renames HybridQuery to AggregateHybridQuery

Author: justin-cechmanek

redisvl/query/__init__.py

@@ -1,4 +1,5 @@
 from redisvl.query.aggregate import (
+    AggregateHybridQuery,
     AggregationQuery,
     HybridQuery,
     MultiVectorQuery,
@@ -25,6 +26,7 @@
     "CountQuery",
     "TextQuery",
     "AggregationQuery",
+    "AggregateHybridQuery",
     "HybridQuery",
     "MultiVectorQuery",
     "Vector",

redisvl/query/aggregate.py

@@ -1,3 +1,4 @@
+import warnings
 from typing import Any, Dict, List, Optional, Set, Tuple, Union
 
 from pydantic import BaseModel, field_validator, model_validator
@@ -53,20 +54,20 @@ def __init__(self, query_string):
         super().__init__(query_string)
 
 
-class HybridQuery(AggregationQuery):
+class AggregateHybridQuery(AggregationQuery):
     """
-    HybridQuery combines text and vector search in Redis.
+    AggregateHybridQuery combines text and vector search in Redis.
     It allows you to perform a hybrid search using both text and vector similarity.
     It scores documents based on a weighted combination of text and vector similarity.
 
     .. code-block:: python
 
-        from redisvl.query import HybridQuery
+        from redisvl.query import AggregateHybridQuery
         from redisvl.index import SearchIndex
 
         index = SearchIndex.from_yaml("path/to/index.yaml")
 
-        query = HybridQuery(
+        query = AggregateHybridQuery(
             text="example text",
             text_field_name="text_field",
             vector=[0.1, 0.2, 0.3],
@@ -105,7 +106,7 @@ def __init__(
         text_weights: Optional[Dict[str, float]] = None,
     ):
         """
-        Instantiates a HybridQuery object.
+        Instantiates a AggregateHybridQuery object.
 
         Args:
             text (str): The text to search for.
@@ -313,6 +314,26 @@ def __str__(self) -> str:
         return " ".join([str(x) for x in self.build_args()])
 
 
+class HybridQuery(AggregateHybridQuery):
+    """Backward compatibility wrapper for AggregateHybridQuery.
+
+    .. deprecated::
+        HybridQuery is a backward compatibility wrapper around AggregateHybridQuery
+        and will eventually be replaced with a new hybrid query implementation.
+        to maintain current functionality please use AggregateHybridQuery directly.",
+    """
+
+    def __init__(self, *args, **kwargs):
+        warnings.warn(
+            "HybridQuery is a backward compatibility wrapper around AggregateHybridQuery "
+            "and will eventually be replaced with a new hybrid query implementation. "
+            "to maintain current functionality please use AggregateHybridQuery directly.",
+            DeprecationWarning,
+            stacklevel=2,
+        )
+        super().__init__(*args, **kwargs)
+
+
 class MultiVectorQuery(AggregationQuery):
     """
     MultiVectorQuery allows for search over multiple vector fields in a document simultaneously.

tests/integration/test_aggregation.py

@@ -1,7 +1,7 @@
 import pytest
 
 from redisvl.index import SearchIndex
-from redisvl.query import HybridQuery, MultiVectorQuery, Vector
+from redisvl.query import AggregateHybridQuery, HybridQuery, MultiVectorQuery, Vector
 from redisvl.query.filter import FilterExpression, Geo, GeoRadius, Num, Tag, Text
 from redisvl.redis.utils import array_to_buffer
 from tests.conftest import skip_if_redis_version_below
@@ -89,7 +89,7 @@ def test_hybrid_query(index):
     vector_field = "user_embedding"
     return_fields = ["user", "credit_score", "age", "job", "location", "description"]
 
-    hybrid_query = HybridQuery(
+    hybrid_query = AggregateHybridQuery(
         text=text,
         text_field_name=text_field,
         vector=vector,
@@ -115,7 +115,7 @@ def test_hybrid_query(index):
         assert doc["job"] in ["engineer", "doctor", "dermatologist", "CEO", "dentist"]
         assert doc["credit_score"] in ["high", "low", "medium"]
 
-    hybrid_query = HybridQuery(
+    hybrid_query = AggregateHybridQuery(
         text=text,
         text_field_name=text_field,
         vector=vector,
@@ -141,7 +141,7 @@ def test_empty_query_string():
 
     # test if text is empty
     with pytest.raises(ValueError):
-        hybrid_query = HybridQuery(
+        hybrid_query = AggregateHybridQuery(
             text=text,
             text_field_name=text_field,
             vector=vector,
@@ -151,7 +151,7 @@ def test_empty_query_string():
     # test if text becomes empty after stopwords are removed
     text = "with a for but and"  # will all be removed as default stopwords
     with pytest.raises(ValueError):
-        hybrid_query = HybridQuery(
+        hybrid_query = AggregateHybridQuery(
             text=text,
             text_field_name=text_field,
             vector=vector,
@@ -169,7 +169,7 @@ def test_hybrid_query_with_filter(index):
     return_fields = ["user", "credit_score", "age", "job", "location", "description"]
     filter_expression = (Tag("credit_score") == ("high")) & (Num("age") > 30)
 
-    hybrid_query = HybridQuery(
+    hybrid_query = AggregateHybridQuery(
         text=text,
         text_field_name=text_field,
         vector=vector,
@@ -195,7 +195,7 @@ def test_hybrid_query_with_geo_filter(index):
     return_fields = ["user", "credit_score", "age", "job", "location", "description"]
     filter_expression = Geo("location") == GeoRadius(-122.4194, 37.7749, 1000, "m")
 
-    hybrid_query = HybridQuery(
+    hybrid_query = AggregateHybridQuery(
         text=text,
         text_field_name=text_field,
         vector=vector,
@@ -219,7 +219,7 @@ def test_hybrid_query_alpha(index, alpha):
     vector = [0.1, 0.1, 0.5]
     vector_field = "user_embedding"
 
-    hybrid_query = HybridQuery(
+    hybrid_query = AggregateHybridQuery(
         text=text,
         text_field_name=text_field,
         vector=vector,
@@ -247,7 +247,7 @@ def test_hybrid_query_stopwords(index):
     vector_field = "user_embedding"
     alpha = 0.5
 
-    hybrid_query = HybridQuery(
+    hybrid_query = AggregateHybridQuery(
         text=text,
         text_field_name=text_field,
         vector=vector,
@@ -282,7 +282,7 @@ def test_hybrid_query_with_text_filter(index):
     filter_expression = Text(text_field) == ("medical")
 
     # make sure we can still apply filters to the same text field we are querying
-    hybrid_query = HybridQuery(
+    hybrid_query = AggregateHybridQuery(
         text=text,
         text_field_name=text_field,
         vector=vector,
@@ -300,7 +300,7 @@ def test_hybrid_query_with_text_filter(index):
     filter_expression = (Text(text_field) == ("medical")) & (
         (Text(text_field) != ("research"))
     )
-    hybrid_query = HybridQuery(
+    hybrid_query = AggregateHybridQuery(
         text=text,
         text_field_name=text_field,
         vector=vector,
@@ -330,7 +330,7 @@ def test_hybrid_query_word_weights(index, scorer):
     weights = {"medical": 3.4, "cancers": 5}
 
     # test we can run a query with text weights
-    weighted_query = HybridQuery(
+    weighted_query = AggregateHybridQuery(
         text=text,
         text_field_name=text_field,
         vector=vector,
@@ -344,7 +344,7 @@ def test_hybrid_query_word_weights(index, scorer):
     assert len(weighted_results) == 7
 
     # test that weights do change the scores on results
-    unweighted_query = HybridQuery(
+    unweighted_query = AggregateHybridQuery(
         text=text,
         text_field_name=text_field,
         vector=vector,
@@ -363,7 +363,7 @@ def test_hybrid_query_word_weights(index, scorer):
 
     # test that weights do change the document score and order of results
     weights = {"medical": 5, "cancers": 3.4}  # switch the weights
-    weighted_query = HybridQuery(
+    weighted_query = AggregateHybridQuery(
         text=text,
         text_field_name=text_field,
         vector=vector,
@@ -377,7 +377,7 @@ def test_hybrid_query_word_weights(index, scorer):
     assert weighted_results != unweighted_results
 
     # test assigning weights on construction is equivalent to setting them on the query object
-    new_query = HybridQuery(
+    new_query = AggregateHybridQuery(
         text=text,
         text_field_name=text_field,
         vector=vector,
@@ -743,3 +743,44 @@ def test_multivector_query_mixed_index(index):
         assert (
             float(r["combined_score"]) - score <= 0.0001
         )  # allow for small floating point error
+
+
+def test_hybrid_query_backward_compatibility(index):
+    skip_if_redis_version_below(index.client, "7.2.0")
+
+    text = "a medical professional with expertise in lung cancer"
+    text_field = "description"
+    vector = [0.1, 0.1, 0.5]
+    vector_field = "user_embedding"
+    return_fields = ["user", "credit_score", "age", "job", "location", "description"]
+
+    hybrid_query = AggregateHybridQuery(
+        text=text,
+        text_field_name=text_field,
+        vector=vector,
+        vector_field_name=vector_field,
+        return_fields=return_fields,
+    )
+
+    results = index.query(hybrid_query)
+    assert len(results) == 7
+    for result in results:
+        assert result["user"] in [
+            "john",
+            "derrick",
+            "nancy",
+            "tyler",
+            "tim",
+            "taimur",
+            "joe",
+            "mary",
+        ]
+
+    with pytest.warns(DeprecationWarning):
+        _ = HybridQuery(
+            text=text,
+            text_field_name=text_field,
+            vector=vector,
+            vector_field_name=vector_field,
+            return_fields=return_fields,
+        )