Add parse CLI to benchmark PGO

crates/oxc_cli/Cargo.toml

@@ -28,6 +28,11 @@ name = "oxlint"
 path = "src/lint/main.rs"
 test = false
 
+[[bin]]
+name = "oxparse"
+path = "src/parse/main.rs"
+test = false
+
 [[bin]]
 name = "oxformat"
 path = "src/format/main.rs"

crates/oxc_cli/src/lib.rs

@@ -1,6 +1,7 @@
 mod command;
 mod format;
 mod lint;
+mod parse;
 mod result;
 mod runner;
 mod walk;
@@ -9,6 +10,7 @@ pub use crate::{
     command::*,
     format::FormatRunner,
     lint::LintRunner,
+    parse::ParseRunner,
     result::{CliRunResult, LintResult},
     runner::Runner,
 };

crates/oxc_cli/src/parse/main.rs

@@ -0,0 +1,26 @@
+#![cfg(not(miri))] // Miri does not support custom allocators
+
+#[cfg(not(debug_assertions))]
+#[cfg(not(target_env = "msvc"))]
+#[global_allocator]
+static GLOBAL: jemallocator::Jemalloc = jemallocator::Jemalloc;
+
+#[cfg(not(debug_assertions))]
+#[cfg(target_os = "windows")]
+#[global_allocator]
+static GLOBAL: mimalloc::MiMalloc = mimalloc::MiMalloc;
+
+use oxc_cli::{CliRunResult, ParseRunner, Runner};
+
+fn main() -> CliRunResult {
+    init_miette();
+
+    let command = oxc_cli::format_command().fallback_to_usage().run();
+    command.handle_threads();
+    ParseRunner::new(command.format_options).run()
+}
+
+// Initialize the data which relies on `is_atty` system calls so they don't block subsequent threads.
+fn init_miette() {
+    miette::set_hook(Box::new(|_| Box::new(miette::MietteHandlerOpts::new().build()))).unwrap();
+}

crates/oxc_cli/src/parse/mod.rs

@@ -0,0 +1,78 @@
+use std::{env, path::Path};
+
+use ignore::gitignore::Gitignore;
+use oxc_allocator::Allocator;
+use oxc_parser::Parser;
+use oxc_span::{SourceType, VALID_EXTENSIONS};
+use rayon::prelude::{IntoParallelRefIterator, ParallelIterator};
+
+use crate::{
+    command::FormatOptions,
+    result::{CliRunResult, ParseResult},
+    walk::{Extensions, Walk},
+    Runner,
+};
+
+pub struct ParseRunner {
+    options: FormatOptions,
+}
+
+impl Runner for ParseRunner {
+    type Options = FormatOptions;
+
+    fn new(options: Self::Options) -> Self {
+        Self { options }
+    }
+
+    fn run(self) -> CliRunResult {
+        let FormatOptions { paths, ignore_options, .. } = self.options;
+
+        let mut paths = paths;
+
+        // The ignore crate whitelists explicit paths, but priority
+        // should be given to the ignore file. Many users lint
+        // automatically and pass a list of changed files explicitly.
+        // To accommodate this, unless `--no-ignore` is passed,
+        // pre-filter the paths.
+        if !paths.is_empty() && !ignore_options.no_ignore {
+            let (ignore, _err) = Gitignore::new(&ignore_options.ignore_path);
+            paths.retain(|p| if p.is_dir() { true } else { !ignore.matched(p, false).is_ignore() });
+        }
+
+        if paths.is_empty() {
+            if let Ok(cwd) = env::current_dir() {
+                paths.push(cwd);
+            } else {
+                return CliRunResult::InvalidOptions {
+                    message: "Failed to get current working directory.".to_string(),
+                };
+            }
+        }
+
+        let extensions = VALID_EXTENSIONS.to_vec();
+
+        let now = std::time::Instant::now();
+
+        let paths =
+            Walk::new(&paths, &ignore_options).with_extensions(Extensions(extensions)).paths();
+
+        let total_duration = paths.par_iter().map(|path| Self::parse(path)).sum();
+
+        CliRunResult::ParseResult(ParseResult {
+            parse_duration: total_duration,
+            duration: now.elapsed(),
+            number_of_files: paths.len(),
+        })
+    }
+}
+
+impl ParseRunner {
+    fn parse(path: &Path) -> std::time::Duration {
+        let source_text = std::fs::read_to_string(path).unwrap();
+        let allocator = Allocator::default();
+        let source_type = SourceType::from_path(path).unwrap();
+        let now = std::time::Instant::now();
+        let _ = Parser::new(&allocator, &source_text, source_type).preserve_parens(false).parse();
+        now.elapsed()
+    }
+}

crates/oxc_cli/src/result.rs

@@ -11,6 +11,7 @@ pub enum CliRunResult {
     PathNotFound { paths: Vec<PathBuf> },
     LintResult(LintResult),
     FormatResult(FormatResult),
+    ParseResult(ParseResult),
     TypeCheckResult { duration: Duration, number_of_diagnostics: usize },
 }
 
@@ -32,6 +33,13 @@ pub struct FormatResult {
     pub number_of_files: usize,
 }
 
+#[derive(Debug)]
+pub struct ParseResult {
+    pub parse_duration: Duration,
+    pub duration: Duration,
+    pub number_of_files: usize,
+}
+
 impl Termination for CliRunResult {
     fn report(self) -> ExitCode {
         match self {
@@ -95,6 +103,16 @@ impl Termination for CliRunResult {
                 );
                 ExitCode::from(0)
             }
+            Self::ParseResult(ParseResult { parse_duration, duration, number_of_files }) => {
+                let threads = rayon::current_num_threads();
+                let time = Self::get_execution_time(&duration);
+                let parse_time = Self::get_execution_time(&parse_duration);
+                let s = if number_of_files == 1 { "" } else { "s" };
+                println!(
+                    "Finished in {time} ({parse_time}) on {number_of_files} file{s} using {threads} threads."
+                );
+                ExitCode::from(0)
+            }
             Self::TypeCheckResult { duration, number_of_diagnostics } => {
                 let time = Self::get_execution_time(&duration);
                 println!("Finished in {time}.");

justfile

@@ -1,5 +1,7 @@
 #!/usr/bin/env -S just --justfile
 
+set windows-shell := ["pwsh.exe", "-c"]
+
 _default:
   @just --list -u
 
@@ -140,3 +142,31 @@ upgrade:
 clone-submodule dir url sha:
   git clone --depth=1 {{url}} {{dir}} || true
   cd {{dir}} && git fetch origin {{sha}} && git reset --hard {{sha}}
+
+
+ecosystem_dir := "C:/source/ecosystem"
+oxlint_bin := "C:/source/rust/oxc/target/release/oxparse.exe"
+threads := "12"
+
+pgo_data_dir := "C:/source/rust/oxc/pgo-data"
+llvm_profdata_bin := "~/.rustup/toolchains/1.78.0-x86_64-pc-windows-msvc/lib/rustlib/x86_64-pc-windows-msvc/bin/llvm-profdata.exe"
+
+build-pgo:
+  just build-pgo-init
+  just oxlint_bin=C:/source/rust/oxc/target/x86_64-pc-windows-msvc/release/oxparse.exe ecosystem
+  {{llvm_profdata_bin}} merge -o {{pgo_data_dir}}/merged.profdata {{pgo_data_dir}}
+  just build-pgo-final
+
+build-pgo-init $RUSTFLAGS="-Cprofile-generate=C:/source/rust/oxc/pgo-data":
+  cargo build --release -p oxc_cli --bin oxparse --features allocator --target x86_64-pc-windows-msvc
+
+build-pgo-final $RUSTFLAGS="-Cprofile-use=C:/source/rust/oxc/pgo-data/merged.profdata -Cllvm-args=-pgo-warn-missing-function":
+  cargo build --release -p oxc_cli --bin oxparse --features allocator --target x86_64-pc-windows-msvc
+
+ecosystem:
+  cd "{{ecosystem_dir}}/DefinitelyTyped" && {{oxlint_bin}} --threads={{threads}}
+  cd "{{ecosystem_dir}}/affine" && {{oxlint_bin}} --threads={{threads}}
+  cd "{{ecosystem_dir}}/napi-rs" && {{oxlint_bin}} --threads={{threads}} --ignore-path=.oxlintignore
+  cd "{{ecosystem_dir}}/preact" && {{oxlint_bin}} --threads={{threads}} oxlint src test debug compat hooks test-utils
+  cd "{{ecosystem_dir}}/rolldown" && {{oxlint_bin}} --threads={{threads}} --ignore-path=.oxlintignore
+  cd "{{ecosystem_dir}}/vscode" && {{oxlint_bin}} --threads={{threads}}