The homepage hero lays out the announcement banner and the search field on
the same flex row. The `.search-field` wrapper was a plain block, so its
`docs-text-field` kept its intrinsic height instead of filling the row,
leaving the two pills at different heights and vertically misaligned.
Make `.search-field` a flex container so the search control stretches to the
row height and matches the banner.
On tablet the social and theme mini-menus didn't line up with the buttons that
open them.
All three mini-menus now share one tablet positioning rule on `.adev-mini-menu`
that centers each panel under its trigger, with `--social`/`--theme`/`--version`
modifiers selecting the anchor; the version picker's on-screen behavior is
unchanged. The social trigger also gains `aria-controls` + a matching menu `id`
for a11y parity with the theme trigger.
Follow-up to #69205. After switching adev's COEP to `credentialless`, the
cross-origin YouTube iframe in `<docs-video>` loads in Chromium and Safari but
not Firefox, whose `credentialless` policy does not extend to nested frames. The
result was a COEP error screen instead of the player.
Render `<docs-video>` as a lightweight thumbnail facade instead of embedding the
iframe directly. The thumbnail is a cross-origin subresource, so it loads under
`credentialless` in every browser. `DocViewer` then upgrades the facade to the
inline player on hydration in browsers that can load the embed (Chromium,
Safari), preserving the previous behavior there. On Firefox the facade stays a
plain link that opens the video on YouTube (with autoplay), which replaces the
error screen.
The thumbnail uses `maxresdefault` and falls back to `hqdefault` when a video
has no max-resolution image.
Add font family, size and weight to the `.docs-primary-btn`. This guarantees that applying the class to non-button elements, like anchors, will results in the same visual representation.
Add an internals doc covering how the injector tree is built from the directive forest's resolution paths, rendered with d3, and how injector data is read from the DI debug APIs.
`animate.leave` was skipped — the element was removed from the DOM
synchronously instead of running its leave animation — whenever a
sibling instance of the same template entered in a different DOM parent
during the same change-detection tick (e.g. an exclusive-expansion
accordion or nav where opening section B collapses section A).
`leavingNodes` is keyed by `TNode`, which is shared by every instance of
a template. When a node was inserted, `cancelLeavingNodes` force-removed
any tracked leaving node whose DOM parent differed from the entering
node's parent (the `leavingParent !== newParent` branch added to
de-duplicate a dynamic component re-rendered into a fresh overlay pane).
For two distinct live sibling instances that merely share a `TNode`,
"different parent" is the normal situation, so the still-animating
sibling was ripped out.
Track the declaration view of each leaving element alongside it, and
only perform the cross-parent removal when the entering element belongs
to the same declaration view as the leaving one — i.e. the same logical
view re-rendered, the case the branch was written for. Two distinct
instances of a shared template have different declaration views, so
their `animate.leave` is now left to run to completion.
This preserves the dynamic-component/overlay de-duplication (#67032) and
the drag-and-drop node-move rescue (#67361), which are unchanged.
Fixes#69291
Serialize transfer cache request parameters without comma-joining repeated values so distinct HttpClient requests cannot reuse the same cached response.
Initialize zoneSymbolEventNames with Object.create(null) instead of {}.
This is hardening only. addEventListener('__proto__', fn) is not
directly attacker-controllable — its presence in an application is
itself an application bug and a prerequisite for any issue here.
Without this change, if that application bug exists, two unexpected
behaviors follow depending on environment:
Browser: zoneSymbolEventNames['__proto__'] reads the __proto__ getter
and returns Object.prototype (truthy), bypassing prepareEventNames.
symbolEventName resolves to undefined and window['undefined'] = []
throws TypeError.
Node.js + --disable-proto=throw: the assignment
zoneSymbolEventNames['__proto__'] = {} inside prepareEventNames
triggers the disabled __proto__ setter and throws.
Using Object.create(null) removes the __proto__ accessor from the
map so the key is treated as a plain missing property in both cases.
`findAnchorFromDocument` interpolates the raw url fragment into
`[name="${target}"]` for the shadow DOM lookup, so a fragment reachable
through the router when `anchorScrolling` is on can break out of the
attribute selector and make `querySelector` throw or match unrelated
nodes, and it also breaks legitimate anchor names containing a quote.
Wrap the value in `CSS.escape` so it stays a single attribute-value token.
The change can be treated as a continuation of #68535.
- Make the table header sticky
- Reduce slightly the table density in order to equalize it with the rest of the UI
- Reduce the font size of the type pills
- Change the position of the filter clear button to match with the rest of the filter inputs
- Add a "No such key" label when there are no data rows as a result of filtering
Improve the name of the public methods of the visualizers by using common terms like "highlighted". This should, hopefully, clear up some ambiguities that came with the term "selected". Additionally, clean up some CSS class names.
This refactoring renames the 'element' (tagName string) property of DevToolsNode to 'tagName' for clarity, as 'element' was ambiguous.
It also restores 'nativeElement' as a flat property (undoing the nesting introduced previously) because there are cases (like control flow blocks) where a tag name is present without an associated DOM node, making nesting inappropriate.
Resolve the approved relative workspace tsdk path to an absolute path by checking workspace folders. This ensures the path is correctly resolved on the server side.
Fixes#69276
This limit breaks ts-ignore comments when using this for our source->source transform.
Rather than overridding it there, it's just removed here since we don't care about the limit
This test flakes occasionally because it is called in production when a platform is created and unregistered when a platform is destroyed. However, not all tests properly clean up their platforms, meaning we can accidentally leak platforms between tests. If this happens, we end up have an event listener created from the production code path and a second event listener from the test. When the test emits the event, both listeners respond and it causes too many responses which fails the test.
Ideally, all tests would clean up the platforms correctly, but this seems difficult to guarantee for all Angular tests and is likely to break over time. The simplest solution is just destroy any leaked platform before the test starts. It's a bit elegant, but the safest option.
Convert the router details panel in DevTools to use a description list (<dl>) instead of a table. This is semantically more correct for key-value metadata and improves accessibility.
Normalize SVG animation attributeName lookup to also recognize lowercase attributename before allowing dynamic animation value bindings.
Add runtime and platform-server SSR regression coverage for lowercase attributename retargeting.
Replace `in` operator with `Object.hasOwn` in
formatDateTime to prevent prototype pollution attacks.
The `in` operator traverses the prototype chain, meaning a polluted
Object.prototype key could be picked up as a valid replacement value.
This is especially critical in SSR environments where a single
prototype pollution attack persists across all subsequent requests in
the shared Node.js process, potentially injecting malicious content
into every user's rendered HTML.
Using `Object.hasOwn` restricts the lookup to
own properties only, blocking prototype chain traversal.
Reading the form model on init is unsafe as it could depend on inputs (eg a required input). We need to delay the read by a tick (after the inputs are set) to ensure that values can be safely read.
fixes#69262