Skip to content

Commit

Permalink
Fix async loader modules in graph traversal (#69077)
Browse files Browse the repository at this point in the history
<!-- Thanks for opening a PR! Your contribution is much appreciated.
To make sure your PR is handled as smoothly as possible we request that you follow the checklist sections below.
Choose the right checklist for the change(s) that you're making:

## For Contributors

### Improving Documentation

- Run `pnpm prettier-fix` to fix formatting issues before opening the PR.
- Read the Docs Contribution Guide to ensure your contribution follows the docs guidelines: https://nextjs.org/docs/community/contribution-guide

### Adding or Updating Examples

- The "examples guidelines" are followed from our contributing doc https://github.com/vercel/next.js/blob/canary/contributing/examples/adding-examples.md
- Make sure the linting passes by running `pnpm build && pnpm lint`. See https://github.com/vercel/next.js/blob/canary/contributing/repository/linting.md

### Fixing a bug

- Related issues linked using `fixes #number`
- Tests added. See: https://github.com/vercel/next.js/blob/canary/contributing/core/testing.md#writing-tests-for-nextjs
- Errors have a helpful link attached, see https://github.com/vercel/next.js/blob/canary/contributing.md

### Adding a feature

- Implements an existing feature request or RFC. Make sure the feature request has been accepted for implementation before opening a PR. (A discussion must be opened, see https://github.com/vercel/next.js/discussions/new?category=ideas)
- Related issues/discussions are linked using `fixes #number`
- e2e tests added (https://github.com/vercel/next.js/blob/canary/contributing/core/testing.md#writing-tests-for-nextjs)
- Documentation added
- Telemetry added. In case of a feature if it's used or not.
- Errors have a helpful link attached, see https://github.com/vercel/next.js/blob/canary/contributing.md


## For Maintainers

- Minimal description (aim for explaining to someone not on the team to understand the PR)
- When linking to a Slack thread, you might want to share details of the conclusion
- Link both the Linear (Fixes NEXT-xxx) and the GitHub issues
- Add review comments if necessary to explain to the reviewer the logic behind a change

### What?

### Why?

### How?

Closes NEXT-
Fixes #

-->
  • Loading branch information
lichu acuña authored Aug 27, 2024
1 parent c077d9e commit 91b979f
Showing 1 changed file with 117 additions and 17 deletions.
134 changes: 117 additions & 17 deletions turbopack/crates/turbopack-ecmascript/src/global_module_id_strategy.rs
Original file line number Diff line number Diff line change
Expand Up @@ -3,15 +3,17 @@ use std::collections::{HashMap, HashSet};
use anyhow::Result;
use turbo_tasks::{
graph::{AdjacencyMap, GraphTraversal},
RcStr, ValueToString, Vc,
RcStr, TryJoinIterExt, ValueToString, Vc,
};
use turbo_tasks_hash::hash_xxh3_hash64;
use turbopack_core::{
chunk::ModuleId,
module::{Module, Modules},
reference::primary_referenced_modules,
reference::ModuleReference,
};

use crate::references::esm::EsmAsyncAssetReference;

#[turbo_tasks::value]
pub struct PreprocessedChildrenIdents {
// ident.to_string() -> full hash
Expand All @@ -20,12 +22,92 @@ pub struct PreprocessedChildrenIdents {
modules_idents: HashMap<RcStr, u64>,
}

#[derive(Clone, Hash)]
#[turbo_tasks::value(shared)]
pub enum ReferencedModule {
Module(Vc<Box<dyn Module>>),
AsyncLoaderModule(Vc<Box<dyn Module>>),
}

impl ReferencedModule {
fn module(&self) -> Vc<Box<dyn Module>> {
match *self {
ReferencedModule::Module(module) => module,
ReferencedModule::AsyncLoaderModule(module) => module,
}
}
}

// TODO(LichuAcu): Reduce type complexity
#[allow(clippy::type_complexity)]
type ModulesAndAsyncLoaders = Vec<(Vec<Vc<Box<dyn Module>>>, Option<Vc<Box<dyn Module>>>)>;

#[turbo_tasks::value(transparent)]
pub struct ReferencedModules(Vec<Vc<ReferencedModule>>);

#[turbo_tasks::function]
async fn referenced_modules(module: Vc<Box<dyn Module>>) -> Result<Vc<ReferencedModules>> {
let references = module.references().await?;

let mut set = HashSet::new();
let modules_and_async_loaders: ModulesAndAsyncLoaders = references
.iter()
.map(|reference| async move {
let async_loader =
if Vc::try_resolve_downcast_type::<EsmAsyncAssetReference>(*reference)
.await?
.is_some()
{
*reference
.resolve_reference()
.resolve()
.await?
.first_module()
.await?
} else {
None
};

let modules = reference
.resolve_reference()
.resolve()
.await?
.primary_modules()
.await?
.clone_value();

Ok((modules, async_loader))
})
.try_join()
.await?;

let mut modules = Vec::new();

for (module_list, async_loader) in modules_and_async_loaders {
for module in module_list {
if set.insert(module) {
modules.push(ReferencedModule::Module(module).cell());
}
}
if let Some(async_loader_module) = async_loader {
if set.insert(async_loader_module) {
modules.push(ReferencedModule::AsyncLoaderModule(async_loader_module).cell());
}
}
}

Ok(Vc::cell(modules))
}

pub async fn get_children_modules(
parent: Vc<Box<dyn Module>>,
) -> Result<impl Iterator<Item = Vc<Box<dyn Module>>> + Send> {
let mut primary_modules = primary_referenced_modules(parent).await?.clone_value();
primary_modules.extend(parent.additional_layers_modules().await?.clone_value());
Ok(primary_modules.into_iter())
parent: Vc<ReferencedModule>,
) -> Result<impl Iterator<Item = Vc<ReferencedModule>> + Send> {
let parent_module = parent.await?.module();
let mut modules = referenced_modules(parent_module).await?.clone_value();
for module in parent_module.additional_layers_modules().await? {
modules.push(ReferencedModule::Module(*module).cell());
}
Ok(modules.into_iter())
}

// NOTE(LichuAcu) Called on endpoint.root_modules(). It would probably be better if this was called
Expand All @@ -38,20 +120,38 @@ pub async fn children_modules_idents(
) -> Result<Vc<PreprocessedChildrenIdents>> {
let children_modules_iter = AdjacencyMap::new()
.skip_duplicates()
.visit(root_modules.await?.iter().copied(), get_children_modules)
.visit(
root_modules
.await?
.iter()
.map(|module| ReferencedModule::Module(*module).cell())
.collect::<Vec<_>>(),
get_children_modules,
)
.await
.completed()?
.into_inner()
.into_reverse_topological();

// module_id -> full hash
let mut modules_idents = HashMap::new();

for module in children_modules_iter {
let module_ident = module.ident();
let ident_str = module_ident.to_string().await?.clone_value();
let hash = hash_xxh3_hash64(&ident_str);
modules_idents.insert(ident_str, hash);
for child_module in children_modules_iter {
match *child_module.await? {
ReferencedModule::Module(module) => {
let module_ident = module.ident();
let ident_str = module_ident.to_string().await?.clone_value();
let hash = hash_xxh3_hash64(&ident_str);
modules_idents.insert(ident_str, hash);
}
ReferencedModule::AsyncLoaderModule(async_loader_module) => {
let async_loader_ident = async_loader_module
.ident()
.with_modifier(Vc::cell("async loader".into()));
let ident_str = async_loader_ident.to_string().await?.clone_value();
let hash = hash_xxh3_hash64(&ident_str);
modules_idents.insert(ident_str, hash);
}
}
}

Ok(PreprocessedChildrenIdents { modules_idents }.cell())
Expand All @@ -60,13 +160,13 @@ pub async fn children_modules_idents(
// Note(LichuAcu): This could be split into two functions: one that merges the preprocessed module
// ids and another that generates the final, optimized module ids. Thoughts?
pub async fn merge_preprocessed_module_ids(
prepared_module_ids: Vec<Vc<PreprocessedChildrenIdents>>,
preprocessed_module_ids: Vec<Vc<PreprocessedChildrenIdents>>,
) -> Result<HashMap<RcStr, Vc<ModuleId>>> {
let mut module_id_map: HashMap<RcStr, Vc<ModuleId>> = HashMap::new();
let mut used_ids: HashSet<u64> = HashSet::new();

for prepared_module_ids in prepared_module_ids {
for (module_ident, full_hash) in prepared_module_ids.await?.modules_idents.iter() {
for preprocessed_module_ids in preprocessed_module_ids {
for (module_ident, full_hash) in preprocessed_module_ids.await?.modules_idents.iter() {
process_module(
module_ident.clone(),
*full_hash,
Expand Down

0 comments on commit 91b979f

Please sign in to comment.