chore: remove lazy_static in favor of LazyLock (delta-io#365)

Turns out we don't use lazy static in many places so potentially removing the dependency is straight forward. Lazy static itself even gives an example from the standard library. https://github.com/rust-lang-nursery/lazy-static.rs?tab=readme-ov-file#standard-library --------- Co-authored-by: Ryan Johnson <[email protected]>
nicklan · Oct 2, 2024 · 0968281 · 0968281
1 parent 4c214c8
commit 0968281
Show file tree

Hide file tree

Showing 5 changed files with 38 additions and 44 deletions.
diff --git a/kernel/Cargo.toml b/kernel/Cargo.toml
@@ -21,7 +21,6 @@ either = "1.13"
 fix-hidden-lifetime-bug = "0.2"
 indexmap = "2.5.0"
 itertools = "0.13"
-lazy_static = "1.5"
 roaring = "0.10.6"
 serde = { version = "1", features = ["derive", "rc"] }
 serde_json = "1"

diff --git a/kernel/src/actions/mod.rs b/kernel/src/actions/mod.rs
@@ -5,11 +5,10 @@ pub mod deletion_vector;
 pub(crate) mod schemas;
 pub(crate) mod visitors;
 
-use std::collections::HashMap;
-
 use delta_kernel_derive::Schema;
-use lazy_static::lazy_static;
 use serde::{Deserialize, Serialize};
+use std::collections::HashMap;
+use std::sync::LazyLock;
 use visitors::{AddVisitor, MetadataVisitor, ProtocolVisitor};
 
 use self::deletion_vector::DeletionVectorDescriptor;
@@ -23,21 +22,19 @@ pub(crate) const METADATA_NAME: &str = "metaData";
 pub(crate) const PROTOCOL_NAME: &str = "protocol";
 pub(crate) const TRANSACTION_NAME: &str = "txn";
 
-lazy_static! {
-    static ref LOG_SCHEMA: StructType = StructType::new(
-        vec![
-            Option::<Add>::get_struct_field(ADD_NAME),
-            Option::<Remove>::get_struct_field(REMOVE_NAME),
-            Option::<Metadata>::get_struct_field(METADATA_NAME),
-            Option::<Protocol>::get_struct_field(PROTOCOL_NAME),
-            Option::<Transaction>::get_struct_field(TRANSACTION_NAME),
-            // We don't support the following actions yet
-            //Option<Cdc>::get_field(CDC_NAME),
-            //Option<CommitInfo>::get_field(COMMIT_INFO_NAME),
-            //Option<DomainMetadata>::get_field(DOMAIN_METADATA_NAME),
-        ]
-    );
-}
+static LOG_SCHEMA: LazyLock<StructType> = LazyLock::new(|| {
+    StructType::new(vec![
+        Option::<Add>::get_struct_field(ADD_NAME),
+        Option::<Remove>::get_struct_field(REMOVE_NAME),
+        Option::<Metadata>::get_struct_field(METADATA_NAME),
+        Option::<Protocol>::get_struct_field(PROTOCOL_NAME),
+        Option::<Transaction>::get_struct_field(TRANSACTION_NAME),
+        // We don't support the following actions yet
+        //Option<Cdc>::get_field(CDC_NAME),
+        //Option<CommitInfo>::get_field(COMMIT_INFO_NAME),
+        //Option<DomainMetadata>::get_field(DOMAIN_METADATA_NAME),
+    ])
+});
 
 pub(crate) fn get_log_schema() -> &'static StructType {
     &LOG_SCHEMA

diff --git a/kernel/src/expressions/scalars.rs b/kernel/src/expressions/scalars.rs
@@ -1,7 +1,7 @@
+use chrono::{DateTime, NaiveDate, NaiveDateTime, TimeZone, Utc};
 use std::cmp::Ordering;
 use std::fmt::{Display, Formatter};
-
-use chrono::{DateTime, NaiveDate, NaiveDateTime, TimeZone, Utc};
+use std::sync::LazyLock;
 
 use crate::schema::{ArrayType, DataType, PrimitiveType, StructField};
 use crate::utils::require;
@@ -296,9 +296,8 @@ impl PrimitiveType {
     pub fn parse_scalar(&self, raw: &str) -> Result<Scalar, Error> {
         use PrimitiveType::*;
 
-        lazy_static::lazy_static! {
-            static ref UNIX_EPOCH: DateTime<Utc> = DateTime::from_timestamp(0, 0).unwrap();
-        }
+        static UNIX_EPOCH: LazyLock<DateTime<Utc>> =
+            LazyLock::new(|| DateTime::from_timestamp(0, 0).unwrap());
 
         if raw.is_empty() {
             return Ok(Scalar::Null(self.data_type()));

diff --git a/kernel/src/scan/data_skipping.rs b/kernel/src/scan/data_skipping.rs
@@ -1,6 +1,6 @@
 use std::collections::HashSet;
 use std::ops::Not;
-use std::sync::Arc;
+use std::sync::{Arc, LazyLock};
 
 use tracing::debug;
 
@@ -180,13 +180,12 @@ impl DataSkippingFilter {
         table_schema: &SchemaRef,
         predicate: &Option<Expr>,
     ) -> Option<Self> {
-        lazy_static::lazy_static!(
-            static ref PREDICATE_SCHEMA: DataType = StructType::new(vec![
-                StructField::new("predicate", DataType::BOOLEAN, true),
-            ]).into();
-            static ref STATS_EXPR: Expr = Expr::column("add.stats");
-            static ref FILTER_EXPR: Expr = Expr::column("predicate").distinct(Expr::literal(false));
-        );
+        static PREDICATE_SCHEMA: LazyLock<DataType> = LazyLock::new(|| {
+            DataType::struct_type(vec![StructField::new("predicate", DataType::BOOLEAN, true)])
+        });
+        static STATS_EXPR: LazyLock<Expr> = LazyLock::new(|| Expr::column("add.stats"));
+        static FILTER_EXPR: LazyLock<Expr> =
+            LazyLock::new(|| Expr::column("predicate").distinct(Expr::literal(false)));
 
         let predicate = match predicate {
             Some(predicate) => predicate,

diff --git a/kernel/src/scan/log_replay.rs b/kernel/src/scan/log_replay.rs
@@ -1,8 +1,8 @@
+use std::clone::Clone;
 use std::collections::HashSet;
-use std::sync::Arc;
+use std::sync::{Arc, LazyLock};
 
 use either::Either;
-use lazy_static::lazy_static;
 use tracing::debug;
 
 use super::data_skipping::DataSkippingFilter;
@@ -77,11 +77,11 @@ impl DataVisitor for AddRemoveVisitor {
     }
 }
 
-lazy_static! {
-    // NB: If you update this schema, ensure you update the comment describing it in the doc comment
-    // for `scan_row_schema` in scan/mod.rs! You'll also need to update ScanFileVisitor as the
-    // indexes will be off
-    pub(crate) static ref SCAN_ROW_SCHEMA: Arc<StructType> = Arc::new(StructType::new(vec!(
+// NB: If you update this schema, ensure you update the comment describing it in the doc comment
+// for `scan_row_schema` in scan/mod.rs! You'll also need to update ScanFileVisitor as the
+// indexes will be off
+pub(crate) static SCAN_ROW_SCHEMA: LazyLock<Arc<StructType>> = LazyLock::new(|| {
+    Arc::new(StructType::new(vec![
         StructField::new("path", DataType::STRING, false),
         StructField::new("size", DataType::LONG, true),
         StructField::new("modificationTime", DataType::LONG, true),
@@ -95,7 +95,7 @@ lazy_static! {
                 StructField::new("sizeInBytes", DataType::INTEGER, false),
                 StructField::new("cardinality", DataType::LONG, false),
             ]),
-            true
+            true,
         ),
         StructField::new(
             "fileConstantValues",
@@ -104,11 +104,11 @@ lazy_static! {
                 MapType::new(DataType::STRING, DataType::STRING, false),
                 true,
             )]),
-            true
+            true,
         ),
-    )));
-    static ref SCAN_ROW_DATATYPE: DataType = SCAN_ROW_SCHEMA.as_ref().clone().into();
-}
+    ]))
+});
+static SCAN_ROW_DATATYPE: LazyLock<DataType> = LazyLock::new(|| SCAN_ROW_SCHEMA.clone().into());
 
 impl LogReplayScanner {
     /// Create a new [`LogReplayScanner`] instance