Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

feat(compiler): perform automatic key insertion in more situations #5594

Merged
merged 1 commit into from Apr 5, 2024

Conversation

alicewriteswrongs
Copy link
Member

@alicewriteswrongs alicewriteswrongs commented Mar 26, 2024

This expands the scope of the automatic key insertion feature that was added in #5143. Previously the feature stopped at the border of any JsxExpression syntax node (a bit of arbitrary JavaScript wrapped in curly braces and then inserted into JSX) so that if you were doing something like:

<div>{ someCondition && <span>when condition is true</span>}</div>

the compiler would insert a key into the <div> syntax tree node but not the node for the <span>. We did this to just be a bit cautious with the feature because there are a lot of different things that could be going on within a JsxExpression node, and some things which could be written within curly braces which would not be safe to transform, such as the following:

<div>{ xs.map(x => <span>{ x }</span>) }</div>

We wouldn't want to insert a key into the <span> syntax tree node because that would result in the dynamically generated <span> tags always having the same key attributes.

That said, there are some situations where it is safe to insert a key, such as the condition || <some-tag /> case shown above. This change adds support for inserting keys in those situations.

STENCIL-1120

What is the current behavior?

GitHub Issue Number: N/A

What is the new behavior?

Documentation

Does this introduce a breaking change?

  • Yes
  • No

Testing

Other information

Copy link
Contributor

github-actions bot commented Mar 26, 2024

--strictNullChecks error report

Typechecking with --strictNullChecks resulted in 1138 errors on this branch.

That's the same number of errors on main, so at least we're not creating new ones!

reports and statistics

Our most error-prone files
Path Error Count
src/dev-server/index.ts 37
src/dev-server/server-process.ts 32
src/compiler/prerender/prerender-main.ts 22
src/testing/puppeteer/puppeteer-element.ts 21
src/runtime/client-hydrate.ts 20
src/screenshot/connector-base.ts 19
src/runtime/vdom/vdom-render.ts 17
src/dev-server/request-handler.ts 15
src/compiler/prerender/prerender-optimize.ts 14
src/compiler/sys/stencil-sys.ts 14
src/sys/node/node-sys.ts 14
src/compiler/prerender/prerender-queue.ts 13
src/compiler/sys/in-memory-fs.ts 13
src/runtime/connected-callback.ts 13
src/runtime/set-value.ts 13
src/compiler/output-targets/output-www.ts 12
src/compiler/transformers/test/parse-vdom.spec.ts 12
src/compiler/transformers/transform-utils.ts 12
src/compiler/transpile/transpile-module.ts 12
src/mock-doc/test/attribute.spec.ts 12
Our most common errors
Typescript Error Code Count
TS2322 361
TS2345 345
TS18048 204
TS18047 82
TS2722 37
TS2532 24
TS2531 21
TS2454 14
TS2790 11
TS2352 9
TS2769 8
TS2538 8
TS2416 7
TS2493 3
TS18046 2
TS2684 1
TS2430 1

Unused exports report

There are 14 unused exports on this PR. That's the same number of errors on main, so at least we're not creating new ones!

Unused exports
File Line Identifier
src/runtime/bootstrap-lazy.ts 21 setNonce
src/screenshot/screenshot-fs.ts 18 readScreenshotData
src/testing/testing-utils.ts 198 withSilentWarn
src/utils/index.ts 145 CUSTOM
src/utils/index.ts 269 normalize
src/utils/index.ts 7 escapeRegExpSpecialCharacters
src/compiler/app-core/app-data.ts 25 BUILD
src/compiler/app-core/app-data.ts 115 Env
src/compiler/app-core/app-data.ts 117 NAMESPACE
src/compiler/fs-watch/fs-watch-rebuild.ts 123 updateCacheFromRebuild
src/compiler/types/validate-primary-package-output-target.ts 61 satisfies
src/compiler/types/validate-primary-package-output-target.ts 61 Record
src/testing/puppeteer/puppeteer-declarations.ts 485 WaitForEventOptions
src/compiler/sys/fetch/write-fetch-success.ts 7 writeFetchSuccessSync

Copy link
Contributor

github-actions bot commented Mar 26, 2024

PR built and packed!

Download the tarball here: https://github.com/ionic-team/stencil/actions/runs/8571121509/artifacts/1388578614

If your browser saves files to ~/Downloads you can install it like so:

unzip -d ~/Downloads ~/Downloads/stencil-core-4.14.1-dev.1712326253.ce157a6.tgz.zip && npm install ~/Downloads/stencil-core-4.14.1-dev.1712326253.ce157a6.tgz

@alicewriteswrongs alicewriteswrongs marked this pull request as ready for review March 26, 2024 20:00
Copy link
Member

@rwaskiewicz rwaskiewicz left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

All in all, looks/works good! Just a couple minor asks

// we're going to encounter the same problem here that we encounter with
// multiple return statements, so we don't try to transform the arms of
// the conditional.
if (ts.isCallExpression(node) || ts.isPropertyAccessExpression(node) || ts.isConditionalExpression(node)) {
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

For isPropertyAccessExpression, our tests here still all pass if I remove that part of the conditional - can we add a test for cases where isPropertyAccessExpression is truthy?

// multiple return statements, so we don't try to transform the arms of
// the conditional.
if (ts.isCallExpression(node) || ts.isPropertyAccessExpression(node) || ts.isConditionalExpression(node)) {
// we're going to encounter the same problems here that we encounter with
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Since we've change this conditional to remove isJsxExpression and add isCallExpression+isPropertyAccessExpression, I think it might make sense to update this comment. I remember you doing this original work, but I think some of the context of this comment is a little lost on me now (TBH, I don't remember what the multiple return statement problem is 😆).

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

yeah, I think that actually the isPropertyAccessExpression check isn't necessary overall, I was thinking of situations like

<div>{ foo.bar(<div />) }</div>

where we might not want to transform the inner <div /> because we don't know what will happen to it (.bar could do anything!) but actually in this case the whole expression foo.bar(<div />) is a CallExpression where the callee is a PropertyAccessExpression. So just checking for isPropertyAccessExpression would only guard against situations like

<div>{ foo.bar }</div>

where the property access expression can't have any children that are JSX nodes anyway...

so anyway yeah I think I will remove it and then fix up the comments

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I updated the code here and also fleshed out a comment in the findRenderMethodVisitor helper function to explain a bit more the 'multiple return' problem, lmk what you think!

Copy link
Member

@rwaskiewicz rwaskiewicz left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM - one non-blocking nit. Comments make sense 🙂

src/compiler/transformers/automatic-key-insertion/index.ts Outdated Show resolved Hide resolved
@rwaskiewicz
Copy link
Member

Actually, @alicewriteswrongs can we hold off on merging this for today? I wanna see if I can get a patch release out to unblock the Lisbon team

@rwaskiewicz
Copy link
Member

v4.14.1 went out this morning, once this has another ✔️ it should be good to merge. Thanks for holding off on the merge!

This expands the scope of the automatic key insertion feature that was
added in #5143. Previously the feature stopped at the border of any
`JsxExpression` syntax node (a bit of arbitrary JavaScript wrapped in
curly braces and then inserted into JSX) so that if you were doing
something like:

```tsx
<div>{ someCondition && <span>when condition is true</span>}</div>
```

the compiler would insert a key into the `<div>` syntax tree node but
_not_ the node for the `<span>`. We did this to just be a bit cautious
with the feature because there are a lot of different things that could
be going on within a `JsxExpression` node, and some things which could
be written within curly braces which would not be safe to transform,
such as the following:

```tsx
<div>{ xs.map(x => <span>{ x }</span>) }</div>
```

We wouldn't want to insert a key into the `<span>` syntax tree node
because that would result in the dynamically generated `<span>` tags
always having the same key attributes.

That said, there are some situations where it is safe to insert a key,
such as the `condition || <some-tag />` case shown above. This change
adds support for inserting keys in those situations.

STENCIL-1120
@alicewriteswrongs alicewriteswrongs added this pull request to the merge queue Apr 5, 2024
Merged via the queue into main with commit 8ee071b Apr 5, 2024
121 checks passed
@alicewriteswrongs alicewriteswrongs deleted the ap/key-insertion-refinement branch April 5, 2024 14:27
@christian-bromann christian-bromann mentioned this pull request May 15, 2024
3 tasks
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

None yet

3 participants