Queue observer records for every same-value attribute write, including setAttributeNS and dataset. Keep meta declaration processing independent so it does not queue a duplicate record.
Report the matched Attr local name and namespace from ordinary and namespace-aware writes, then match refresh attributes exactly. Cover unrelated uppercase/namespaced mutations and ordinary HTML case folding in both main and child Documents.
Share a lazy inert owner across each document's template contents and keep
HTML/XML semantics, interface identity, and template-owner state separate.
Cloning preserves interfaces while creating independent owner state.
Separate portable Document metadata from arena-local relationships. Use
one metadata-copy boundary for Document cloning and main-document
replacement, clearing template-owner and CSS-target handles while keeping
whole-DOM snapshots and copy-on-write storage relationships intact. Metadata
copies also discard base-element href/target caches while retaining about
base URLs and explicit overrides; only DOM node cloning freezes computed
base URLs. Detached Document URL readers use native metadata without JS
URL/documentURI/baseURI shadow state, including selector and bridge readers.
Apply the DOM-specific document cloning rules separately: freeze the source
computed base URL as an override, separately retain the inherited about
base URL, and rebuild base-element caches from the cloned tree. Preserve
main-document replacement metadata semantics.
Establish exact HTML template identity and associated fragment allocation
in one native entry shared by DOM and parser creation. Clone actual native
contents rather than guessing from names, preserving foreign and mixed-case
elements through serialization, deep cloning, import and adoption. Keep HTML
createElement normalization at its binding boundary; namespace-aware creation
and wrapper selection preserve the exact local name, including registry paths.
Plan per-node adoption targets across ordinary children, shadow roots and
template contents. Retarget custom-element registries and callbacks with
the actual destination documents, and clear cross-document shadow sheets
while retaining sheets within a document and its own template owner.
Avoid preparing unused inert style sources.
Cover cross-arena ID collisions, metadata preservation, snapshot isolation,
nested templates, deep adoption, iframe/detached documents, cloned
interfaces, retained sheet arrays, registry associations and URL behavior.
Track the inert template-document role separately from HTML/XML semantics.
Choose the Document wrapper for that role while preserving the source kind
for HTML name normalization and XML element creation.
Cover parser/createElement/innerHTML paths in main and child documents,
DOMParser and createHTMLDocument sources, and XHTML contents. Ordinary
Document interfaces retain their existing identity.
Validation: cargo fmt --all; strict workspace/all-targets/all-features
Clippy; 86 template-related tests across three stress iterations; full
Nextest 21,167 passed, 19 skipped, without retries.
Process parser and script-created refresh declarations in order and retain them
on their owning Document across load callbacks and adoption. Share candidate
priority and deadline state between main and child documents, and arm child
timers after owner load completes.
Consume queued work by exact Document identity and revision, classify navigation
when it fires, and keep the registered target URL. Reprocess same-value meta
attributes using the matched Attr identity so XML case, namespaces, observer
records, and refresh deadlines remain correct.
Cover competing declarations, stale tasks, independent frames, load/adoption
boundaries, sandbox/shadow handling, saved targets, and fragment navigation.
Retarget ShadowRoot ownership and descendants along with the adopted host,
and collect stylesheet owners from nested shadow trees. Cover closed and
nested roots, inactive documents, and deep iterative adoption.
Keep press and current mouse targets in host-owned state so synchronous
DOM removals invalidate or retarget them before click dispatch. Resolve
click and auxclick targets through the flat tree, preserve captured click
targets, and dispatch mouseup before lostpointercapture.
Cover sibling targets, Shadow DOM, slots, removal and reinsertion, and
pointer capture with deterministic input regressions.
Group ignored reasons, values and nonempty properties behind one optional box. Use nonzero backend IDs and serialize fixed role/name kinds on demand, reducing the common AX node from 296 to 96 bytes while preserving payload shape and field order.
Add reproducible CLI and native construction benchmarks. Generated 40000-row text dumps use 382.6 to 341.5 MiB peak RSS; 200000-row text dumps use 1753.5 to 1612.8 MiB. Record the roughly 5.5 percent native construction cost for dense ignored nodes alongside the memory gains.
Validation: cargo fmt --all; workspace clippy with all targets/features and warnings denied; cargo nextest run --no-fail-fast (19890 passed, 17 skipped). All 188 formal benchmark samples match their corresponding before/after outputs.
Large semantic dumps retain a JSON object graph with allocated field names
and string node references for every AX node before formatting the output.
Build compact typed nodes with numeric references and pass them directly
from the renderer to the CLI. Keep one AX traversal for both typed nodes
and CDP payloads, converting protocol nodes individually as they are built.
Preserve omitted/empty fields, field insertion order, list markers, and
frame-root attachment order. Render text iteratively and trim in place.
Add compatibility coverage and a standalone benchmark that stages both
executables on the same filesystem, alternates runs, records peak RSS and
wall time, and rejects changes to dump output bytes.
Against main at 5f5e1fff3 on a generated 40,000-row list (three interleaved
runs, medians):
- semantic_tree_text: 718.42 -> 390.96 MiB; 2.919 -> 2.204 s.
- semantic_tree: 815.71 -> 483.81 MiB; 2.484 -> 2.369 s.
- markdown control: 95.52 -> 95.59 MiB; 0.895 -> 0.913 s.
A generated 200,000-row list also completes with identical bytes (one run):
- semantic_tree_text: 3473.02 -> 1775.07 MiB; 13.245 -> 10.015 s.
- semantic_tree: 3803.45 -> 2137.26 MiB; 12.626 -> 10.397 s.
JSON/text comparisons also match the baseline on a mixed fixture containing
hidden/ignored nodes, form values, list markers, shadow DOM, and child frames.
These generated fixtures are independent of the external MOL-269 evidence.
Validation:
- cargo fmt --all
- cargo clippy --workspace --all-targets --all-features -- -D warnings
- cargo nextest run --no-fail-fast (19,885 passed, 17 skipped)
- cargo build --release -p moli
Refs: MOL-269
Add the native DOMString value representation that preserves unpaired UTF-16 code units and exposes a scalar projection for native consumers. Equality and hashing retain original units; parser scalar appends preserve allocation capacity. Unit coverage includes surrogate identity and append boundaries. This PR provides the native value API; migrating all DOM producers is separate work.
Completes native SVG value and element interface exposure in one independently mergeable PR against main, refreshing the former SVG constructor PR. Animated scalar values, angles, preserve-aspect-ratio values and live point/string/number/length/transform lists provide the shared native DOM foundation for the added elements.
Adds the remaining component-transfer, flood, image, merge, merge-node and tile filter interfaces; SVGSwitchElement and SVGMPathElement; and the animation, animate, set, animate-motion and animate-transform interfaces. Native factories, inheritance, SVGTests lists, attribute reflection, receiver validation and animation event handlers use native objects and the existing DOM backend.
Tests cover main/iframe realms, live and windowless documents, native identity, value/list mutation, units, Trusted Types, animation target resolution, handlers and timing argument validation. All internal SVG dependencies are included here, with no external PR merge order.
Filter raster rendering and SMIL scheduling remain unimplemented. Timing controls reject NotSupportedError, getStartTime reports no interval, and getCurrentTime remains zero. The separate SVG path parser changes and animated-string UTF-16 storage migration are excluded. The retained length synchronization is included; a separate ignored regression records the lone-surrogate limitation in the current DOM attribute storage.
SVG event-handler properties reuse main's existing shared event backend. Tests verify replacement, clearing, canonical event types, cancellation and realm/receiver identity; complete handler/listener registration-order semantics remain part of the separate event-handler backend migration.
Create script auxiliary Pages synchronously and let their CDP targets adopt
those exact Pages. Preserve Page and WindowProxy identity through document
navigation, sharing browsing-context names, opener references, message endpoints
and close state with related windows.
Commit browser responses into the existing Page with exact document admission,
cancellation and failure disposition. Capture accepted navigation origin, base,
policy, storage and referrer, preserve auxiliary history and restore related
browsing-context groups without issuing duplicate popup requests.
Reuse the initial same-origin Window while replacing its Document, and centralize
Document owner transitions, descendant retirement and Inspector publication.
Give related Pages independent Inspector routes and shared per-Page startup
pause state, route isolate foreground work through surviving Pages, and cache
cross-origin Window functions in each accessing realm with receiver-aware access.
Cover actual popup state, navigation and parser rebinding, accepted environments,
history, cross-origin access, multiple frontends, cancellation and task retirement.
Validation: cargo fmt --all; cargo clippy --workspace --all-targets --all-features -- -D warnings; cargo nextest run --no-fail-fast.
Introduce moli-webmcp for native form schema synthesis, complete input validation,
numeric conversion, and JSON-LD results. Preserve requested selection values in
fill intents so the renderer can apply them to current controls after author
input events.
Add native tool activity states, selector support, control helpers, and
a namespace-aware form index that includes open and closed shadow trees.
Pin the Stylo revision providing the activity pseudo-classes.
Validation on this commit tree: cargo fmt --all;
cargo clippy --workspace --all-targets --all-features -- -D warnings;
cargo nextest run --no-fail-fast.
Summary [ 98.804s] 19248 tests run: 19248 passed, 16 skipped
Count UTF-16 code units when masking password values to match Chromium.
Cover initial and live values across ASCII, BMP, supplementary characters,
combining sequences, ZWJ emoji, and empty passwords, verified against
Chromium 145 in 14 cases.
Separate parser provenance from active child construction. Share one
construction session across HTML/XML parsers and runtime adapters, with
completion ownership, normal-EOF checks, and cancellation that does not
publish incomplete styles.
Use ordinary creation for generated document shells and remove obsolete
per-element completion forwarding. Cover parser ownership, adoption,
duplicate completion, missing callbacks, and cancellation in tests.
Process parser-created style sources after their closing tag or EOF, and
prepare later source changes through the stylesheet owner lifecycle.
Keep synchronization and CSSOM reads on prepared sources so CSSOM edits
survive reads and document initialization.
Complete styles in inert XML parsing and build the XML viewer through
ordinary DOM creation. Cover streaming, EOF, child documents, fragments,
and XML stylesheet behavior in regression tests.
ldm0
Maxim M
lanyue-llk
Donough Liu