perf(core) Reduce script name and script code copies (denoland#18298)

Reduce the number of copies and allocations of script code by carrying
around ownership/reference information from creation time.

As an advantage, this allows us to maintain the identity of `&'static
str`-based scripts and use v8's external 1-byte strings (to avoid
incorrectly passing non-ASCII strings, debug `assert!`s gate all string
reference paths).

Benchmark results:

Perf improvements -- ~0.1 - 0.2ms faster, but should reduce garbage
w/external strings and reduces data copies overall. May also unlock some
more interesting optimizations in the future.

This requires adding some generics to functions, but manual
monomorphization has been applied (outer/inner function) to avoid code
bloat.

bench_util/js_runtime.rs

@@ -103,7 +103,8 @@ pub fn bench_js_async_with(
  opts.benching_inner
  };
  let looped = loop_code(inner_iters, src);
- let src = looped.as_ref();
+ // Get a &'static str by leaking -- this is fine because it's benchmarking code
+ let src = Box::leak(looped.into_boxed_str());
  if is_profiling() {
  for _ in 0..opts.profiling_outer {
  tokio_runtime.block_on(inner_async(src, &mut runtime));
@@ -115,7 +116,7 @@ pub fn bench_js_async_with(
  }
 }
 
-async fn inner_async(src: &str, runtime: &mut JsRuntime) {
+async fn inner_async(src: &'static str, runtime: &mut JsRuntime) {
  runtime.execute_script("inner_loop", src).unwrap();
  runtime.run_event_loop(false).await.unwrap();
 }

cli/emit.rs

@@ -5,6 +5,7 @@ use crate::cache::FastInsecureHasher;
 use crate::cache::ParsedSourceCache;
 
 use deno_core::error::AnyError;
+use deno_core::ModuleCode;
 use deno_core::ModuleSpecifier;
 use deno_graph::MediaType;
 use std::sync::Arc;
@@ -27,11 +28,11 @@ pub fn emit_parsed_source(
  source: &Arc<str>,
  emit_options: &deno_ast::EmitOptions,
  emit_config_hash: u64,
-) -> Result<String, AnyError> {
+) -> Result<ModuleCode, AnyError> {
  let source_hash = get_source_hash(source, emit_config_hash);
 
  if let Some(emit_code) = emit_cache.get_emit_code(specifier, source_hash) {
- Ok(emit_code)
+ Ok(emit_code.into())
  } else {
  // this will use a cached version if it exists
  let parsed_source = parsed_source_cache.get_or_parse_module(
@@ -42,6 +43,6 @@ pub fn emit_parsed_source(
  let transpiled_source = parsed_source.transpile(emit_options)?;
  debug_assert!(transpiled_source.source_map.is_none());
  emit_cache.set_emit_code(specifier, source_hash, &transpiled_source.text);
- Ok(transpiled_source.text)
+ Ok(transpiled_source.text.into())
  }
 }

cli/lsp/tsc.rs

@@ -2854,7 +2854,7 @@ fn start(runtime: &mut JsRuntime, debug: bool) -> Result<(), AnyError> {
  let init_config = json!({ "debug": debug });
  let init_src = format!("globalThis.serverInit({init_config});");
 
- runtime.execute_script(&located_script_name!(), &init_src)?;
+ runtime.execute_script(located_script_name!(), init_src)?;
  Ok(())
 }
 
@@ -3442,7 +3442,7 @@ pub fn request(
  };
  let mark = performance.mark("request", Some(request_params.clone()));
  let request_src = format!("globalThis.serverRequest({request_params});");
- runtime.execute_script(&located_script_name!(), &request_src)?;
+ runtime.execute_script(located_script_name!(), request_src)?;
 
  let op_state = runtime.op_state();
  let mut op_state = op_state.borrow_mut();

cli/module_loader.rs

@@ -14,6 +14,7 @@ use deno_core::error::AnyError;
 use deno_core::futures::future::FutureExt;
 use deno_core::futures::Future;
 use deno_core::resolve_url;
+use deno_core::ModuleCode;
 use deno_core::ModuleLoader;
 use deno_core::ModuleSource;
 use deno_core::ModuleSpecifier;
@@ -30,7 +31,7 @@ use std::rc::Rc;
 use std::str;
 
 struct ModuleCodeSource {
- pub code: String,
+ pub code: ModuleCode,
  pub found_url: ModuleSpecifier,
  pub media_type: MediaType,
 }
@@ -91,7 +92,7 @@ impl CliModuleLoader {
  specifier,
  ..
  })) => Ok(ModuleCodeSource {
- code: source.to_string(),
+ code: source.into(),
  found_url: specifier.clone(),
  media_type: *media_type,
  }),
@@ -101,13 +102,15 @@ impl CliModuleLoader {
  specifier,
  ..
  })) => {
- let code = match media_type {
+ let code: ModuleCode = match media_type {
  MediaType::JavaScript
  | MediaType::Unknown
  | MediaType::Cjs
  | MediaType::Mjs
- | MediaType::Json => source.to_string(),
- MediaType::Dts | MediaType::Dcts | MediaType::Dmts => "".to_string(),
+ | MediaType::Json => source.into(),
+ MediaType::Dts | MediaType::Dcts | MediaType::Dmts => {
+ Default::default()
+ }
  MediaType::TypeScript
  | MediaType::Mts
  | MediaType::Cts
@@ -191,7 +194,7 @@ impl CliModuleLoader {
  )?
  };
  ModuleCodeSource {
- code,
+ code: code.into(),
  found_url: specifier.clone(),
  media_type: MediaType::from_specifier(specifier),
  }
@@ -208,7 +211,7 @@ impl CliModuleLoader {
  code_without_source_map(code_source.code)
  };
  Ok(ModuleSource {
- code: code.into_bytes().into_boxed_slice(),
+ code,
  module_url_specified: specifier.to_string(),
  module_url_found: code_source.found_url.to_string(),
  module_type: match code_source.media_type {

cli/standalone.rs

@@ -178,7 +178,7 @@ impl ModuleLoader for EmbeddedModuleLoader {
  async move {
  if let Some((source, _)) = is_data_uri {
  return Ok(deno_core::ModuleSource {
- code: source.into_bytes().into_boxed_slice(),
+ code: source.into(),
  module_type: deno_core::ModuleType::JavaScript,
  module_url_specified: module_specifier.to_string(),
  module_url_found: module_specifier.to_string(),
@@ -192,7 +192,7 @@ impl ModuleLoader for EmbeddedModuleLoader {
  .to_owned();
 
  Ok(deno_core::ModuleSource {
- code: code.into_bytes().into_boxed_slice(),
+ code: code.into(),
  module_type: match module.kind {
  eszip::ModuleKind::JavaScript => deno_core::ModuleType::JavaScript,
  eszip::ModuleKind::Json => deno_core::ModuleType::Json,
@@ -384,16 +384,16 @@ pub async fn run(
  options,
  );
  worker.execute_main_module(main_module).await?;
- worker.dispatch_load_event(&located_script_name!())?;
+ worker.dispatch_load_event(located_script_name!())?;
 
  loop {
  worker.run_event_loop(false).await?;
- if !worker.dispatch_beforeunload_event(&located_script_name!())? {
+ if !worker.dispatch_beforeunload_event(located_script_name!())? {
  break;
  }
  }
 
- worker.dispatch_unload_event(&located_script_name!())?;
+ worker.dispatch_unload_event(located_script_name!())?;
  std::process::exit(0);
 }
 

cli/tools/coverage/mod.rs

@@ -20,6 +20,7 @@ use deno_core::serde_json;
 use deno_core::sourcemap::SourceMap;
 use deno_core::url::Url;
 use deno_core::LocalInspectorSession;
+use deno_core::ModuleCode;
 use regex::Regex;
 use std::fs;
 use std::fs::File;
@@ -170,16 +171,16 @@ struct CoverageReport {
 
 fn generate_coverage_report(
  script_coverage: &ScriptCoverage,
- script_source: &str,
+ script_source: String,
  maybe_source_map: &Option<Vec<u8>>,
  output: &Option<PathBuf>,
 ) -> CoverageReport {
  let maybe_source_map = maybe_source_map
  .as_ref()
  .map(|source_map| SourceMap::from_slice(source_map).unwrap());
- let text_lines = TextLines::new(script_source);
+ let text_lines = TextLines::new(&script_source);
 
- let comment_ranges = deno_ast::lex(script_source, MediaType::JavaScript)
+ let comment_ranges = deno_ast::lex(&script_source, MediaType::JavaScript)
  .into_iter()
  .filter(|item| {
  matches!(item.inner, deno_ast::TokenOrComment::Comment { .. })
@@ -680,22 +681,22 @@ pub async fn cover_files(
  })?;
 
  // Check if file was transpiled
- let original_source = &file.source;
- let transpiled_code = match file.media_type {
+ let original_source = file.source.clone();
+ let transpiled_code: ModuleCode = match file.media_type {
  MediaType::JavaScript
  | MediaType::Unknown
  | MediaType::Cjs
  | MediaType::Mjs
- | MediaType::Json => file.source.as_ref().to_string(),
- MediaType::Dts | MediaType::Dmts | MediaType::Dcts => "".to_string(),
+ | MediaType::Json => file.source.into(),
+ MediaType::Dts | MediaType::Dmts | MediaType::Dcts => Default::default(),
  MediaType::TypeScript
  | MediaType::Jsx
  | MediaType::Mts
  | MediaType::Cts
  | MediaType::Tsx => {
  let source_hash = get_source_hash(&file.source, ps.emit_options_hash);
  match ps.emit_cache.get_emit_code(&file.specifier, source_hash) {
- Some(code) => code,
+ Some(code) => code.into(),
  None => {
  return Err(anyhow!(
  "Missing transpiled source code for: \"{}\".
@@ -710,15 +711,16 @@ pub async fn cover_files(
  }
  };
 
+ let source_map = source_map_from_code(&transpiled_code);
  let coverage_report = generate_coverage_report(
  &script_coverage,
- &transpiled_code,
- &source_map_from_code(&transpiled_code),
+ transpiled_code.take_as_string(),
+ &source_map,
  &out_mode,
  );
 
  if !coverage_report.found_lines.is_empty() {
- reporter.report(&coverage_report, original_source)?;
+ reporter.report(&coverage_report, &original_source)?;
  }
  }
 

cli/tsc/mod.rs

@@ -824,9 +824,9 @@ pub fn exec(request: Request) -> Result<Response, AnyError> {
  });
 
  runtime
- .execute_script(&located_script_name!(), startup_source)
+ .execute_script(located_script_name!(), startup_source)
  .context("Could not properly start the compiler runtime.")?;
- runtime.execute_script(&located_script_name!(), &exec_source)?;
+ runtime.execute_script(located_script_name!(), exec_source)?;
 
  let op_state = runtime.op_state();
  let mut op_state = op_state.borrow_mut();

cli/util/text_encoding.rs

@@ -1,5 +1,6 @@
 // Copyright 2018-2023 the Deno authors. All rights reserved. MIT license.
 
+use deno_core::ModuleCode;
 use encoding_rs::*;
 use std::borrow::Cow;
 use std::io::Error;
@@ -53,11 +54,12 @@ pub fn strip_bom(text: &str) -> &str {
  }
 }
 
-static SOURCE_MAP_PREFIX: &str =
- "//# sourceMappingURL=data:application/json;base64,";
+static SOURCE_MAP_PREFIX: &[u8] =
+ b"//# sourceMappingURL=data:application/json;base64,";
 
-pub fn source_map_from_code(code: &str) -> Option<Vec<u8>> {
- let last_line = code.rsplit(|u| u == '\n').next()?;
+pub fn source_map_from_code(code: &ModuleCode) -> Option<Vec<u8>> {
+ let bytes = code.as_bytes();
+ let last_line = bytes.rsplit(|u| *u == b'\n').next()?;
  if last_line.starts_with(SOURCE_MAP_PREFIX) {
  let input = last_line.split_at(SOURCE_MAP_PREFIX.len()).1;
  let decoded_map = base64::decode(input)
@@ -68,17 +70,18 @@ pub fn source_map_from_code(code: &str) -> Option<Vec<u8>> {
  }
 }
 
-pub fn code_without_source_map(mut code: String) -> String {
- if let Some(last_line_index) = code.rfind('\n') {
- if code[last_line_index + 1..].starts_with(SOURCE_MAP_PREFIX) {
- code.truncate(last_line_index + 1);
- code
- } else {
- code
+/// Truncate the source code before the source map.
+pub fn code_without_source_map(mut code: ModuleCode) -> ModuleCode {
+ let bytes = code.as_bytes();
+ for i in (0..bytes.len()).rev() {
+ if bytes[i] == b'\n' {
+ if bytes[i + 1..].starts_with(SOURCE_MAP_PREFIX) {
+ code.truncate(i + 1);
+ }
+ return code;
  }
- } else {
- code
  }
+ code
 }
 
 #[cfg(test)]
@@ -155,8 +158,11 @@ mod tests {
  "\n",
  );
 
- fn run_test(input: &str, output: &str) {
- assert_eq!(code_without_source_map(input.to_string()), output);
+ fn run_test(input: &'static str, output: &'static str) {
+ assert_eq!(
+ code_without_source_map(input.into()).take_as_string(),
+ output
+ );
  }
  }
 }