Merge branch '0.6.0' into feat/rerun-060-notebooks

justin-cechmanek · justin-cechmanek · commit fde29831c636 · 2025-04-30T19:24:10.000-06:00
diff --git a/redisvl/cli/index.py b/redisvl/cli/index.py
@@ -8,6 +8,7 @@
 from redisvl.redis.utils import convert_bytes, make_dict
 from redisvl.schema.schema import IndexSchema
 from redisvl.utils.log import get_logger
+from redisvl.utils.utils import lazy_import
 
 logger = get_logger("[RedisVL]")
 
diff --git a/redisvl/cli/stats.py b/redisvl/cli/stats.py
@@ -6,6 +6,7 @@
 from redisvl.index import SearchIndex
 from redisvl.schema.schema import IndexSchema
 from redisvl.utils.log import get_logger
+from redisvl.utils.utils import lazy_import
 
 logger = get_logger("[RedisVL]")
 
@@ -85,6 +86,8 @@ def _connect_to_index(self, args: Namespace) -> SearchIndex:
 
 
 def _display_stats(index_info, output_format="rounded_outline"):
+    tabulate = lazy_import("tabulate")
+
     # Extracting the statistics
     stats_data = [(key, str(index_info.get(key))) for key in STATS_KEYS]
 
diff --git a/redisvl/query/aggregate.py b/redisvl/query/aggregate.py
@@ -5,6 +5,10 @@
 from redisvl.query.filter import FilterExpression
 from redisvl.redis.utils import array_to_buffer
 from redisvl.utils.token_escaper import TokenEscaper
+from redisvl.utils.utils import lazy_import
+
+nltk = lazy_import("nltk")
+nltk_stopwords = lazy_import("nltk.corpus.stopwords")
 
 
 class AggregationQuery(AggregateRequest):
@@ -162,17 +166,13 @@ def _set_stopwords(self, stopwords: Optional[Union[str, Set[str]]] = "english"):
         if not stopwords:
             self._stopwords = set()
         elif isinstance(stopwords, str):
-            # Lazy import because nltk is an optional dependency
             try:
-                import nltk
-                from nltk.corpus import stopwords as nltk_stopwords
+                nltk.download("stopwords", quiet=True)
+                self._stopwords = set(nltk_stopwords.words(stopwords))
             except ImportError:
                 raise ValueError(
                     f"Loading stopwords for {stopwords} failed: nltk is not installed."
                 )
-            try:
-                nltk.download("stopwords", quiet=True)
-                self._stopwords = set(nltk_stopwords.words(stopwords))
             except Exception as e:
                 raise ValueError(f"Error trying to load {stopwords} from nltk. {e}")
         elif isinstance(stopwords, (Set, List, Tuple)) and all(  # type: ignore
diff --git a/redisvl/query/query.py b/redisvl/query/query.py
@@ -6,7 +6,10 @@
 from redisvl.query.filter import FilterExpression
 from redisvl.redis.utils import array_to_buffer
 from redisvl.utils.token_escaper import TokenEscaper
-from redisvl.utils.utils import denorm_cosine_distance
+from redisvl.utils.utils import denorm_cosine_distance, lazy_import
+
+nltk = lazy_import("nltk")
+nltk_stopwords = lazy_import("nltk.corpus.stopwords")
 
 
 class BaseQuery(RedisQuery):
@@ -893,17 +896,13 @@ def _set_stopwords(self, stopwords: Optional[Union[str, Set[str]]] = "english"):
         if not stopwords:
             self._stopwords = set()
         elif isinstance(stopwords, str):
-            # Lazy import because nltk is an optional dependency
             try:
-                import nltk
-                from nltk.corpus import stopwords as nltk_stopwords
+                nltk.download("stopwords", quiet=True)
+                self._stopwords = set(nltk_stopwords.words(stopwords))
             except ImportError:
                 raise ValueError(
                     f"Loading stopwords for {stopwords} failed: nltk is not installed."
                 )
-            try:
-                nltk.download("stopwords", quiet=True)
-                self._stopwords = set(nltk_stopwords.words(stopwords))
             except Exception as e:
                 raise ValueError(f"Error trying to load {stopwords} from nltk. {e}")
         elif isinstance(stopwords, (Set, List, Tuple)) and all(  # type: ignore
diff --git a/redisvl/redis/utils.py b/redisvl/redis/utils.py
@@ -1,8 +1,10 @@
 import hashlib
 from typing import Any, Dict, List, Optional
 
-import numpy as np
-from ml_dtypes import bfloat16
+from redisvl.utils.utils import lazy_import
+
+# Lazy import numpy
+np = lazy_import("numpy")
 
 from redisvl.schema.fields import VectorDataType
 
@@ -41,6 +43,13 @@ def array_to_buffer(array: List[float], dtype: str) -> bytes:
         raise ValueError(
             f"Invalid data type: {dtype}. Supported types are: {[t.lower() for t in VectorDataType]}"
         )
+
+    # Special handling for bfloat16 which requires explicit import from ml_dtypes
+    if dtype.lower() == "bfloat16":
+        from ml_dtypes import bfloat16
+
+        return np.array(array, dtype=bfloat16).tobytes()
+
     return np.array(array, dtype=dtype.lower()).tobytes()
 
 
@@ -52,6 +61,14 @@ def buffer_to_array(buffer: bytes, dtype: str) -> List[Any]:
         raise ValueError(
             f"Invalid data type: {dtype}. Supported types are: {[t.lower() for t in VectorDataType]}"
         )
+
+    # Special handling for bfloat16 which requires explicit import from ml_dtypes
+    # because otherwise the (lazily imported) numpy is unaware of the type
+    if dtype.lower() == "bfloat16":
+        from ml_dtypes import bfloat16
+
+        return np.frombuffer(buffer, dtype=bfloat16).tolist()  # type: ignore[return-value]
+
     return np.frombuffer(buffer, dtype=dtype.lower()).tolist()  # type: ignore[return-value]
 
 
diff --git a/redisvl/utils/optimize/cache.py b/redisvl/utils/optimize/cache.py
@@ -1,6 +1,8 @@
 from typing import Any, Callable, Dict, List
 
-import numpy as np
+from redisvl.utils.utils import lazy_import
+
+np = lazy_import("numpy")
 from ranx import Qrels, Run, evaluate
 
 from redisvl.extensions.cache.llm.semantic import SemanticCache
diff --git a/redisvl/utils/optimize/router.py b/redisvl/utils/optimize/router.py
@@ -1,7 +1,9 @@
 import random
 from typing import Any, Callable, Dict, List
 
-import numpy as np
+from redisvl.utils.utils import lazy_import
+
+np = lazy_import("numpy")
 from ranx import Qrels, Run, evaluate
 
 from redisvl.extensions.router.semantic import SemanticRouter
diff --git a/redisvl/utils/optimize/utils.py b/redisvl/utils/optimize/utils.py
@@ -1,6 +1,8 @@
 from typing import List
 
-import numpy as np
+from redisvl.utils.utils import lazy_import
+
+np = lazy_import("numpy")
 from ranx import Qrels
 
 from redisvl.utils.optimize.schema import LabeledData
diff --git a/redisvl/utils/utils.py b/redisvl/utils/utils.py
@@ -321,15 +321,17 @@ def __getattr__(self, name: str) -> Any:
                     else:
                         # This means we couldn't find the attribute in the module path
                         raise AttributeError(
-                            f"{self._parts[0]} has no attribute '{self._parts[1]}'"
+                            f"module '{self._parts[0]}' has no attribute '{self._parts[1]}'"
                         )
 
             # If we have a module, get the requested attribute
             if hasattr(self._module, name):
                 return getattr(self._module, name)
 
             # If the attribute doesn't exist, raise AttributeError
-            raise AttributeError(f"{self._module_path} has no attribute '{name}'")
+            raise AttributeError(
+                f"module '{self._module_path}' has no attribute '{name}'"
+            )
 
         def __call__(self, *args: Any, **kwargs: Any) -> Any:
             # Import the module if it hasn't been imported yet
diff --git a/schemas/semantic_router.yaml b/schemas/semantic_router.yaml
@@ -1,4 +1,4 @@
-name: test-router-01JSHK4MJ79HH51PS6WEK6M9MF
+name: test-router
 routes:
 - name: greeting
   references:
diff --git a/tests/integration/test_semantic_router.py b/tests/integration/test_semantic_router.py
@@ -200,6 +200,7 @@ def test_from_dict(semantic_router):
 
 def test_to_yaml(semantic_router):
     yaml_file = str(get_base_path().joinpath("../../schemas/semantic_router.yaml"))
+    semantic_router.name = "test-router"
     semantic_router.to_yaml(yaml_file, overwrite=True)
     assert pathlib.Path(yaml_file).exists()
 
diff --git a/tests/integration/test_threshold_optimizer.py b/tests/integration/test_threshold_optimizer.py
@@ -5,6 +5,8 @@
 if sys.version_info.major == 3 and sys.version_info.minor < 10:
     pytest.skip("Test requires Python 3.10 or higher", allow_module_level=True)
 
+from redis import Redis
+
 from redisvl.extensions.cache.llm import SemanticCache
 from redisvl.extensions.router import Route, SemanticRouter
 from redisvl.extensions.router.schema import RoutingConfig
@@ -35,13 +37,13 @@ def routes():
 
 
 @pytest.fixture
-def semantic_router(client, routes, hf_vectorizer):
+def semantic_router(redis_url, routes, hf_vectorizer):
     router = SemanticRouter(
         name="test-router",
         routes=routes,
         vectorizer=hf_vectorizer,
         routing_config=RoutingConfig(max_k=2),
-        redis_client=client,
+        redis_url=redis_url,
         overwrite=False,
     )
     yield router
@@ -87,9 +89,10 @@ def test_data_optimization():
 
 
 def test_routes_different_distance_thresholds_optimizer_default(
-    semantic_router, routes, redis_url, test_data_optimization, hf_vectorizer
+    routes, redis_url, test_data_optimization, hf_vectorizer
 ):
-    redis_version = semantic_router._index.client.info()["redis_version"]
+    redis = Redis.from_url(redis_url)
+    redis_version = redis.info()["redis_version"]
     if not compare_versions(redis_version, "7.0.0"):
         pytest.skip("Not using a late enough version of Redis")
 
@@ -121,10 +124,10 @@ def test_routes_different_distance_thresholds_optimizer_default(
 
 
 def test_routes_different_distance_thresholds_optimizer_precision(
-    semantic_router, routes, redis_url, test_data_optimization, hf_vectorizer
+    routes, redis_url, test_data_optimization, hf_vectorizer
 ):
-
-    redis_version = semantic_router._index.client.info()["redis_version"]
+    redis = Redis.from_url(redis_url)
+    redis_version = redis.info()["redis_version"]
     if not compare_versions(redis_version, "7.0.0"):
         pytest.skip("Not using a late enough version of Redis")
 
@@ -158,9 +161,10 @@ def test_routes_different_distance_thresholds_optimizer_precision(
 
 
 def test_routes_different_distance_thresholds_optimizer_recall(
-    semantic_router, routes, redis_url, test_data_optimization, hf_vectorizer
+    routes, redis_url, test_data_optimization, hf_vectorizer, client
 ):
-    redis_version = semantic_router._index.client.info()["redis_version"]
+    redis = Redis.from_url(redis_url)
+    redis_version = redis.info()["redis_version"]
     if not compare_versions(redis_version, "7.0.0"):
         pytest.skip("Not using a late enough version of Redis")
 
diff --git a/tests/unit/test_utils.py b/tests/unit/test_utils.py

Original file line number	Diff line number	Diff line change
`@@ -1,4 +1,4 @@`
`1`		`-name: test-router-01JSHK4MJ79HH51PS6WEK6M9MF`
	`1`	`+name: test-router`
`2`	`2`	`routes:`
`3`	`3`	`- name: greeting`
`4`	`4`	`references:`