Chore/upgrade datafusion 44 (#973)

* Bump DataFusion version to 44

* Trait definition for plan properties now returns LexOrdering

* find_df_window_func was removed upstream

* Prepare and Execute variants were removed from LogicalPlan

* Substrait functions now take SessionState instead of SessionContext

* Remove unused import

* RuntimeConfig is now deprecated

* Switch from RuntimeConfig to RuntimeEnvBuilder

* Update return types on unit tests

* DF 44 changes the execution plan properties to have boundedness and emission type

* Initcap now returns stringview

* Bump datafusion version in example

Author: timsaucer

Cargo.lock

@@ -38,11 +38,10 @@ tokio = { version = "1.41", features = ["macros", "rt", "rt-multi-thread", "sync
 pyo3 = { version = "0.22", features = ["extension-module", "abi3", "abi3-py38"] }
 pyo3-async-runtimes = { version = "0.22", features = ["tokio-runtime"]}
 arrow = { version = "53", features = ["pyarrow"] }
-datafusion = { version = "43.0.0", features = ["pyarrow", "avro", "unicode_expressions"] }
-datafusion-substrait = { version = "43.0.0", optional = true }
-datafusion-proto = { version = "43.0.0" }
-datafusion-ffi = { version = "43.0.0" }
-datafusion-functions-window-common = { version = "43.0.0" }
+datafusion = { version = "44.0.0", features = ["pyarrow", "avro", "unicode_expressions"] }
+datafusion-substrait = { version = "44.0.0", optional = true }
+datafusion-proto = { version = "44.0.0" }
+datafusion-ffi = { version = "44.0.0" }
 prost = "0.13" # keep in line with `datafusion-substrait`
 uuid = { version = "1.11", features = ["v4"] }
 mimalloc = { version = "0.1", optional = true, default-features = false, features = ["local_dynamic_tls"] }

Cargo.toml

@@ -87,7 +87,7 @@ It is possible to configure runtime (memory and disk settings) and configuration
 
 ```python
 runtime = (
-    RuntimeConfig()
+    RuntimeEnvBuilder()
     .with_disk_manager_os()
     .with_fair_spill_pool(10000000)
 )

README.md

@@ -22,7 +22,7 @@
 from datafusion import (
     col,
     functions as f,
-    RuntimeConfig,
+    RuntimeEnvBuilder,
     SessionConfig,
     SessionContext,
 )
@@ -85,7 +85,9 @@ def execute(df):
 
 # create a session context with explicit runtime and config settings
 runtime = (
-    RuntimeConfig().with_disk_manager_os().with_fair_spill_pool(64 * 1024 * 1024 * 1024)
+    RuntimeEnvBuilder()
+    .with_disk_manager_os()
+    .with_fair_spill_pool(64 * 1024 * 1024 * 1024)
 )
 config = (
     SessionConfig()

benchmarks/db-benchmark/groupby-datafusion.py

@@ -28,7 +28,7 @@ def bench(data_path, query_path):
 
         # create context
         # runtime = (
-        #     RuntimeConfig()
+        #     RuntimeEnvBuilder()
         #     .with_disk_manager_os()
         #     .with_fair_spill_pool(10000000)
         # )

benchmarks/tpch/tpch.py

@@ -19,18 +19,18 @@ Configuration
 =============
 
 Let's look at how we can configure DataFusion. When creating a :py:class:`~datafusion.context.SessionContext`, you can pass in
-a :py:class:`~datafusion.context.SessionConfig` and :py:class:`~datafusion.context.RuntimeConfig` object. These two cover a wide range of options.
+a :py:class:`~datafusion.context.SessionConfig` and :py:class:`~datafusion.context.RuntimeEnvBuilder` object. These two cover a wide range of options.
 
 .. code-block:: python
 
-    from datafusion import RuntimeConfig, SessionConfig, SessionContext
+    from datafusion import RuntimeEnvBuilder, SessionConfig, SessionContext
 
     # create a session context with default settings
     ctx = SessionContext()
     print(ctx)
 
     # create a session context with explicit runtime and config settings
-    runtime = RuntimeConfig().with_disk_manager_os().with_fair_spill_pool(10000000)
+    runtime = RuntimeEnvBuilder().with_disk_manager_os().with_fair_spill_pool(10000000)
     config = (
         SessionConfig()
         .with_create_default_catalog_and_schema(True)
@@ -48,4 +48,4 @@ a :py:class:`~datafusion.context.SessionConfig` and :py:class:`~datafusion.conte
 
 
 You can read more about available :py:class:`~datafusion.context.SessionConfig` options in the `rust DataFusion Configuration guide <https://arrow.apache.org/datafusion/user-guide/configs.html>`_,
-and about :code:`RuntimeConfig` options in the rust `online API documentation <https://docs.rs/datafusion/latest/datafusion/execution/runtime_env/struct.RuntimeConfig.html>`_.
+and about :code:`RuntimeEnvBuilder` options in the rust `online API documentation <https://docs.rs/datafusion/latest/datafusion/execution/runtime_env/struct.RuntimeEnvBuilder.html>`_.

docs/source/user-guide/configuration.rst

@@ -15,14 +15,14 @@
 # specific language governing permissions and limitations
 # under the License.
 
-from datafusion import RuntimeConfig, SessionConfig, SessionContext
+from datafusion import RuntimeEnvBuilder, SessionConfig, SessionContext
 
 # create a session context with default settings
 ctx = SessionContext()
 print(ctx)
 
 # create a session context with explicit runtime and config settings
-runtime = RuntimeConfig().with_disk_manager_os().with_fair_spill_pool(10000000)
+runtime = RuntimeEnvBuilder().with_disk_manager_os().with_fair_spill_pool(10000000)
 config = (
     SessionConfig()
     .with_create_default_catalog_and_schema(True)

examples/create-context.py

@@ -21,8 +21,8 @@ version = "0.1.0"
 edition = "2021"
 
 [dependencies]
-datafusion = { version = "43.0.0" }
-datafusion-ffi = { version = "43.0.0" }
+datafusion = { version = "44.0.0" }
+datafusion-ffi = { version = "44.0.0" }
 pyo3 = { version = "0.22.6", features = ["extension-module", "abi3", "abi3-py38"] }
 arrow = { version = "53.2.0" }
 arrow-array = { version = "53.2.0" }

examples/ffi-table-provider/Cargo.toml

@@ -29,7 +29,7 @@
 from .context import (
     SessionContext,
     SessionConfig,
-    RuntimeConfig,
+    RuntimeEnvBuilder,
     SQLOptions,
 )
 
@@ -66,7 +66,7 @@
     "SessionContext",
     "SessionConfig",
     "SQLOptions",
-    "RuntimeConfig",
+    "RuntimeEnvBuilder",
     "Expr",
     "ScalarUDF",
     "WindowFrame",

python/datafusion/__init__.py

@@ -20,7 +20,7 @@
 from __future__ import annotations
 
 from ._internal import SessionConfig as SessionConfigInternal
-from ._internal import RuntimeConfig as RuntimeConfigInternal
+from ._internal import RuntimeEnvBuilder as RuntimeEnvBuilderInternal
 from ._internal import SQLOptions as SQLOptionsInternal
 from ._internal import SessionContext as SessionContextInternal
 
@@ -265,56 +265,58 @@ def set(self, key: str, value: str) -> SessionConfig:
         return self
 
 
-class RuntimeConfig:
+class RuntimeEnvBuilder:
     """Runtime configuration options."""
 
     def __init__(self) -> None:
-        """Create a new :py:class:`RuntimeConfig` with default values."""
-        self.config_internal = RuntimeConfigInternal()
+        """Create a new :py:class:`RuntimeEnvBuilder` with default values."""
+        self.config_internal = RuntimeEnvBuilderInternal()
 
-    def with_disk_manager_disabled(self) -> RuntimeConfig:
+    def with_disk_manager_disabled(self) -> RuntimeEnvBuilder:
         """Disable the disk manager, attempts to create temporary files will error.
 
         Returns:
-            A new :py:class:`RuntimeConfig` object with the updated setting.
+            A new :py:class:`RuntimeEnvBuilder` object with the updated setting.
         """
         self.config_internal = self.config_internal.with_disk_manager_disabled()
         return self
 
-    def with_disk_manager_os(self) -> RuntimeConfig:
+    def with_disk_manager_os(self) -> RuntimeEnvBuilder:
         """Use the operating system's temporary directory for disk manager.
 
         Returns:
-            A new :py:class:`RuntimeConfig` object with the updated setting.
+            A new :py:class:`RuntimeEnvBuilder` object with the updated setting.
         """
         self.config_internal = self.config_internal.with_disk_manager_os()
         return self
 
-    def with_disk_manager_specified(self, *paths: str | pathlib.Path) -> RuntimeConfig:
+    def with_disk_manager_specified(
+        self, *paths: str | pathlib.Path
+    ) -> RuntimeEnvBuilder:
         """Use the specified paths for the disk manager's temporary files.
 
         Args:
             paths: Paths to use for the disk manager's temporary files.
 
         Returns:
-            A new :py:class:`RuntimeConfig` object with the updated setting.
+            A new :py:class:`RuntimeEnvBuilder` object with the updated setting.
         """
         paths_list = [str(p) for p in paths]
         self.config_internal = self.config_internal.with_disk_manager_specified(
             paths_list
         )
         return self
 
-    def with_unbounded_memory_pool(self) -> RuntimeConfig:
+    def with_unbounded_memory_pool(self) -> RuntimeEnvBuilder:
         """Use an unbounded memory pool.
 
         Returns:
-            A new :py:class:`RuntimeConfig` object with the updated setting.
+            A new :py:class:`RuntimeEnvBuilder` object with the updated setting.
         """
         self.config_internal = self.config_internal.with_unbounded_memory_pool()
         return self
 
-    def with_fair_spill_pool(self, size: int) -> RuntimeConfig:
+    def with_fair_spill_pool(self, size: int) -> RuntimeEnvBuilder:
         """Use a fair spill pool with the specified size.
 
         This pool works best when you know beforehand the query has multiple spillable
@@ -335,16 +337,16 @@ def with_fair_spill_pool(self, size: int) -> RuntimeConfig:
             size: Size of the memory pool in bytes.
 
         Returns:
-            A new :py:class:`RuntimeConfig` object with the updated setting.
+            A new :py:class:`RuntimeEnvBuilder` object with the updated setting.
 
         Examples usage::
 
-            config = RuntimeConfig().with_fair_spill_pool(1024)
+            config = RuntimeEnvBuilder().with_fair_spill_pool(1024)
         """
         self.config_internal = self.config_internal.with_fair_spill_pool(size)
         return self
 
-    def with_greedy_memory_pool(self, size: int) -> RuntimeConfig:
+    def with_greedy_memory_pool(self, size: int) -> RuntimeEnvBuilder:
         """Use a greedy memory pool with the specified size.
 
         This pool works well for queries that do not need to spill or have a single
@@ -355,32 +357,39 @@ def with_greedy_memory_pool(self, size: int) -> RuntimeConfig:
             size: Size of the memory pool in bytes.
 
         Returns:
-            A new :py:class:`RuntimeConfig` object with the updated setting.
+            A new :py:class:`RuntimeEnvBuilder` object with the updated setting.
 
         Example usage::
 
-            config = RuntimeConfig().with_greedy_memory_pool(1024)
+            config = RuntimeEnvBuilder().with_greedy_memory_pool(1024)
         """
         self.config_internal = self.config_internal.with_greedy_memory_pool(size)
         return self
 
-    def with_temp_file_path(self, path: str | pathlib.Path) -> RuntimeConfig:
+    def with_temp_file_path(self, path: str | pathlib.Path) -> RuntimeEnvBuilder:
         """Use the specified path to create any needed temporary files.
 
         Args:
             path: Path to use for temporary files.
 
         Returns:
-            A new :py:class:`RuntimeConfig` object with the updated setting.
+            A new :py:class:`RuntimeEnvBuilder` object with the updated setting.
 
         Example usage::
 
-            config = RuntimeConfig().with_temp_file_path("/tmp")
+            config = RuntimeEnvBuilder().with_temp_file_path("/tmp")
         """
         self.config_internal = self.config_internal.with_temp_file_path(str(path))
         return self
 
 
+@deprecated("Use `RuntimeEnvBuilder` instead.")
+class RuntimeConfig(RuntimeEnvBuilder):
+    """See `RuntimeEnvBuilder`."""
+
+    pass
+
+
 class SQLOptions:
     """Options to be used when performing SQL queries."""
 
@@ -454,7 +463,9 @@ class SessionContext:
     """
 
     def __init__(
-        self, config: SessionConfig | None = None, runtime: RuntimeConfig | None = None
+        self,
+        config: SessionConfig | None = None,
+        runtime: RuntimeEnvBuilder | None = None,
     ) -> None:
         """Main interface for executing queries with DataFusion.