Add parser benchmark (#3990)

crates/ruff_benchmark/Cargo.toml

@@ -16,6 +16,10 @@ bench = false
 name = "linter"
 harness = false
 
+[[bench]]
+name = "parser"
+harness = false
+
 [dependencies]
 once_cell.workspace = true
 serde.workspace = true
@@ -25,7 +29,9 @@ ureq = "2.6.2"
 
 [dev-dependencies]
 ruff.path = "../ruff"
+ruff_python_ast.path = "../ruff_python_ast"
 criterion = { version = "0.4.0"}
+rustpython-parser.workspace = true
 
 [target.'cfg(target_os = "windows")'.dev-dependencies]
 mimalloc = "0.1.34"

crates/ruff_benchmark/benches/parser.rs

@@ -0,0 +1,82 @@
+use criterion::measurement::WallTime;
+use criterion::{criterion_group, criterion_main, BenchmarkId, Criterion, Throughput};
+use ruff_benchmark::{TestCase, TestCaseSpeed, TestFile, TestFileDownloadError};
+use ruff_python_ast::visitor::{walk_stmt, Visitor};
+use rustpython_parser::ast::Stmt;
+use std::time::Duration;
+
+#[cfg(target_os = "windows")]
+#[global_allocator]
+static GLOBAL: mimalloc::MiMalloc = mimalloc::MiMalloc;
+
+#[cfg(all(
+    not(target_os = "windows"),
+    not(target_os = "openbsd"),
+    any(
+        target_arch = "x86_64",
+        target_arch = "aarch64",
+        target_arch = "powerpc64"
+    )
+))]
+#[global_allocator]
+static GLOBAL: tikv_jemallocator::Jemalloc = tikv_jemallocator::Jemalloc;
+
+fn create_test_cases() -> Result<Vec<TestCase>, TestFileDownloadError> {
+    Ok(vec![
+        TestCase::fast(TestFile::try_download("numpy/globals.py", "https://raw.githubusercontent.com/numpy/numpy/89d64415e349ca75a25250f22b874aa16e5c0973/numpy/_globals.py")?),
+        TestCase::normal(TestFile::try_download(
+            "pydantic/types.py",
+            "https://raw.githubusercontent.com/pydantic/pydantic/83b3c49e99ceb4599d9286a3d793cea44ac36d4b/pydantic/types.py",
+        )?),
+        TestCase::normal(TestFile::try_download("numpy/ctypeslib.py", "https://raw.githubusercontent.com/numpy/numpy/e42c9503a14d66adfd41356ef5640c6975c45218/numpy/ctypeslib.py")?),
+        TestCase::slow(TestFile::try_download(
+            "large/dataset.py",
+            "https://raw.githubusercontent.com/DHI/mikeio/b7d26418f4db2909b0aa965253dbe83194d7bb5b/tests/test_dataset.py",
+        )?),
+    ])
+}
+
+struct CountVisitor {
+    count: usize,
+}
+
+impl<'a> Visitor<'a> for CountVisitor {
+    fn visit_stmt(&mut self, stmt: &'a Stmt) {
+        walk_stmt(self, stmt);
+        self.count += 1;
+    }
+}
+
+fn benchmark_parser(criterion: &mut Criterion<WallTime>) {
+    let test_cases = create_test_cases().unwrap();
+    let mut group = criterion.benchmark_group("parser");
+
+    for case in test_cases {
+        group.throughput(Throughput::Bytes(case.code().len() as u64));
+        group.measurement_time(match case.speed() {
+            TestCaseSpeed::Fast => Duration::from_secs(10),
+            TestCaseSpeed::Normal => Duration::from_secs(20),
+            TestCaseSpeed::Slow => Duration::from_secs(45),
+        });
+
+        group.bench_with_input(
+            BenchmarkId::from_parameter(case.name()),
+            &case,
+            |b, case| {
+                b.iter(|| {
+                    let parsed =
+                        rustpython_parser::parse_program(case.code(), case.name()).unwrap();
+
+                    let mut visitor = CountVisitor { count: 0 };
+                    visitor.visit_body(&parsed);
+                    visitor.count
+                });
+            },
+        );
+    }
+
+    group.finish();
+}
+
+criterion_group!(parser, benchmark_parser);
+criterion_main!(parser);