Merge #531

531: fix the mixed dataset geosearch indexing bug r=Kerollmops a=irevoire

port #529 to main

Co-authored-by: Tamo <tamo@meilisearch.com>

milli/src/update/index_documents/extract/extract_geo_points.rs

@@ -35,26 +35,28 @@ pub fn extract_geo_points<R: io::Read + io::Seek>(
         };
 
         // first we get the two fields
-        let lat = obkv
+        let lat = obkv.get(lat_fid);
-            .get(lat_fid)
+        let lng = obkv.get(lng_fid);
-            .ok_or_else(|| GeoError::MissingLatitude { document_id: primary_key() })?;
-        let lng = obkv
-            .get(lng_fid)
-            .ok_or_else(|| GeoError::MissingLongitude { document_id: primary_key() })?;
 
-        // then we extract the values
+        if let Some((lat, lng)) = lat.zip(lng) {
-        let lat = extract_float_from_value(
+            // then we extract the values
-            serde_json::from_slice(lat).map_err(InternalError::SerdeJson)?,
+            let lat = extract_float_from_value(
-        )
+                serde_json::from_slice(lat).map_err(InternalError::SerdeJson)?,
-        .map_err(|lat| GeoError::BadLatitude { document_id: primary_key(), value: lat })?;
+            )
+            .map_err(|lat| GeoError::BadLatitude { document_id: primary_key(), value: lat })?;
 
-        let lng = extract_float_from_value(
+            let lng = extract_float_from_value(
-            serde_json::from_slice(lng).map_err(InternalError::SerdeJson)?,
+                serde_json::from_slice(lng).map_err(InternalError::SerdeJson)?,
-        )
+            )
-        .map_err(|lng| GeoError::BadLongitude { document_id: primary_key(), value: lng })?;
+            .map_err(|lng| GeoError::BadLongitude { document_id: primary_key(), value: lng })?;
 
-        let bytes: [u8; 16] = concat_arrays![lat.to_ne_bytes(), lng.to_ne_bytes()];
+            let bytes: [u8; 16] = concat_arrays![lat.to_ne_bytes(), lng.to_ne_bytes()];
-        writer.insert(docid_bytes, bytes)?;
+            writer.insert(docid_bytes, bytes)?;
+        } else if lat.is_none() && lng.is_some() {
+            return Err(GeoError::MissingLatitude { document_id: primary_key() })?;
+        } else if lat.is_some() && lng.is_none() {
+            return Err(GeoError::MissingLongitude { document_id: primary_key() })?;
+        }
     }
 
     Ok(writer_into_reader(writer)?)

milli/src/update/index_documents/mod.rs

@@ -1018,6 +1018,43 @@ mod tests {
         wtxn.commit().unwrap();
     }
 
+    #[test]
+    fn mixed_geo_documents() {
+        let path = tempfile::tempdir().unwrap();
+        let mut options = EnvOpenOptions::new();
+        options.map_size(10 * 1024 * 1024); // 10 MB
+        let index = Index::new(options, &path).unwrap();
+
+        // We send 6 documents and mix the ones that have _geo and those that don't have it.
+        let mut wtxn = index.write_txn().unwrap();
+        let documents = documents!([
+          { "id": 2, "price": 3.5, "_geo": { "lat": 12, "lng": 42 } },
+          { "id": 456 },
+          { "id": 1 },
+          { "id": 1344 },
+          { "id": 4 },
+          { "id": 42, "_geo": { "lat": 35, "lng": 23 } }
+        ]);
+        let config = IndexerConfig::default();
+        let indexing_config = IndexDocumentsConfig {
+            update_method: IndexDocumentsMethod::ReplaceDocuments,
+            ..Default::default()
+        };
+        let mut builder =
+            IndexDocuments::new(&mut wtxn, &index, &config, indexing_config, |_| ()).unwrap();
+        builder.add_documents(documents).unwrap();
+        builder.execute().unwrap();
+        wtxn.commit().unwrap();
+
+        let mut wtxn = index.write_txn().unwrap();
+        let mut builder = update::Settings::new(&mut wtxn, &index, &config);
+
+        let faceted_fields = hashset!(S("_geo"));
+        builder.set_filterable_fields(faceted_fields);
+        builder.execute(|_| ()).unwrap();
+        wtxn.commit().unwrap();
+    }
+
     #[test]
     fn index_all_flavour_of_geo() {
         let path = tempfile::tempdir().unwrap();