Introduce HashMap and HashSet type aliases

datafusion/common/src/lib.rs

@@ -66,6 +66,7 @@ pub use functional_dependencies::{
     get_target_functional_dependencies, Constraint, Constraints, Dependency,
     FunctionalDependence, FunctionalDependencies,
 };
+use hashbrown::hash_map::DefaultHashBuilder;
 pub use join_type::{JoinConstraint, JoinSide, JoinType};
 pub use param_value::ParamValues;
 pub use scalar::{ScalarType, ScalarValue};
@@ -87,6 +88,9 @@ pub use error::{
     _substrait_datafusion_err,
 };
 
+// The HashMap implementation that should be the uniform default
+pub type HashMap<K, V, S = DefaultHashBuilder> = hashbrown::HashMap<K, V, S>;
+
 /// Downcast an Arrow Array to a concrete type, return an `DataFusionError::Internal` if the cast is
 /// not possible. In normal usage of DataFusion the downcast should always succeed.
 ///

datafusion/core/src/catalog_common/listing_schema.rs

@@ -18,14 +18,16 @@
 //! [`ListingSchemaProvider`]: [`SchemaProvider`] that scans ObjectStores for tables automatically
 
 use std::any::Any;
-use std::collections::{HashMap, HashSet};
+use std::collections::HashSet;
 use std::path::Path;
 use std::sync::{Arc, Mutex};
 
 use crate::catalog::{SchemaProvider, TableProvider, TableProviderFactory};
 use crate::execution::context::SessionState;
 
-use datafusion_common::{Constraints, DFSchema, DataFusionError, TableReference};
+use datafusion_common::{
+    Constraints, DFSchema, DataFusionError, HashMap, TableReference,
+};
 use datafusion_expr::CreateExternalTable;
 
 use async_trait::async_trait;

datafusion/core/src/datasource/file_format/parquet.rs

@@ -63,7 +63,7 @@ use datafusion_physical_plan::metrics::MetricsSet;
 
 use async_trait::async_trait;
 use bytes::Bytes;
-use hashbrown::HashMap;
+use datafusion_common::HashMap;
 use log::debug;
 use object_store::buffered::BufWriter;
 use parquet::arrow::arrow_writer::{

datafusion/core/tests/fuzz_cases/aggregate_fuzz.rs

@@ -42,8 +42,8 @@ use test_utils::{add_empty_batches, StringBatchGenerator};
 use crate::fuzz_cases::aggregation_fuzzer::{
     AggregationFuzzerBuilder, ColumnDescr, DatasetGeneratorConfig, QueryBuilder,
 };
+use datafusion_common::HashMap;
 use datafusion_physical_expr_common::sort_expr::LexOrdering;
-use hashbrown::HashMap;
 use rand::rngs::StdRng;
 use rand::{Rng, SeedableRng};
 use tokio::task::JoinSet;

datafusion/core/tests/fuzz_cases/window_fuzz.rs

@@ -45,10 +45,10 @@ use datafusion_physical_expr::{PhysicalExpr, PhysicalSortExpr};
 use test_utils::add_empty_batches;
 
 use datafusion::functions_window::row_number::row_number_udwf;
+use datafusion_common::HashMap;
 use datafusion_functions_window::lead_lag::{lag_udwf, lead_udwf};
 use datafusion_functions_window::rank::{dense_rank_udwf, rank_udwf};
 use datafusion_physical_expr_common::sort_expr::LexOrdering;
-use hashbrown::HashMap;
 use rand::distributions::Alphanumeric;
 use rand::rngs::StdRng;
 use rand::{Rng, SeedableRng};

datafusion/core/tests/user_defined/user_defined_scalar_functions.rs

@@ -16,7 +16,6 @@
 // under the License.
 
 use std::any::Any;
-use std::collections::HashMap;
 use std::hash::{DefaultHasher, Hash, Hasher};
 use std::sync::Arc;
 
@@ -39,7 +38,8 @@ use datafusion_common::cast::{as_float64_array, as_int32_array};
 use datafusion_common::tree_node::{Transformed, TreeNode};
 use datafusion_common::{
     assert_batches_eq, assert_batches_sorted_eq, assert_contains, exec_err, internal_err,
-    not_impl_err, plan_err, DFSchema, DataFusionError, ExprSchema, Result, ScalarValue,
+    not_impl_err, plan_err, DFSchema, DataFusionError, ExprSchema, HashMap, Result,
+    ScalarValue,
 };
 use datafusion_expr::simplify::{ExprSimplifyResult, SimplifyInfo};
 use datafusion_expr::{

datafusion/execution/src/memory_pool/pool.rs

@@ -16,8 +16,8 @@
 // under the License.
 
 use crate::memory_pool::{MemoryConsumer, MemoryPool, MemoryReservation};
+use datafusion_common::HashMap;
 use datafusion_common::{resources_datafusion_err, DataFusionError, Result};
-use hashbrown::HashMap;
 use log::debug;
 use parking_lot::Mutex;
 use std::{

datafusion/expr/src/execution_props.rs

@@ -18,7 +18,7 @@
 use crate::var_provider::{VarProvider, VarType};
 use chrono::{DateTime, TimeZone, Utc};
 use datafusion_common::alias::AliasGenerator;
-use std::collections::HashMap;
+use datafusion_common::HashMap;
 use std::sync::Arc;
 
 /// Holds per-query execution properties and data (such as statement

datafusion/expr/src/registry.rs

@@ -20,8 +20,8 @@
 use crate::expr_rewriter::FunctionRewrite;
 use crate::planner::ExprPlanner;
 use crate::{AggregateUDF, ScalarUDF, UserDefinedLogicalNode, WindowUDF};
-use datafusion_common::{not_impl_err, plan_datafusion_err, Result};
-use std::collections::{HashMap, HashSet};
+use datafusion_common::{not_impl_err, plan_datafusion_err, HashMap, Result};
+use std::collections::HashSet;
 use std::fmt::Debug;
 use std::sync::Arc;
 

datafusion/expr/src/utils.rs

@@ -18,7 +18,7 @@
 //! Expression utilities
 
 use std::cmp::Ordering;
-use std::collections::{HashMap, HashSet};
+use std::collections::HashSet;
 use std::ops::Deref;
 use std::sync::Arc;
 
@@ -36,7 +36,7 @@ use datafusion_common::tree_node::{
 use datafusion_common::utils::get_at_indices;
 use datafusion_common::{
     internal_err, plan_datafusion_err, plan_err, Column, DFSchema, DFSchemaRef,
-    DataFusionError, Result, TableReference,
+    DataFusionError, HashMap, Result, TableReference,
 };
 
 use indexmap::IndexSet;

datafusion/functions-aggregate/src/regr.rs

@@ -24,8 +24,10 @@ use arrow::{
     datatypes::DataType,
     datatypes::Field,
 };
-use datafusion_common::{downcast_value, plan_err, unwrap_or_internal_err, ScalarValue};
-use datafusion_common::{DataFusionError, Result};
+use datafusion_common::{
+    downcast_value, plan_err, unwrap_or_internal_err, DataFusionError, HashMap, Result,
+    ScalarValue,
+};
 use datafusion_expr::aggregate_doc_sections::DOC_SECTION_STATISTICAL;
 use datafusion_expr::function::{AccumulatorArgs, StateFieldsArgs};
 use datafusion_expr::type_coercion::aggregates::NUMERICS;
@@ -34,7 +36,6 @@ use datafusion_expr::{
     Accumulator, AggregateUDFImpl, Documentation, Signature, Volatility,
 };
 use std::any::Any;
-use std::collections::HashMap;
 use std::fmt::Debug;
 use std::mem::size_of_val;
 use std::sync::OnceLock;

datafusion/functions/src/unicode/translate.rs

@@ -22,7 +22,7 @@ use arrow::array::{
     ArrayAccessor, ArrayIter, ArrayRef, AsArray, GenericStringArray, OffsetSizeTrait,
 };
 use arrow::datatypes::DataType;
-use hashbrown::HashMap;
+use datafusion_common::HashMap;
 use unicode_segmentation::UnicodeSegmentation;
 
 use crate::utils::{make_scalar_function, utf8_to_str_type};

datafusion/optimizer/src/decorrelate.rs

@@ -17,7 +17,7 @@
 
 //! [`PullUpCorrelatedExpr`] converts correlated subqueries to `Joins`
 
-use std::collections::{BTreeSet, HashMap};
+use std::collections::BTreeSet;
 use std::ops::Deref;
 use std::sync::Arc;
 
@@ -27,7 +27,7 @@ use crate::utils::collect_subquery_cols;
 use datafusion_common::tree_node::{
     Transformed, TransformedResult, TreeNode, TreeNodeRecursion, TreeNodeRewriter,
 };
-use datafusion_common::{plan_err, Column, DFSchemaRef, Result, ScalarValue};
+use datafusion_common::{plan_err, Column, DFSchemaRef, HashMap, Result, ScalarValue};
 use datafusion_expr::expr::Alias;
 use datafusion_expr::simplify::SimplifyContext;
 use datafusion_expr::utils::{conjunction, find_join_exprs, split_conjunction};

datafusion/physical-expr-common/src/binary_view_map.rs

@@ -393,7 +393,7 @@ where
 #[cfg(test)]
 mod tests {
     use arrow::array::{BinaryViewArray, GenericByteViewArray, StringViewArray};
-    use hashbrown::HashMap;
+    use datafusion_common::HashMap;
 
     use super::*;
 

datafusion/physical-expr/src/expressions/in_list.rs

@@ -44,8 +44,8 @@ use datafusion_expr::ColumnarValue;
 use datafusion_physical_expr_common::datum::compare_with_eq;
 
 use ahash::RandomState;
+use datafusion_common::HashMap;
 use hashbrown::hash_map::RawEntryMut;
-use hashbrown::HashMap;
 
 /// InList
 pub struct InListExpr {

datafusion/physical-expr/src/utils/guarantee.rs

@@ -20,9 +20,9 @@
 
 use crate::utils::split_disjunction;
 use crate::{split_conjunction, PhysicalExpr};
-use datafusion_common::{Column, ScalarValue};
+use datafusion_common::{Column, HashMap, ScalarValue};
 use datafusion_expr::Operator;
-use std::collections::{HashMap, HashSet};
+use std::collections::HashSet;
 use std::fmt::{self, Display, Formatter};
 use std::sync::Arc;
 

datafusion/physical-expr/src/utils/mod.rs

@@ -20,7 +20,6 @@ pub use guarantee::{Guarantee, LiteralGuarantee};
 use hashbrown::HashSet;
 
 use std::borrow::Borrow;
-use std::collections::HashMap;
 use std::sync::Arc;
 
 use crate::expressions::{BinaryExpr, Column};
@@ -32,7 +31,7 @@ use arrow::datatypes::SchemaRef;
 use datafusion_common::tree_node::{
     Transformed, TransformedResult, TreeNode, TreeNodeRecursion,
 };
-use datafusion_common::Result;
+use datafusion_common::{HashMap, Result};
 use datafusion_expr::Operator;
 
 use datafusion_physical_expr_common::sort_expr::{LexOrdering, LexOrderingRef};

datafusion/physical-plan/src/metrics/mod.rs

@@ -28,7 +28,7 @@ use std::{
     sync::Arc,
 };
 
-use hashbrown::HashMap;
+use datafusion_common::HashMap;
 
 // public exports
 pub use baseline::{BaselineMetrics, RecordOutput};

datafusion/physical-plan/src/repartition/mod.rs

@@ -50,10 +50,10 @@ use datafusion_execution::TaskContext;
 use datafusion_physical_expr::{EquivalenceProperties, PhysicalExpr};
 
 use crate::execution_plan::CardinalityEffect;
+use datafusion_common::HashMap;
 use datafusion_physical_expr_common::sort_expr::{LexOrdering, PhysicalSortExpr};
 use futures::stream::Stream;
 use futures::{FutureExt, StreamExt, TryStreamExt};
-use hashbrown::HashMap;
 use log::trace;
 use parking_lot::Mutex;
 

datafusion/physical-plan/src/topk/mod.rs

@@ -27,14 +27,14 @@ use std::{cmp::Ordering, collections::BinaryHeap, sync::Arc};
 use crate::{stream::RecordBatchStreamAdapter, SendableRecordBatchStream};
 use arrow_array::{Array, ArrayRef, RecordBatch};
 use arrow_schema::SchemaRef;
+use datafusion_common::HashMap;
 use datafusion_common::Result;
 use datafusion_execution::{
     memory_pool::{MemoryConsumer, MemoryReservation},
     runtime_env::RuntimeEnv,
 };
 use datafusion_physical_expr::PhysicalSortExpr;
 use datafusion_physical_expr_common::sort_expr::LexOrdering;
-use hashbrown::HashMap;
 
 use super::metrics::{BaselineMetrics, Count, ExecutionPlanMetricsSet, MetricBuilder};
 

datafusion/physical-plan/src/unnest.rs

@@ -18,7 +18,6 @@
 //! Define a plan for unnesting values in columns that contain a list type.
 
 use std::cmp::{self, Ordering};
-use std::collections::HashMap;
 use std::{any::Any, sync::Arc};
 
 use super::metrics::{self, ExecutionPlanMetricsSet, MetricBuilder, MetricsSet};
@@ -40,7 +39,7 @@ use arrow::record_batch::RecordBatch;
 use arrow_array::{Int64Array, Scalar, StructArray};
 use arrow_ord::cmp::lt;
 use datafusion_common::{
-    exec_datafusion_err, exec_err, internal_err, Result, UnnestOptions,
+    exec_datafusion_err, exec_err, internal_err, HashMap, Result, UnnestOptions,
 };
 use datafusion_execution::TaskContext;
 use datafusion_physical_expr::EquivalenceProperties;

datafusion/physical-plan/src/windows/bounded_window_agg_exec.rs

@@ -22,7 +22,7 @@
 
 use std::any::Any;
 use std::cmp::{min, Ordering};
-use std::collections::{HashMap, VecDeque};
+use std::collections::VecDeque;
 use std::pin::Pin;
 use std::sync::Arc;
 use std::task::{Context, Poll};
@@ -51,7 +51,9 @@ use datafusion_common::stats::Precision;
 use datafusion_common::utils::{
     evaluate_partition_ranges, get_at_indices, get_row_at_idx,
 };
-use datafusion_common::{arrow_datafusion_err, exec_err, DataFusionError, Result};
+use datafusion_common::{
+    arrow_datafusion_err, exec_err, DataFusionError, HashMap, Result,
+};
 use datafusion_execution::TaskContext;
 use datafusion_expr::window_state::{PartitionBatchState, WindowAggState};
 use datafusion_expr::ColumnarValue;

datafusion/sql/src/unparser/rewrite.rs

@@ -15,15 +15,12 @@
 // specific language governing permissions and limitations
 // under the License.
 
-use std::{
-    collections::{HashMap, HashSet},
-    sync::Arc,
-};
+use std::{collections::HashSet, sync::Arc};
 
 use arrow_schema::Schema;
 use datafusion_common::{
     tree_node::{Transformed, TransformedResult, TreeNode, TreeNodeRewriter},
-    Column, Result, TableReference,
+    Column, HashMap, Result, TableReference,
 };
 use datafusion_expr::{expr::Alias, tree_node::transform_sort_vec};
 use datafusion_expr::{Expr, LogicalPlan, Projection, Sort, SortExpr};

datafusion/sql/src/utils.rs

@@ -17,7 +17,6 @@
 
 //! SQL Utility Functions
 
-use std::collections::HashMap;
 use std::vec;
 
 use arrow_schema::{
@@ -27,8 +26,8 @@ use datafusion_common::tree_node::{
     Transformed, TransformedResult, TreeNode, TreeNodeRecursion, TreeNodeRewriter,
 };
 use datafusion_common::{
-    exec_err, internal_err, plan_err, Column, DFSchemaRef, DataFusionError, Result,
-    ScalarValue,
+    exec_err, internal_err, plan_err, Column, DFSchemaRef, DataFusionError, HashMap,
+    Result, ScalarValue,
 };
 use datafusion_expr::builder::get_struct_unnested_columns;
 use datafusion_expr::expr::{Alias, GroupingSet, Unnest, WindowFunction};

datafusion/substrait/src/extensions.rs

@@ -15,8 +15,7 @@
 // specific language governing permissions and limitations
 // under the License.
 
-use datafusion::common::{plan_err, DataFusionError};
-use std::collections::HashMap;
+use datafusion::common::{plan_err, DataFusionError, HashMap};
 use substrait::proto::extensions::simple_extension_declaration::{
     ExtensionFunction, ExtensionType, ExtensionTypeVariation, MappingType,
 };