Hit,
     MetadataFilter,
     Passage,
     SearchParameters,
     SearchResponse,
     sort_fields,
 )
 from cpr_sdk.search_adaptors import VespaSearchAdapter
+from cpr_sdk.utils import dig
+from cpr_sdk.vespa import build_vespa_request_body
 
 
 def vespa_search(
@@ -455,6 +458,7 @@ def test_vespa_search_adaptor__corpus_type_name(
     response = vespa_search(test_vespa, request)
     assert response.total_family_hits > 0
     for family in response.families:
+        assert len(family.hits) > 0
         for hit in family.hits:
             assert hit.corpus_type_name not in [None, []]
             assert hit.corpus_type_name in corpus_type_names
@@ -553,3 +557,39 @@ def test_vespa_search_adaptor__filters(test_vespa, query_string, filters):
                 attribute_value_from_hit = getattr(hit, filter_name)
                 assert attribute_value_from_hit not in [None, []]
                 assert all([val in attribute_value_from_hit for val in filter_values])
+
+
+@pytest.mark.vespa
+@pytest.mark.parametrize("query_string", ["e"])
+@pytest.mark.parametrize("exact_match", [True, False])
+@pytest.mark.parametrize(
+    "metadata_filters", [None, [{"name": "family.sector", "value": "Price"}]]
+)
+@pytest.mark.parametrize("geographies", [None, {"family_geographies": ["BIH"]}])
+def test_vespa_search_response__geographies(
+    test_vespa, query_string, exact_match, metadata_filters, geographies
+) -> None:
+    """Test that the search response includes geographies"""
+    parameters = SearchParameters(
+        query_string=query_string,
+        exact_match=exact_match,
+        filters=Filters.model_validate(geographies) if geographies else None,
+        metadata=(
+            [
+                MetadataFilter.model_validate(metadata_filter)
+                for metadata_filter in metadata_filters
+            ]
+            if metadata_filters
+            else None
+        ),
+    )
+
+    vespa_response = test_vespa.client.query(body=build_vespa_request_body(parameters))
+
+    root = vespa_response.json["root"]
+    response_families = dig(root, "children", 0, "children", 0, "children", default=[])
+    for family in response_families:
+        for hit in dig(family, "children", 0, "children", default=[]):
+            hit = Hit.from_vespa_response(response_hit=hit)
+            assert hit.family_geography not in [None, []]
+            assert hit.family_geographies not in [None, []]