Remove unnecessary ValueDebugFormat item, hide Vc field #5567
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
The latest updates on your projects. Learn more about Vercel for Git ↗︎
10 Ignored Deployments
|
|
✅ This change can build |
|
Linux Benchmark for e40c341Click to view benchmark
|
jridgewell
approved these changes
Jul 19, 2023
sokra
added a commit
to vercel/next.js
that referenced
this pull request
Jul 21, 2023
* vercel/turborepo#5567 <!-- Alex Kirszenberg - Remove unnecessary ValueDebugFormat item, hide Vc field --> * vercel/turborepo#5576 <!-- Alex Kirszenberg - Extract shared HMR utils to their own modules/crates --> * vercel/turborepo#5503 <!-- OJ Kwon - feat(turbopack_core): define trait for diagnostics --> * vercel/turborepo#5487 <!-- Will Binns-Smith - turbopack-cli: modularize code to support turbopack build --> * vercel/turborepo#5488 <!-- Will Binns-Smith - turbopack-cli: implement `turbopack build` --> * vercel/turborepo#5450 <!-- Leah - feat: async modules / top level await --> --------- Co-authored-by: kodiakhq[bot] <49736102+kodiakhq[bot]@users.noreply.github.com>
NicholasLYang
pushed a commit
that referenced
this pull request
Jul 21, 2023
commit 6c178d2 Author: Alex Kirszenberg <alex.kirszenberg@vercel.com> Date: Thu Jul 20 18:06:03 2023 +0200 Extract shared HMR utils to their own modules/crates (#5576) ### Description These will also be used in Next.js by the Nexturbo dev API. ### Testing Instructions N/A commit e5f43a5 Author: Chris Olszewski <chris.olszewski@vercel.com> Date: Wed Jul 19 18:03:12 2023 -0700 fix: pnpm alias workspace deps (#5569) ### Description Fixes #5441 Adds support for [referencing workspaces through aliases](https://pnpm.io/workspaces#referencing-workspace-packages-through-aliases) by properly resolving them to the correct workspace. Before we would mark a package as being an external dependency (or if the alias was a valid workspace depend on the incorrect one). This PR now recognizes when `workspace:` dependency references a different package than the name that's used in the `package.json`. Note for reviewers: This probably isn't the cleanest solution in either Rust or Go, but while we need to maintain two codepaths this keeps the code roughly equivalent. ### Testing Instructions Added unit tests on the Go side. Tested manually with a repository where `web` specified it's dependency on `@scope/ui` as `"ui": "workspace:@scope/ui@*" and verified that: - `turbo run build`: `@scope/ui` finished building before building `web`, this hits the Go impl - `turbo prune --scope=web`: `@scope/ui` was included in the pruned repository, this hits the Rust impl --------- Co-authored-by: Chris Olszewski <Chris Olszewski> commit 396bf45 Author: Chris Olszewski <chris.olszewski@vercel.com> Date: Wed Jul 19 13:31:35 2023 -0700 feat: port prune to rust (#5531) ### Description Port prune from Go to Rust and remove the old Go implementation. To achieve this the following was done: - Parsing the lockfile during package graph construction - Porting utility functions - Some minor changes in `turborepo_paths`, these were primarily moving methods from their owned to borrowed counterparts - Expanding `package.json` parsing to grab some of the information pruning requires - Porting of the prune command itself. I added some additional structure compared to the Go version, but not enough that comparing it to the Go version should be difficult Notes for reviewers: I apologize that this PR ended up touching as much as it did. Reviewing the PR by commit should at least make all of the changes and their impacts obvious. Commits before `2d4154c` are already on main and can be skipped. ### Testing Instructions Existing unit tests and integration tests for file copying and package graph traversal. Actual lockfile behavior is mostly covered by unit tests that were ported when the lockfile were ported. Also did various manual testing with pruning monorepos. --------- Co-authored-by: Chris Olszewski <Chris Olszewski> commit 007b574 Author: Alex Kirszenberg <alex.kirszenberg@vercel.com> Date: Wed Jul 19 22:17:39 2023 +0200 Remove unnecessary ValueDebugFormat item, hide Vc field (#5567) ### Description This removes unnecessary items from the IDE's autosuggestion. I'd also like to get rid of all the `*_inline` suggestions, but RA will ignore `#[doc(hidden)]` when inside the same crate. I don't think there's a way to indicate "never, ever suggest this item". We could move some of these to be private to some generated module, but that module would probably still show up as a top level suggestion. ### Testing Instructions Automated tests. commit d13b812 Author: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com> Date: Wed Jul 19 11:32:07 2023 -0700 release(turborepo): 1.10.9 (#5565) Co-authored-by: Turbobot <turbobot@vercel.com> Co-authored-by: Chris Olszewski <Chris Olszewski> commit 229a2a4 Author: Tobias Koppers <tobias.koppers@googlemail.com> Date: Wed Jul 19 16:08:16 2023 +0200 make with_layer return Vc<Self> (#5563) ### Description Using the new ability to return `Vc<Self>` Co-authored-by: Alex Kirszenberg <alexkirsz@users.noreply.github.com> commit c78593b Author: Greg Soltis <greg.soltis@vercel.com> Date: Tue Jul 18 20:11:16 2023 -0700 feat(turborepo): Add proxy support to create-turbo and turbo-gen (#5554) Co-authored-by: Greg Soltis <Greg Soltis> commit 5ab8ac0 Author: Chris Olszewski <chris.olszewski@vercel.com> Date: Tue Jul 18 15:06:38 2023 -0700 fix(lockfile): Fix directory resolution variant (#5551) ### Description Fixes #5529 During the Rust migration I must've messed up the directory field name. Double checked against [`@pnpm/lockfile-types`](https://github.com/pnpm/pnpm/blob/main/lockfile/lockfile-types/src/index.ts#L86) to make sure all of the fields are correct now. `PackageResolution` should be an enum, but the fact that tarballs are an untagged variant makes that tricky to communicate to `serde`. A struct does enough for us. ### Testing Instructions Added new unit test to make sure we don't lose any fields for the various variants of the `resolution` field --------- Co-authored-by: Chris Olszewski <Chris Olszewski> commit ca7e3e4 Author: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com> Date: Tue Jul 18 14:51:34 2023 -0700 release(turborepo): 1.10.9-canary.0 (#5559) Co-authored-by: Turbobot <turbobot@vercel.com> commit fc5e2b0 Author: Chris Olszewski <chris.olszewski@vercel.com> Date: Tue Jul 18 13:16:16 2023 -0700 chore: no longer convert in relative unix path constructor (#5552) ### Description To quote @gsoltis: > In general: > - constructors should validate, to the extent they can (not much for `unix` paths, but can verify relative) > - conversions should be explicit. You need to know where you're starting from. If this were an AnchoredSystemPath on windows, the `\` -> `/` makes sense. If it's a literal from e.g. a tar file, then it doesn't. Reviewers Notes: - Opening up this PR in VSCode and using `Find All References` on the constructor is useful for double checking that I didn't miss a conversion. - Clippy error appeared on local when I made these changes. Fixed it just in case that would block CI - Moved `to_unix` to `AnchoredSystemPath` instead of `AnchoredSystemPathBuf` now that we have deref coercion which will automatically convert `&AnchroedSystemPathBuf` to `&AnchoredSystemPath` and moving the method allows it to be called from either type. ### Testing Instructions Looked through all uses of `RelativeUnixPathBuf::new` to see if there were places that depended on the conversion. The only use that was obvious was the usage in `cache_archive/create.rs`. `dotEnv` was the only other place where we possibly were converting a system path to a relative unix. We don't specify that `dotEnv` entries should be unix relative, so we might've been accidentally supporting system paths, but --------- Co-authored-by: Chris Olszewski <Chris Olszewski> commit 3eb3a5f Author: Tobias Koppers <tobias.koppers@googlemail.com> Date: Tue Jul 18 21:21:51 2023 +0200 Ensure output assets reference only output assets (#5557) ### Description Ensure output assets reference only output assets next.js PR: vercel/next.js#52832 commit 22f0bf4 Author: Nathan Hammond <nathan.hammond@vercel.com> Date: Wed Jul 19 02:59:12 2023 +0800 Remove binary optimization step. (#5543) Co-authored-by: Nathan Hammond <Nathan Hammond> commit 5955625 Author: mknichel <7355009+mknichel@users.noreply.github.com> Date: Tue Jul 18 10:09:28 2023 -0700 fix(turborepo): Allow users to select a Vercel team when linking a repository to a Space (#5533) commit 814f0a2 Author: Leah <github.leah@hrmny.sh> Date: Tue Jul 18 17:31:33 2023 +0200 feat(turbopack-ecmascript): implement acyclic SCC graph for ESM imports (#5506) commit 46bb9b7 Author: Tobias Koppers <tobias.koppers@googlemail.com> Date: Tue Jul 18 13:41:01 2023 +0200 move references() to specific traits (#5555) ### Description preparation for making different typed references next.js PR: vercel/next.js#52822 commit 56edd9e Author: Greg Soltis <greg.soltis@vercel.com> Date: Mon Jul 17 21:37:26 2023 -0700 feat(turborepo): Add task ids to failure reports (#5535) Co-authored-by: Greg Soltis <Greg Soltis> commit ea934d1 Author: Greg Soltis <greg.soltis@vercel.com> Date: Mon Jul 17 13:52:59 2023 -0700 chore(turborepo): Turborepo owns the examples-tests directory (#5540) Co-authored-by: Greg Soltis <Greg Soltis> commit 5279282 Author: Alex Kirszenberg <alex.kirszenberg@vercel.com> Date: Mon Jul 17 19:55:22 2023 +0200 Add missing feature to syn (#5547) ### Description Cargo check currently fails with: ``` error[E0277]: `syn::Type` doesn't implement `Debug` --> crates/turbo-tasks-macros-shared/src/primitive_input.rs:8:5 | 6 | #[derive(Debug)] | ----- in this derive macro expansion 7 | pub struct PrimitiveInput { 8 | pub ty: Type, | ^^^^^^^^^^^^ `syn::Type` cannot be formatted using `{:?}` because it doesn't implement `Debug` | = help: the trait `Debug` is not implemented for `syn::Type` = note: this error originates in the derive macro `Debug` (in Nightly builds, run with -Z macro-backtrace for more info) For more information about this error, try `rustc --explain E0277`. error: could not compile `turbo-tasks-macros-shared` (lib) due to previous error ``` This is because because turbo-tasks-macros-shared (which is missing a syn feature to enable the debug trait) is part of the workspace default members, while turbo-tasks-macros (which has the feature) isn’t. ### Testing Instructions `cargo check` commit 78f6cc8 Author: Chris Olszewski <chris.olszewski@vercel.com> Date: Mon Jul 17 10:40:13 2023 -0700 docs: document log order (#5463) ### Description Adds documentation for `--log-order` #3916 including a callout of our special behavior on Github Actions. ### Testing Instructions Eyes --------- Co-authored-by: Chris Olszewski <Chris Olszewski> commit 0771b80 Author: Amit Gurbani <amit1994.gurbani@gmail.com> Date: Mon Jul 17 23:09:44 2023 +0530 Update storybook.mdx (#5537) ### Description Storybook now can be built with node 18, hence removing this from documentation. Co-authored-by: Anthony Shew <anthony.shew@vercel.com> commit 91ca2ae Author: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com> Date: Mon Jul 17 10:11:06 2023 -0700 release(turborepo): 1.10.8 (#5546) Co-authored-by: Turbobot <turbobot@vercel.com> commit d24075f Author: Greg Soltis <greg.soltis@vercel.com> Date: Mon Jul 17 08:12:38 2023 -0700 fix(turborepo): Set corepack dir for examples (#5539) ### Description - set a corepack install directory per example to avoid concurrency issues - set the `PATH` for the test to include the corepack directory ### Testing Instructions Examples tests --------- Co-authored-by: Greg Soltis <Greg Soltis> commit 46d0945 Author: Nicholas Yang <nicholas.yang@vercel.com> Date: Mon Jul 17 11:07:45 2023 -0400 feat(turborepo): FS Cache (#5473) ### Description Implements the FS cache on top of CacheItem. ~~This is stacked on top of #5065~~ ### Testing Instructions Uses the same round-trip tests of HTTP cache. --------- Co-authored-by: --global <Nicholas Yang> Co-authored-by: Chris Olszewski <chris.olszewski@vercel.com> commit 135c08f Author: Tobias Koppers <tobias.koppers@googlemail.com> Date: Mon Jul 17 16:44:56 2023 +0200 add direct cycle detection (#5544) ### Description very simple detection of dumb mistakes commit 02f55d9 Author: Tobias Koppers <tobias.koppers@googlemail.com> Date: Mon Jul 17 16:33:34 2023 +0200 move Asset::ident to more specific traits (#5528) ### Description * `ident()` is no longer on Asset, but on `Module`, `Source`, `OutputAsset` or `Chunk` * On the way, more AssetVc types needed to be switched to more specific traits next.js PR: vercel/next.js#52683 commit b069545 Author: Alex Kirszenberg <alex.kirszenberg@vercel.com> Date: Mon Jul 17 13:41:18 2023 +0200 AdjacencyMap::reverse_topological (+ fixes) (#5527) ### Description This adds `AdjacencyMap::reverse_topological`, which is similar to `AdjacencyMap::into_reverse_topological` but doesn't consume the graph. This also: * Makes `AdjacencyMap` storable in `turbo_tasks::value`s; * Fixes ValueDebugFormat and TraceRawVcs derive macros so they support generic argument and bounds properly. ### Testing Instructions N/A commit 8433a32 Author: Alex Kirszenberg <alex.kirszenberg@vercel.com> Date: Sun Jul 16 14:07:49 2023 +0200 Vc<T> and Turbo Engine type system improvements (#4587) This PR changes our Turbo Engine code generation from generating additional `TypeVc` types to allowing the `Vc<Type>` notation. It also brings other improvements to the Turbo Engine type system, like more type-safe downcasting and upcasting, better support for primitives, the possibility to accept `&self` in `#[value_impl]` implementations everywhere, and a bunch of other changes. link WEB-379 commit fdc358a Author: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com> Date: Fri Jul 14 20:31:07 2023 -0700 release(turborepo): 1.10.8-canary.2 (#5534) Co-authored-by: Turbobot <turbobot@vercel.com> commit 26fee25 Author: Greg Soltis <greg.soltis@vercel.com> Date: Fri Jul 14 11:13:59 2023 -0700 fix(turborepo): Export and match on our copy of BasicUI (#5532) Co-authored-by: Greg Soltis <Greg Soltis> commit b0ea0a8 Author: Greg Soltis <greg.soltis@vercel.com> Date: Fri Jul 14 10:39:52 2023 -0700 fix(turborepo): Rebuild turbo if Go code has changed (#5530) Co-authored-by: Greg Soltis <Greg Soltis> commit b6bb8fe Author: Chris Olszewski <chris.olszewski@vercel.com> Date: Fri Jul 14 06:47:13 2023 -0700 chore: use fs-err in turborepo fs related libs (#5517) ### Description Swaps our usage of various `fs` methods to use `fs-err` instead. To quote the `fs-err` docs: > Using [std::fs](https://doc.rust-lang.org/stable/std/fs/), if this code fails: > > `let file = File::open("does not exist.txt")?;` > > The error message that Rust gives you isn't very useful: > > `The system cannot find the file specified. (os error 2)` > > ...but if we use `fs-err` instead, our error contains more actionable information: > > ```failed to open file `does not exist.txt` > caused by: The system cannot find the file specified. (os error 2)``` ### Testing Instructions Existing unit tests pass Co-authored-by: Chris Olszewski <Chris Olszewski> commit f3a36e7 Author: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com> Date: Thu Jul 13 15:27:21 2023 -0700 release(turborepo): 1.10.8-canary.1 (#5523) Co-authored-by: Turbobot <turbobot@vercel.com> commit cabca3c Author: Nicholas Yang <nicholas.yang@vercel.com> Date: Thu Jul 13 16:12:46 2023 -0400 chore: Extend examples timeout (#5504) ### Description We get a lot of timeout failures on examples tests. Since this almost always requires re-running the test, which adds another 30 minutes of compute time, it's probably better to extend the timeout to 40 minutes. ### Testing Instructions <!-- Give a quick description of steps to test your changes. --> Co-authored-by: nicholaslyang <Nicholas Yang> commit 7c18c80 Author: Greg Soltis <greg.soltis@vercel.com> Date: Thu Jul 13 12:58:14 2023 -0700 Implement hashing fallback (#5505) Co-authored-by: Greg Soltis <Greg Soltis> commit 67e71c7 Author: Anthony Shew <anthony.shew@vercel.com> Date: Thu Jul 13 11:35:08 2023 -0700 Better spot for link. (#5520) commit a14180e Author: Anthony Shew <anthony.shew@vercel.com> Date: Thu Jul 13 11:24:26 2023 -0700 Fix link. (#5518) commit b7aaa7b Author: Nicholas Yang <nicholas.yang@vercel.com> Date: Thu Jul 13 13:18:44 2023 -0400 chore: Added clippy deny all to crates (#5514) Co-authored-by: nicholaslyang <Nicholas Yang> commit e3c68fa Author: Alex Kirszenberg <alex.kirszenberg@vercel.com> Date: Thu Jul 13 16:34:27 2023 +0200 Add any_content_changed_of_output_assets (#5513) ### Description Required for vercel/next.js#52259 ### Testing Instructions N/A Co-authored-by: Tobias Koppers <sokra@users.noreply.github.com> commit 4022f2b Author: Tobias Koppers <tobias.koppers@googlemail.com> Date: Thu Jul 13 14:54:55 2023 +0200 OutputAsset trait (#5507) ### Description adding a trait to all assets in the output graph next.js PR: vercel/next.js#52606
ForsakenHarmony
pushed a commit
to vercel/next.js
that referenced
this pull request
Jul 25, 2024
…repo#5567) ### Description This removes unnecessary items from the IDE's autosuggestion. I'd also like to get rid of all the `*_inline` suggestions, but RA will ignore `#[doc(hidden)]` when inside the same crate. I don't think there's a way to indicate "never, ever suggest this item". We could move some of these to be private to some generated module, but that module would probably still show up as a top level suggestion. ### Testing Instructions Automated tests.
ForsakenHarmony
pushed a commit
to vercel/next.js
that referenced
this pull request
Jul 29, 2024
…repo#5567) ### Description This removes unnecessary items from the IDE's autosuggestion. I'd also like to get rid of all the `*_inline` suggestions, but RA will ignore `#[doc(hidden)]` when inside the same crate. I don't think there's a way to indicate "never, ever suggest this item". We could move some of these to be private to some generated module, but that module would probably still show up as a top level suggestion. ### Testing Instructions Automated tests.
ForsakenHarmony
pushed a commit
to vercel/next.js
that referenced
this pull request
Aug 1, 2024
…repo#5567) ### Description This removes unnecessary items from the IDE's autosuggestion. I'd also like to get rid of all the `*_inline` suggestions, but RA will ignore `#[doc(hidden)]` when inside the same crate. I don't think there's a way to indicate "never, ever suggest this item". We could move some of these to be private to some generated module, but that module would probably still show up as a top level suggestion. ### Testing Instructions Automated tests.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Description
This removes unnecessary items from the IDE's autosuggestion.
I'd also like to get rid of all the
*_inlinesuggestions, but RA will ignore#[doc(hidden)]when inside the same crate. I don't think there's a way to indicate "never, ever suggest this item". We could move some of these to be private to some generated module, but that module would probably still show up as a top level suggestion.Testing Instructions
Automated tests.