style

Author: myrrc

vortex-bench/src/tpcds/mod.rs

@@ -11,7 +11,7 @@ pub mod tpcds_benchmark;
 pub use tpcds_benchmark::TpcDsBenchmark;
 
 pub fn tpcds_queries() -> impl Iterator<Item = (usize, String)> {
-    (1..=99).map(|idx| (idx, tpcds_query(idx)))
+    (2..=2).map(|idx| (idx, tpcds_query(idx)))
 }
 
 // A few tpcds queries have multiple statements, this handles that

vortex-datafusion/src/convert/mod.rs

@@ -6,7 +6,6 @@ use vortex::error::VortexResult;
 pub(crate) mod exprs;
 mod scalars;
 pub(crate) mod schema;
-pub(crate) mod stats;
 
 /// First-party trait for implementing conversion from DataFusion types to Vortex types.
 pub(crate) trait FromDataFusion<D: ?Sized>: Sized {

vortex-datafusion/src/convert/stats.rs

@@ -36,11 +36,8 @@ use datafusion_physical_plan::filter_pushdown::PushedDown;
 use datafusion_physical_plan::stream::RecordBatchStreamAdapter;
 use futures::StreamExt;
 use futures::TryStreamExt;
-use futures::future::try_join_all;
 use vortex::array::VortexSessionExecute;
 use vortex::array::arrow::ArrowArrayExecutor;
-use vortex::dtype::DType;
-use vortex::dtype::FieldPath;
 use vortex::dtype::Nullability;
 use vortex::error::VortexExpect;
 use vortex::error::VortexResult;
@@ -61,7 +58,6 @@ use crate::convert::exprs::DefaultExpressionConvertor;
 use crate::convert::exprs::ExpressionConvertor;
 use crate::convert::exprs::ProcessedProjection;
 use crate::convert::exprs::make_vortex_predicate;
-use crate::convert::stats::stats_set_to_df;
 
 /// A builder for a [`VortexDataSource`].
 pub struct VortexDataSourceBuilder {
@@ -137,35 +133,14 @@ impl VortexDataSourceBuilder {
             ));
         }
 
-        let DType::Struct(fields, ..) = projection.return_dtype(self.data_source.dtype())? else {
-            vortex_bail!("Projection does not evaluate to a struct");
-        };
-
-        // We now compute initial statistics.
-        let field_paths: Vec<_> = fields
-            .names()
-            .iter()
-            .cloned()
-            .map(FieldPath::from_name)
-            .collect();
-        let statistics = try_join_all(
-            field_paths
-                .iter()
-                .map(|path| self.data_source.field_statistics(path)),
-        )
-        .await?
-        .iter()
-        .zip(fields.fields())
-        .map(|(stats, dtype)| stats_set_to_df(stats, &dtype))
-        .collect::<VortexResult<Vec<_>>>()?;
-
+        let statistics = Vec::new();
         Ok(VortexDataSource {
             data_source: self.data_source,
             session: self.session,
             initial_schema: Arc::clone(&arrow_schema),
             initial_projection: projection.clone(),
             initial_statistics: statistics.clone(),
-            projected_projection: projection.clone(),
+            projected_projection: projection,
             projected_schema: Arc::clone(&arrow_schema),
             projected_statistics: statistics.clone(),
             leftover_projection: None,

vortex-datafusion/src/v2/source.rs

@@ -205,7 +205,7 @@ impl<T: DataSourceTableFunction> TableFunction for T {
             result.add_result_column(column_name, column_type);
             let stats_set = &file_stats.stats_sets()[i];
             let dtype = &file_stats.dtypes()[i];
-            stats.push(ColumnStatistics::new(&stats_set, dtype.clone()));
+            stats.push(ColumnStatistics::new(stats_set, dtype.clone()));
         }
 
         Ok(DataSourceBindData {

vortex-duckdb/src/datasource.rs

@@ -27,7 +27,7 @@ pub(crate) unsafe extern "C-unwind" fn statistics<T: TableFunction>(
             .cast::<T::BindData>()
             .as_ref()
             .vortex_expect("bind_data null pointer");
-        let stats_ref = T::statistics(client_context, &bind_data, column_index);
+        let stats_ref = T::statistics(client_context, bind_data, column_index);
         let dtype = stats_ref.minmax_dtype;
 
         if let Some(value) = stats_ref.min {

vortex-duckdb/src/duckdb/table_function/statistics.rs

@@ -156,7 +156,7 @@ impl FileStatistics {
             other.stats_sets().iter(),
             dtypes.iter()
         ) {
-            let owned = std::mem::replace(this, StatsSet::default());
+            let owned = std::mem::take(this);
             *this = owned.merge_unordered(other_stat, dtype);
         }
 

vortex-file/src/footer/file_statistics.rs

@@ -192,7 +192,7 @@ impl MultiFileDataSource {
                 .iter()
                 .map(|(f, fs)| {
                     Arc::new(VortexFileReaderFactory {
-                        fs: Arc::clone(&fs),
+                        fs: Arc::clone(fs),
                         file: f.clone(),
                         session: self.session.clone(),
                         open_options_fn: Arc::clone(&self.open_options_fn),

vortex-file/src/multi/mod.rs

@@ -1,7 +1,6 @@
 // SPDX-License-Identifier: Apache-2.0
 // SPDX-FileCopyrightText: Copyright the Vortex contributors
 
-use std::any::Any;
 use std::collections::BTreeSet;
 use std::ops::Range;
 use std::sync::Arc;

vortex-layout/src/reader.rs

@@ -16,12 +16,10 @@ use futures::stream::StreamExt;
 use vortex_array::IntoArray;
 use vortex_array::arrays::ConstantArray;
 use vortex_array::dtype::DType;
-use vortex_array::dtype::FieldPath;
 use vortex_array::dtype::Nullability;
 use vortex_array::expr::Expression;
 use vortex_array::expr::stats::Precision;
 use vortex_array::scalar::Scalar;
-use vortex_array::stats::StatsSet;
 use vortex_array::stream::ArrayStreamAdapter;
 use vortex_array::stream::ArrayStreamExt;
 use vortex_array::stream::SendableArrayStream;