* test(interception): add failing sibling-interception graph tests
* feat(routing): discover sibling-style interception markers
* feat(routing): emit sibling interception manifest facts
* feat(app-router): serialize sibling intercepts in route manifest
* feat(app-router): match sibling intercepts in route matcher
* feat(app-router): render sibling intercept as full page response
* test: add siblingIntercepts field to AppRoute test fixtures
* test(interception): add sibling interception fixtures
* fix(app-router): emit null proof for sibling intercepts to bypass slot-validation
* test(interception): add sibling interception e2e specs
* fix(interception): correct (...)→(..)(..) marker in interception-from-root fixture
* fix(interception): resolve sourceRouteIndex to slot-owner route in intercept lookup
When a route inherits a @slot from an ancestor via discoverInheritedParallelSlots,
the inherited slot's interceptingRoutes are copied from the ancestor. Previously,
createInterceptLookup used routeIndex (the inheriting descendant's index) as
sourceRouteIndex, causing resolveAppPageInterceptState to see sourceRoute ===
currentRoute → kind='current-route' instead of kind='source-route'.
Fix: build a patternToIndex map and resolve sourceRouteIndex via the intercept's
sourceMatchPattern (which names the actual slot-owner route), falling back to
routeIndex only when no match is found.
* fix(interception): use locale-prefixed href so middleware interception fires correctly
* fix(interception): address codex review issues in sibling intercept implementation
- Pass effectiveParams to resolveAppPageHead so generateMetadata/generateViewport
sees the intercepted route's params instead of the source route's params
- Rename SIBLING_INTERCEPT_SLOT_NAME from __page to __vinext_sibling_intercept
to prevent collision with a user @__page parallel route directory
- Restore middleware-rewrite fixture link without locale prefix so the
interception-dynamic-segment-middleware spec exercises the intended
middleware rewrite path
* fix(test): update slotId assertions and restore locale-prefixed fixture link
- Update test assertions from slot:__page to slot:__vinext_sibling_intercept
following the rename in commit 2248b8fe
- Restore locale-prefixed href in interception-mw fixture so the
interception-dynamic-segment-middleware E2E tests pass (reverts the
accidental revert from 2248b8fe back to the working state from 02e80c26)
* fix(interception): address review findings on sibling interception PR
- Restore locale-less href in interception-mw fixture and fix the
underlying middleware-rewrite interception gap: add a Referer-based
fallback in app-rsc-handler so interception fires when middleware
rewrites the URL before the server matches it (the client-side manifest
check sees the pre-rewrite target and skips setting the interception
context header; the Referer carries the source page the server needs)
- Apply interceptLayouts for sibling intercepts in buildPageElements so
a layout.tsx under the interception marker dir wraps the intercepting
page, matching Next.js segment-layout wrapping behaviour; add fixture
layout and E2E assertion
- Add unit test proving findOwnerRouteForDir ancestor walk attaches a
sibling intercept to the nearest ancestor route when the marker's
parent dir has no page.tsx (not dropped silently)
* fix(interception): send current pathname as fallback interception context
The client-side manifest check compares the pre-middleware target URL
against declared interception patterns. When middleware rewrites the URL
(e.g. /foo/p/1 → /en/foo/p/1), the pre-rewrite URL has fewer segments
and doesn't match the pattern, so no X-Vinext-Interception-Context
header is sent and the server cannot fire interception.
Fix: when the manifest check yields no match and there is no prior
intercepted navigation in browser state, fall back to sending the
current page pathname as the interception context. This mirrors Next.js
sending Next-URL on every RSC navigation. The server's findIntercept is
double-gated on both source and target matching, so false positives
cannot occur. The previousNextUrl is also populated so back/forward
traversal can restore intercepted state.
Revert the Referer-based server-side fallback (unreliable across
environments) in favour of this explicit browser-side fix.
* fix(interception): gate middleware-rewrite fallback on declared source pattern
The previous fix sent window.location.pathname as fallback interception
context on every unmatched navigation, which broke prefetch cache reuse:
a prefetch stored with null context wouldn't be found when the navigation
used a non-null context key.
Refine the fallback to only fire when the current page URL matches at
least one sourcePatternParts declared in the route manifest — i.e., the
current page is a known interception source. This preserves prefetch
cache reuse for ordinary navigations while still enabling middleware-
rewrite interception for pages that declare an interception relationship.
* fix(interception): support middleware-rewrite targets in fallback context and planner
The previous fallback-context gate only checked whether the current page
is a declared interception source; it did not verify the target URL
could plausibly be a middleware-rewritten version of the declared target
pattern. This meant the client sent interception context when it
shouldn't (pre-middleware target had no static anchors in common with the
pattern) and the planner hard-navigated when the proof target matched
only after the rewrite.
Three coordinated fixes:
1. Client fallback gate (app-browser-entry / app-browser-interception-context):
replace the broad source-only check with matchRoutePatternWithOptionalDynamicSegments
on the target side — the pre-middleware URL must still share every
static anchor of the declared target pattern, in order, while allowing
dynamic segments to be absent.
2. Planner interception proof (navigation-planner):
findRouteManifestInterceptionForProof now accepts a proof whose target
matched URL equals the pre-rewrite canonical pathname (shorter than the
full pattern) when matchRoutePatternWithOptionalDynamicSegments passes.
Exact pattern match still skips the targetRouteId guard; the relaxed
path skips it too so the middleware-added locale segment does not
invalidate an otherwise valid proof.
3. displayPathname threading (app-rsc-handler / app-rsc-entry / app-page-element-builder):
the handler now passes canonicalPathname as displayPathname separately
from cleanPathname. buildPageElements uses cleanPathname for all
internal routing work (slot param extraction, slot overrides) and
displayPathname only for render identity — keeping the browser-visible
URL in RSC metadata without affecting internal matching.
* fix(interception): guard siblingInterceptElement construction and document reserved slot name
* fix(interception): surface missing default export on sibling intercept pages
A sibling-style interception whose intercepting page module is missing
its `default` export previously fell back to the source route's page
component (`effectivePageModule?.default ?? PageComponent`), silently
rendering a *different* page than the one requested.
Resolve the effective component from the intercepting page only and
extend the existing no-default-export guard to cover sibling intercepts,
so a missing export surfaces the same explicit "Page has no default
export" error the source/slot paths already produce. For a normal
request this is identical to the previous behavior since
`effectivePageModule === pageModule`.
Addresses bonk review feedback on #1804.
* test(app-router): guard canonical pathname commit on soft-nav config rewrites
The `displayPathname = canonicalPathname` change feeds the RSC payload
identity the client planner uses to commit a navigation, so it affects
every rewritten route, not just interceptions. Existing coverage
(nextjs-compat/hooks.test.ts) only exercises the hard-nav (SSR) path.
Add an e2e that soft-navigates (client-side) to an afterFiles-rewritten
route and asserts the committed URL and `usePathname()` stay canonical
(pre-rewrite), with a marker check proving no full page reload occurred.
Addresses bonk review feedback on #1804.
* fix(interception): restore __pageLoader on slot intercept lookup entries
The main merge auto-resolved incorrectly: it moved the `__pageLoader`
field that #1738 added to the *slot* intercept push block onto the new
*sibling* push block instead (the two `interceptLookup.push({...})` calls
are textually similar). As a result slot intercept lookup entries lost
`__pageLoader`, so resolveAppPageInterceptState never loaded the lazy
intercepting page (`page` stayed null) and the modal slot fell back to
default.tsx — breaking every slot interception.
Restore `__pageLoader: intercept.__pageLoader` on the slot push block.
The sibling block already carries it (see merge commit), and both
intercept types declare the optional field.
Fixes the app-router-dev-server intercept tests and the interception
e2e suites that regressed after merging main.
---------
Co-authored-by: James Anderson <james@eli.cx>
* refactor(router): quarantine navigation runtime topology
* fix(router): restore manifest-backed interception runtime
App Router browser navigations lost first-hop intercepted route context after snapshot topology stopped acting as fallback proof. That made /feed to /photos/:id request the direct page payload instead of the declared modal payload.
The violated invariant was that route topology must come from the manifest, while snapshots only carry runtime proof. Resolve first-hop interception from the manifest, move RSC bootstrap streams onto the symbol-backed runtime, teach prerender extraction that script shape, and harden the runtime contract around route manifest and RSC data.
Coverage includes runtime seam tests, prerender extraction tests, planner/browser navigation tests, package rebuild, and the app-router intercepted routes E2E slice.
* fix(router): narrow prerender RSC script parsing
The prerender extractor accepted any script starting with the symbol-runtime bootstrap expression, then parsed it as though it had the exact RSC chunk-push prefix. That was harmless at the current head, but it made future runtime scripts a foot-gun.
Only the exact runtime chunk prefix and exact done marker are now accepted. A focused extractor test covers non-chunk runtime scripts that share the bootstrap expression prefix.
* fix(router): address navigation runtime review
* fix(router): resolve navigation runtime review threads
James Anderson