chore: Benchmark deps cleanup (apache#14793)

* Synchronize criterion version across workspace

Use same version everywhere. This means update for some crates.

* Synchronize and update rand version across workspace

Use same version everywhere and update.

* Avoid linear search for character in bench helper

Get the char in O(1) instead of O(n).

Author: findepi

Cargo.toml

@@ -95,6 +95,7 @@ async-trait = "0.1.73"
 bigdecimal = "0.4.7"
 bytes = "1.10"
 chrono = { version = "0.4.38", default-features = false }
+criterion = "0.5.1"
 ctor = "0.2.9"
 dashmap = "6.0.1"
 datafusion = { path = "datafusion/core", version = "45.0.0", default-features = false }
@@ -144,7 +145,7 @@ pbjson-types = "0.7"
 # Should match arrow-flight's version of prost.
 prost = "0.13.1"
 prost-derive = "0.13.1"
-rand = "0.8"
+rand = "0.8.5"
 recursive = "0.1.1"
 regex = "1.8"
 rstest = "0.24.0"

datafusion/core/Cargo.toml

@@ -131,7 +131,7 @@ zstd = { version = "0.13", optional = true, default-features = false }
 
 [dev-dependencies]
 async-trait = { workspace = true }
-criterion = { version = "0.5", features = ["async_tokio"] }
+criterion = { workspace = true, features = ["async_tokio"] }
 ctor = { workspace = true }
 dashmap = "6.1.0"
 datafusion-doc = { workspace = true }

datafusion/functions-aggregate-common/Cargo.toml

@@ -44,7 +44,7 @@ datafusion-expr-common = { workspace = true }
 datafusion-physical-expr-common = { workspace = true }
 
 [dev-dependencies]
-criterion = "0.5"
+criterion = { workspace = true }
 rand = { workspace = true }
 
 [[bench]]

datafusion/functions-aggregate/Cargo.toml

@@ -54,7 +54,7 @@ paste = "1.0.14"
 
 [dev-dependencies]
 arrow = { workspace = true, features = ["test_utils"] }
-criterion = "0.5"
+criterion = { workspace = true }
 rand = { workspace = true }
 
 [[bench]]

datafusion/functions-nested/Cargo.toml

@@ -53,8 +53,8 @@ log = { workspace = true }
 paste = "1.0.14"
 
 [dev-dependencies]
-criterion = { version = "0.5", features = ["async_tokio"] }
-rand = "0.8.5"
+criterion = { workspace = true, features = ["async_tokio"] }
+rand = { workspace = true }
 
 [[bench]]
 harness = false

datafusion/functions/Cargo.toml

@@ -89,7 +89,7 @@ uuid = { version = "1.13", features = ["v4"], optional = true }
 
 [dev-dependencies]
 arrow = { workspace = true, features = ["test_utils"] }
-criterion = "0.5"
+criterion = { workspace = true }
 rand = { workspace = true }
 tokio = { workspace = true, features = ["macros", "rt", "sync"] }
 

datafusion/functions/benches/helper.rs

@@ -35,7 +35,7 @@ pub fn gen_string_array(
     let rng_ref = &mut rng;
 
     let corpus = "DataFusionДатаФусион数据融合📊🔥"; // includes utf8 encoding with 1~4 bytes
-    let corpus_char_count = corpus.chars().count();
+    let corpus = corpus.chars().collect::<Vec<_>>();
 
     let mut output_string_vec: Vec<Option<String>> = Vec::with_capacity(n_rows);
     for _ in 0..n_rows {
@@ -46,8 +46,7 @@ pub fn gen_string_array(
             // Generate random UTF8 string
             let mut generated_string = String::with_capacity(str_len_chars);
             for _ in 0..str_len_chars {
-                let idx = rng_ref.gen_range(0..corpus_char_count);
-                let char = corpus.chars().nth(idx).unwrap();
+                let char = corpus[rng_ref.gen_range(0..corpus.len())];
                 generated_string.push(char);
             }
             output_string_vec.push(Some(generated_string));

datafusion/physical-expr/Cargo.toml

@@ -55,7 +55,7 @@ petgraph = "0.7.1"
 
 [dev-dependencies]
 arrow = { workspace = true, features = ["test_utils"] }
-criterion = "0.5"
+criterion = { workspace = true }
 datafusion-functions = { workspace = true }
 rand = { workspace = true }
 rstest = { workspace = true }

datafusion/physical-plan/Cargo.toml

@@ -65,7 +65,7 @@ pin-project-lite = "^0.2.7"
 tokio = { workspace = true }
 
 [dev-dependencies]
-criterion = { version = "0.5", features = ["async_futures"] }
+criterion = { workspace = true, features = ["async_futures"] }
 datafusion-functions-aggregate = { workspace = true }
 datafusion-functions-window = { workspace = true }
 rand = { workspace = true }