11404 Commits (a4f0ce17cecb3be4f6d7b31b43def4e22df60984)

 

All Branches   

Search

Author	SHA1	Message	Date
paoloricciuti	a4f0ce17ce	feat: accept `foreign` object in custom renderer to determine how to interleave custom renderers	1 month ago
paoloricciuti	030c665a6f	fix: store `segments` and use them to remove nodes when interleaving is happening	1 month ago
paoloricciuti	fe31c69943	chore: create error for missing foreign	1 month ago
paoloricciuti	235fb213cb	fix: push renderer before executing a snippet	1 month ago
paoloricciuti	36ccbe289e	fix: wrap implicit children snippets in `renderer_snippet`	1 month ago
paoloricciuti	2945f67653	fix: set null renderer for snippets created by DOM components	1 month ago
paoloricciuti	4accea6377	fix: make sure pop renderer is always before `return`	1 month ago
paoloricciuti	9489461319	fix: always validate snippets if custom renderer is enabled	1 month ago
paoloricciuti	3eb795c80f	fix: store and restore parent renderer in effect	1 month ago
paoloricciuti	7e58da3433	fix: skip single node optimization in custom renderer	1 month ago
Paolo Ricciuti	f915aa39cf	Merge branch 'main' into svelte-custom-renderer	1 month ago
paoloricciuti	cd5e2ee6a5	fix: wrap implicit children into `renderer_snippet`	1 month ago
Simon H	3d5c4ab620	fix: prevent false-positive reactivity loss warning (#18373) ... By checking that current/previous batch are null we filter out false positives. `reactivity_loss_tracker` is only reset after a microtask, so if a flush happens before that, we get warnings for things we shouldn't warn on. Fixes #18370 --------- Co-authored-by: Rich Harris <rich.harris@vercel.com>	1 month ago
Simon H	bdb1a0f847	fix: ensure async block assigns correct nodes to effect (#18371) ... If a block has a sole component child, we apply an optimization to not need a template. But if that sole child is wrapped in `$.async`, it can happens that the block's effect is assigned the wrong end node. In the reproduction this happens because we have two components inside a Child which is rendered only after an async block resolves. In `$.append` we have logic to not reassign nodes when the active effect didn't already run, which it has by the time we come across it (because the async work had to resolve first; we added this in https://github.com/sveltejs/svelte/pull/17120 to prevent the opposite, nodes being "rewinded" to an earlier position) and so the second sibling component's `$.append` will not set its end node to the effect. As a result the end node is wrong (too early). To fix this we assign the nodes in `$.async`. We could also use `compareDocumentPosition` in `$.append` to only ever go forward, but that feels a bit heavier performance-wise. --------- Co-authored-by: paoloricciuti <ricciutipaolo@gmail.com> Co-authored-by: Rich Harris <rich.harris@vercel.com>	1 month ago
Simon H	e4bfc5f45a	chore: bump esrap dependency (#18372) ... Avoid installing the broken 2.2.10 version Closes #18369	1 month ago
paoloricciuti	ee2f74ebf3	fix: lint	1 month ago
paoloricciuti	0490721eef	fix: lint and check	1 month ago
Paolo Ricciuti	114be1c593	Merge branch 'main' into svelte-custom-renderer	1 month ago
Paolo Ricciuti	56856f0b10	fix: `customRenderer` as a global option and remove `without_renderer` (#18359) ... Co-authored-by: Rich Harris <rich.harris@vercel.com> Co-authored-by: Rich Harris <hello@rich-harris.dev>	1 month ago
Minh Lê	94bd27953e	docs: fix invalid `rows` CSS property in {#each} chess-board example (#18367) ... ## Problem The `{#each}` docs page has a runnable "chess board" example whose `<style>` block uses an invalid CSS property: ```css .chess-board { display: grid; grid-template-columns: repeat(8, 1fr); rows: repeat(8, 1fr); /* ← not a real CSS property */ ... } ``` There is no `rows` shorthand or longhand in CSS, so this line is silently dropped by every browser. The example clearly intends an 8×8 grid (note the matching `grid-template-columns` line directly above). ## Fix `rows` → `grid-template-rows`. One line, in `documentation/docs/03-template-syntax/03-each.md`. No prose changes.	1 month ago
Simon H	1b4c43b8e5	fix: ignore declaration tags for animation directive (#18366) ... Fixes #18365	1 month ago
Mathias Picker	85dcb91f7c	chore: upgrade to vitest v4 (#18265) ... ## Summary Bumps `vitest` and `@vitest/coverage-v8` from `^2.1.9` to `^4.1.7` (two major versions). Three small test-harness updates compensate for vitest 4 behavior changes; all 7569 tests still pass. - **`packages/svelte/tests/runtime-browser/test.ts`** — vitest 4 removed the deprecated `describe(name, fn, opts)` signature. Pass options as the second argument. - **`packages/svelte/tests/runtime-legacy/shared.ts`** — vitest 4's jsdom env binds `virtualConsole` to the original `globalThis.console` reference *before* vitest wraps the console, so inline-`<script>` logs and `jsdomError` events no longer reach per-test `console.{log,error}` overrides. Restore vitest 2's behavior by re-routing the virtual console through the live `console` in `beforeAll`. Also promote the window-error listener to a named function and remove it in `finally` — previously leaked listeners from earlier tests kept writing to module-level `unhandled_rejection`, polluting later tests. - **`vitest.config.js`** — bump `testTimeout` to 10s. The 5s default trips a handful of dev-mode tests that exercise `effect_update_depth_exceeded`, whose ~1000 Error-stack captures per flush are slower under vitest 4's deeper async stacks. ## Test plan - [x] `pnpm test` — 7569 passed, 63 skipped, 0 failed --------- Co-authored-by: Claude Opus 4.7 (1M context) <noreply@anthropic.com> Co-authored-by: Rich Harris <rich.harris@vercel.com>	1 month ago
Simon H	a81f96549d	fix: reject pending async deriveds on discard (#18308) ... The rejects all async deriveds of a batch as `OBSOLETE`, so they don't hang around and bail early without triggering the batch. If we don't do this, an async derived can trigger the already done batch, which schedules an effect that is never flushed. Because it is never flushed the branches it touched on its way up are never cleared, and so anything else in that subtree is now unreactive.	1 month ago
github-actions[bot]	3ef761b87b	Version Packages (#18346) ... This PR was opened by the [Changesets release](https://github.com/changesets/action) GitHub action. When you're ready to do a release, you can merge this and the packages will be published to npm automatically. If you're not ready to do a release yet, that's fine, whenever you add more changesets to main, this PR will be updated. # Releases ## svelte@5.56.1 ### Patch Changes - fix: error at compile time on duplicate snippet/declaration tag definitions ([#18351](https://github.com/sveltejs/svelte/pull/18351)) - fix: parse declaration tag contents more robustly ([#18353](https://github.com/sveltejs/svelte/pull/18353)) - fix: correctly transform references to earlier declarators in a declaration tag (e.g. `{let a = $state(0), b = $derived(a * 2)}`) ([#18348](https://github.com/sveltejs/svelte/pull/18348)) - fix: avoid spurious `state_referenced_locally` warnings for `$derived` declarations in declaration tags ([#18348](https://github.com/sveltejs/svelte/pull/18348)) - fix: tolerate whitespace before `let`/`const` in declaration tags ([#18348](https://github.com/sveltejs/svelte/pull/18348)) - fix: prevent infinite loop when a tag's expression ends with a trailing `/` at the end of the input ([#18350](https://github.com/sveltejs/svelte/pull/18350)) - fix: more robust parsing of declaration tags with regards to `type` ([#18330](https://github.com/sveltejs/svelte/pull/18330)) - fix: preserve newlines in spread input values when the `type` attribute is applied after `value` ([#18345](https://github.com/sveltejs/svelte/pull/18345)) - fix: update `SvelteURLSearchParams` when setting duplicate keys to the same joined value ([#18336](https://github.com/sveltejs/svelte/pull/18336)) - fix: check references for blockers on server, too ([#18352](https://github.com/sveltejs/svelte/pull/18352)) Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>	1 month ago
Simon H	5b8db1be35	fix: error at compile time on duplicate snippet/declaration tag definitions (#18351)	1 month ago
Simon H	56013a2c01	fix: check references for blockers on server, too (#18352) ... We forgot the server in #18309 Fixes #18344	1 month ago
hjaber	2afb895ffa	fix: parse declaration tags with a division operator in the initializer (#18353) ... Mimic what `parseExpressionAt` does --------- Co-authored-by: hjaber <hjaber@users.noreply.github.com> Co-authored-by: Simon Holthausen <simon.holthausen@vercel.com>	1 month ago
paoloricciuti	e5beb7297d	fix: push renderer in effects even if it's `null` ... This is needed because effects record the renderer they were created with. But if we interlieve a DOM renderer with a Custom one, current_renderer could be a custom one when the effect reruns and if we don't push `null` it will use the custom_renderer methods instead of the DOM ones.	1 month ago
paoloricciuti	131fae3c2d	fix: update snapshot	1 month ago
paoloricciuti	f781fe485c	chore: fix type	1 month ago
paoloricciuti	a72c3650c9	Merge remote-tracking branch 'origin/main' into svelte-custom-renderer	1 month ago
paoloricciuti	c58c5565c0	fix: noop on the server for custom renderer components	1 month ago
Yuichiro Yamashita	b471c15e61	fix: don't hang on a tag whose expression ends with a trailing slash (#18350) ... close https://github.com/sveltejs/svelte/issues/18349	1 month ago
Yuichiro Yamashita	c74f44fff9	fix: don't mistake `type` identifier expressions for TS `type` declarations in tags (#18330) ... Detect TypeScript `type` declarations by actually parsing instead of by character-class blacklists, so that expressions like `{type === 'all' ? a : b}` or `{type instanceof Foo}` aren't misclassified as malformed declarations. Closes #18328 --------- Co-authored-by: Simon Holthausen <simon.holthausen@vercel.com> Co-authored-by: Simon H <5968653+dummdidumm@users.noreply.github.com>	1 month ago
Rich Harris	b76b937e00	fix: various declaration tag bugs (#18348) ... closes #18334	1 month ago
jdoughty04	378bb25097	fix: set input type before spread value (#18345) ... ### Before submitting the PR, please make sure you do the following - [x] It's really useful if your PR references an issue where it is discussed ahead of time. In many cases, features are absent for a reason. For large changes, please create an RFC: https://github.com/sveltejs/rfcs - [x] Prefix your PR title with `feat:`, `fix:`, `chore:`, or `docs:`. - [x] This message body should clearly illustrate what problems it solves. - [x] Ideally, include a test that fails without this PR but passes with it. - [x] If this PR changes code within `packages/svelte/src`, add a changeset (`npx changeset`). ### Tests and linting - [x] Run the tests with `pnpm test` and lint the project with `pnpm lint` Fixes #18332. When spread props put `value` before a later `type="hidden"` on an `<input>`, `set_attributes` currently writes the value while the element is still a text input. Browsers sanitize text input values by removing newlines, so the hidden input permanently loses them before `type` is applied. This sets an input's `type` first whenever the same spread update also includes `value` or `__value`, so the existing value handling runs with the final input type. The new runtime-browser sample covers both the problematic spread-first order and the already-working type-first order. Validation: - `corepack pnpm lint` - `CI=1 corepack pnpm test` - `corepack pnpm vitest run packages/svelte/tests/runtime-browser/test.ts -t input-type-before-value-spread` --------- Co-authored-by: Rich Harris <rich.harris@vercel.com> Co-authored-by: Rich Harris <hello@rich-harris.dev>	1 month ago
jiyujie2006	2f6307af65	Fix searchParams.set duplicate updates (#18336) ... ## Summary - detect `SvelteURLSearchParams#set` changes by comparing duplicate value lists instead of joined strings - update reactive subscribers when duplicate params collapse to a single value with the same concatenated text - cover both `SvelteURLSearchParams` and `SvelteURL.searchParams` synchronization ## Tests - pnpm exec vitest run packages/svelte/src/reactivity/url-search-params.test.ts -t "URLSearchParams.set updates when duplicate values collapse to the same joined string" - pnpm exec vitest run packages/svelte/src/reactivity/url-search-params.test.ts packages/svelte/src/reactivity/url.test.ts --------- Co-authored-by: Rich Harris <hello@rich-harris.dev>	1 month ago
ottomated	11985c020f	docs: desloppify browser support page (#18333) ... The text in https://svelte.dev/docs/svelte/browser-support is quite redundant and has a lot of AI smell. I did my best to make it more concise and remove some slop comments in the generation script. Requires a companion PR in svelte.dev: https://github.com/sveltejs/svelte.dev/pull/2013 <img width="620" height="809" alt="image" src="https://github.com/user-attachments/assets/7be9d5ff-1623-4897-a2d4-9a601b20f54d" /> ### Before submitting the PR, please make sure you do the following - [ ] It's really useful if your PR references an issue where it is discussed ahead of time. In many cases, features are absent for a reason. For large changes, please create an RFC: https://github.com/sveltejs/rfcs - [x] Prefix your PR title with `feat:`, `fix:`, `chore:`, or `docs:`. - [x] This message body should clearly illustrate what problems it solves. - [ ] Ideally, include a test that fails without this PR but passes with it. - [x] If this PR changes code within `packages/svelte/src`, add a changeset (`npx changeset`). ### Tests and linting - [x] Run the tests with `pnpm test` and lint the project with `pnpm lint` --------- Co-authored-by: Rich Harris <rich.harris@vercel.com>	1 month ago
Rich Harris	05a3bce6bb	chore: minor tidy-ups (#18326) ... some more stuff extracted from #18035	2 months ago
paoloricciuti	dfd18af3c3	chore: add comment	2 months ago
paoloricciuti	585e444bcc	fix: treeshake hydration	2 months ago
Rich Harris	e027d8bf22	chore: annotate `each_array` return value (#18325) ... tiny self-mergeable change extracted from #18035 — just turns `each_array` from a `Derived<any[]>` to a `Derived<V[]>`	2 months ago
github-actions[bot]	70afafe18e	Version Packages (#18315) ... This PR was opened by the [Changesets release](https://github.com/changesets/action) GitHub action. When you're ready to do a release, you can merge this and the packages will be published to npm automatically. If you're not ready to do a release yet, that's fine, whenever you add more changesets to main, this PR will be updated. # Releases ## svelte@5.56.0 ### Minor Changes - feat: allow declarations in the template ([#18282](https://github.com/sveltejs/svelte/pull/18282)) ### Patch Changes - perf: use `createElement` instead of `createElementNS` for HTML elements ([#18262](https://github.com/sveltejs/svelte/pull/18262)) - perf: store `current_sources` as a `Set` for O(1) membership checks ([#18278](https://github.com/sveltejs/svelte/pull/18278)) - perf: deduplicate identical hoisted templates within a component ([#18320](https://github.com/sveltejs/svelte/pull/18320)) - perf: hoist `rest_props` exclude list as a module-scope `Set` ([#18252](https://github.com/sveltejs/svelte/pull/18252)) Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>	2 months ago
Rich Harris	6d26dce265	allow class/function expressions in tags (#18324) ... We shouldn't error on encountering `class` and `function`, because they are treated as expressions rather than declarations	2 months ago
7nik	6a955758a6	docs: mention since version of delcaration tags (#18322)	2 months ago
Mathias Picker	2fae91af6d	perf: deduplicate identical hoisted templates within a component (#18320) ... Byte-identical templates hoisted from different fragments, elements, or branches now share a single `$.from_html` factory instead of each emitting its own module-scope variable - this shrinks generated output and avoids redundant runtime template parsing. Dedup is keyed on `(content, flags)` and is skipped in dev mode (templates there are wrapped in `$.add_locations`, which embeds per-call-site info). ### Before submitting the PR, please make sure you do the following - [ ] It's really useful if your PR references an issue where it is discussed ahead of time. In many cases, features are absent for a reason. For large changes, please create an RFC: https://github.com/sveltejs/rfcs - [x] Prefix your PR title with `feat:`, `fix:`, `chore:`, or `docs:`. - [x] This message body should clearly illustrate what problems it solves. - [x] Ideally, include a test that fails without this PR but passes with it. - [x] If this PR changes code within `packages/svelte/src`, add a changeset (`npx changeset`). ### Tests and linting - [x] Run the tests with `pnpm test` and lint the project with `pnpm lint` --------- Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com> Co-authored-by: Rich Harris <rich.harris@vercel.com>	2 months ago
Mathias Picker	a40c745fd9	perf: hoist rest_props exclude list as a module-scope Set (#18252) ... The compiler emitted an inline string array as the second argument to `$.rest_props(...)`, and the runtime did a linear `Array.prototype.includes` on it on every property access via the rest-props Proxy. The exclude list only depends on the component definition, not on the instance, so it can be hoisted to module scope and shared by every instance. Switching it to a `Set` at the same time makes each lookup O(1). For a component like `<Button ...rest />` rendered N times, this turns one per-instance allocation (plus a linear search on every rest-prop access) into one module-scope allocation plus O(1) lookups. The legacy `$$restProps` path is unchanged — it mutates the exclude list in its `deleteProperty` trap, so it can't share a hoisted Set across instances. --------- Co-authored-by: Rich Harris <rich.harris@vercel.com>	2 months ago
Simon H	980c7e2321	fix: don't error on `{type}` in declaration tags (#18321) ... The `\b` in the regex matched on `}`, too, so you would get a very confusing "invalid declaration tag" error on `{type}`. Match on whitespace instead (because that's what has to come afterwards) --------- Co-authored-by: vercel[bot] <35613825+vercel[bot]@users.noreply.github.com> Co-authored-by: Vercel <vercel[bot]@users.noreply.github.com> Co-authored-by: Rich-Harris <hello@rich-harris.dev>	2 months ago
paoloricciuti	1711d7bf9e	fix: set hydrating to false in `push_renderer`	2 months ago
Rich Harris	5300843e86	chore: bump playwright (#18319) ... hopefully this will fix the CI timeouts	2 months ago