* ✨ Delay building schemas in check-fn
check-fn eagerly builds its schemas which contributes to the startup time. But
we can delay this to when the check function is actually called, which seems to
be when an edit is made. On the Travel A2B - Locofy Sample project file this
reduced the time to the first viewport render by ~15%.
* ✨ Make schema resolution lazy only on production builds
Add a lazy-delay macro that defers the expression only on builds that
set the app.common.data.macros/lazy-delay? closure define, and use it
in check-fn to resolve the schema.
The previous change made the schema resolution unconditionally lazy,
which hid wrong schema definitions until the first check call. With
this change development builds, tests and storybook resolve the schema
when the namespace loads, so a bad definition fails early, and
production builds keep the deferred resolution and the startup win.
The define is enabled in the release builds of frontend (main and
worker), exporter and library. On the JVM the expression is always
evaluated immediately.
AI-assisted-by: deepseek-v4.1-flash
---------
Co-authored-by: Andrey Antukh <niwi@niwi.nz>
* 🐛 Fix token export writing the wrong active themes
The export passed only the tokens library, so the active themes and
sets in `$metadata` were the ones of the imported file, kept in the
library, instead of the ones activated in the file. The multi-file
export now needs the file's status, so no caller falls back to them.
Signed-off-by: Florian Schroedl <flo.schroedl@gmail.com>
* 🐛 Say no tokens were found when importing a file without tokens
Importing `{}`, or a JSON file without any token, said the token data
was invalid.
Signed-off-by: Florian Schroedl <flo.schroedl@gmail.com>
---------
Signed-off-by: Florian Schroedl <flo.schroedl@gmail.com>
* 🐛 Show the error page for critical WASM errors
Since #8651 the :wasm-error handler showed the error page for :panic
only, so a :wasm-critical error (a WASM trap with no error code) fell
through to a toast and the user kept working on a broken renderer.
Only the codes in recoverable-wasm-codes now show a toast: :non-blocking
from Rust, plus :wasm-reload-context-failure and :viewer-canvas-failed,
raised by the frontend. Any other code shows the error page. This covers
errors from the render loop, event handlers, promises and store events;
errors thrown inside React effects still reach the error boundary.
The h/call docstring now describes what it does: it always throws.
Closes#12990
AI-assisted-by: claude-opus-5-5
* ✨ Add a guard for React effects that call WASM
React sends any error thrown in an effect, or in its cleanup, to the
error boundary at the app root, which replaces the whole workspace with
the error page. That holds even for recoverable WASM errors, which
should only show a toast.
errors/wrap-wasm-effect runs an effect fn and the cleanup fn it
returns. A recoverable WASM error goes to on-error (toast); any other
error is rethrown, so it still reaches the boundary. The macro
with-wasm-effect wraps mf/with-effect with it.
The macro lives in its own namespace, app.main.ui.hooks.wasm-effect:
app.main.ui.hooks cannot require app.main.errors, because the renderer
loads it and that makes a circular dependency.
AI-assisted-by: claude-opus-5-5
* 🐛 Show a toast for recoverable WASM errors in the viewport
A recoverable WASM error thrown inside a viewport effect reached the
React error boundary, which replaced the whole workspace with the
error page.
Every viewport effect that calls WASM in its body or cleanup now uses
with-wasm-effect, so a recoverable error shows a toast and the
workspace stays; panics and critical errors still show the error page.
Effects that call WASM only from a listener, ts/asap or ts/raf keep
mf/with-effect: those errors never reach React.
AI-assisted-by: claude-opus-5-5
* 🐛 Show a toast for recoverable WASM errors in the text editor
The WASM text editor calls WASM from two React effects: on mount (select
all, render) and in the first caret blink. A recoverable error there
reached the error boundary and replaced the workspace with the error
page.
Both effects now use with-wasm-effect. The mount effect also guards its
select-all and render step on its own: it opens an undo transaction
with no timeout first, and without that inner guard a recoverable error
would stop the effect before it returned the cleanup that commits the
transaction and disposes the WASM editor.
AI-assisted-by: claude-opus-5-5
- Extract modifier encoding and FFI operations into common bindings
for renderer benchmarks and application callers.
- Keep readiness checks and render scheduling in the frontend API.
- Release propagation output after decoding failures.
- Cover payload ownership, heap growth, snapping and interactive
lifecycle behaviour with shared fixtures and integration tests.
AI-assisted-by: gpt-6.1-sol, deepseek-flash
Token import and export move to tokenscript.document.read and
tokenscript.document.write in the tokenscript-interpreter package
(penpot/tokenscript-interpreter#4).
tokens-lib keeps an adapter that maps its documents to a TokensLib and
back: penpot's token types, the keys of composite values and the
activation of themes and sets. About 600 lines of format detection,
legacy rewriting and tree walking go.
- Imports read DTCG 2025.10 files: {value, unit} dimensions, color
objects in their color space (oklch(l, c, h)), the fontFamily and
fontWeight type names, $root, $extends and $ref aliases. Legacy and
DTCG nodes can be mixed in one file.
- What penpot does not store is dropped and listed in one warning
toast: $extensions, $deprecated, group descriptions, values the
reader cannot use (such as a Rec. 2020 color without hex), groups
whose $extends fails and tokens of types penpot does not have.
- Sets a theme marks disabled are no longer part of the theme, and
activating a theme deactivates the other themes of its group.
- Exports are unchanged by default. export-dtcg-json and
export-dtcg-multi-file take {:profile :dtcg} for 2025.10 values and
types. Such a file reads back with the DTCG types: spacing and
sizing tokens come back as dimensions.
- The package moves from frontend/deps.edn to common/deps.edn, since
the backend and the SDK read and write token files too.
Signed-off-by: Florian Schroedl <flo.schroedl@gmail.com>
* 🐛 Fix duplicate-file crash on nil keys in deleted component objects
* ♻️ Withdraw the legacy-36 migration in favor of 0036
legacy-36 and 0036-clear-nil-keys-in-objects perform the same
cleanup of nil keys in `:objects`. Migrations only clean up
invalid data here, so they carry no order dependency between
them and running both is wasteful.
Remove the legacy-36 defmethod and its id from
`available-migrations`, the same way 0021-fix-shape-svg-attrs
was withdrawn in favor of 0035. Files that recorded legacy-36
re-migrate through 0036; their persisted `:migrations` lists
keep the old id, and the playwright fixtures recording it stay
untouched.
Add a migrate-file test proving a file that recorded legacy-36
is re-armed for 0036 and passes schema validation after.
Document the withdrawal pattern in the migration subtleties
memory.
AI-assisted-by: inkling-2-small-free
---------
Co-authored-by: Andrey Antukh <niwi@niwi.nz>
* 🐛 Fix duplicate-file crash on nil keys in deleted component objects
* ♻️ Withdraw the legacy-36 migration in favor of 0036
legacy-36 and 0036-clear-nil-keys-in-objects perform the same
cleanup of nil keys in `:objects`. Migrations only clean up
invalid data here, so they carry no order dependency between
them and running both is wasteful.
Remove the legacy-36 defmethod and its id from
`available-migrations`, the same way 0021-fix-shape-svg-attrs
was withdrawn in favor of 0035. Files that recorded legacy-36
re-migrate through 0036; their persisted `:migrations` lists
keep the old id, and the playwright fixtures recording it stay
untouched.
Add a migrate-file test proving a file that recorded legacy-36
is re-armed for 0036 and passes schema validation after.
Document the withdrawal pattern in the migration subtleties
memory.
AI-assisted-by: inkling-2-small-free
---------
Co-authored-by: Andrey Antukh <niwi@niwi.nz>
* ✨ Reference the file data storage object with a foreign key
The file data pointer lived in `file_data.metadata.storage-ref-id`, a JSON
key. The storage GC parsed it to find live references, and nothing stopped it
from physically deleting an object a live row still pointed at: a failure in
that scan lost the blob.
Migration 0159 moves the pointer to a nullable `storage_object_id` column on
`file_data`, backfills it from the metadata key, removes the key from the
rows it backfills, indexes the column and rebuilds the foreign key to
`storage_object(id)` as `NO ACTION DEFERRABLE INITIALLY IMMEDIATE`.
`INITIALLY IMMEDIATE` is deliberate: the GC catches the violation inside the
chunk transaction, while a deferred check would fire at commit, outside it,
and the sweep would never move past the blocked chunk.
The column is the only source of truth. The `storage-ref-id` entry is gone
from the file metadata schema, and the single point that serializes the
metadata drops it, so no write can resurrect it. `upsert!` does not query
the row: the pointer of the row being replaced travels in its params, and
the object is touched before the row is written. Fragments and snapshots
pass nil because a modified pointer always gets a new id; the offload passes
its own row; `insert-file!` drops it, since an INSERT never replaces a
persisted object.
`update-file!` takes the file map read from the database plus the fields to
change and merges them, so the pointer comes from the read map and no caller
can lose it by building a partial map. A map without `:data` raises
`:missing-file-data`: it is not `assert`, which is compiled out in the
deployed binary. The params cannot override the pointer or retarget the
write with `:id`/`:file-id`.
`bfc/file->response` strips the pointer from every response (`get-file`,
`create-file`, `duplicate-file`, the view-only bundle, the team access
request, the comments read, the library summary, the snapshot read and the
debug download). A `file-data` object is authentication-only, so a leaked id
would let any logged-in profile fetch the blob from `/assets/by-id`.
`get-file-fragment` resolves through the file data backend, so a fragment on
cold storage no longer hands nil to `blob/decode`, and answers only the keys
of its result schema.
The storage GC reference scan, the restore union and `objects-gc` read the
column. The offload reads `SELECT fd.*`, so its rows go through a
`decode-row` that decodes jsonb metadata and drops a JSON null, which the row
schema rejected; a row without a readable object raises
`:missing-storage-object` with the file and row ids. A chunk blocked by the
foreign key (SQLState 23503) is a real anomaly, since the reference scan
should have frozen the object: it stays an error with its own hint and the
chunk is retried on the next run until the reference is fixed.
The release is a coordinated restart, not a rolling deploy, and not
rollback-safe for the rows it writes: the previous code writes and scans the
metadata pointer, so a key it writes misses the backfill and removing it
loses the reference. `operations/2.20.0-file-data-storage-object-id.md`
builds the column, the index and the foreign key online and runs the
backfill with every instance stopped, then marks the step as applied; a
fresh or small instance does the whole job in the migration. The runbook
keeps the validation scan out of the stopped window and adds a lock timeout
and retry to the online statements.
Tests cover the foreign key shape, the backfill and its corrupt and orphan
guards, the rebuild of a missing index and of a wrong, non-deferrable or
deferred constraint, the deletion refused by a live reference, the metadata
cleanup, the offload row normalizations, the touched objects of the unshare
and the restore, the response stripping on every call site and the chunk
blocked by the foreign key. The backend memories and the developer guide
document the pointer, the GC behavior and the two migration paths.
Closes#12768
AI-assisted-by: deepseek-v4.1-flash
AI-assisted-by: inkling-2-small-free
* 🐛 Hold the file row lock across the migration write
`get-file` applies the pending migrations in memory and, when it finds
any, persists the result. That write read the row without a lock and
`update-file!` is an unconditional UPDATE, so a save that committed in
between was overwritten: the saved content, `revn` and `modified-at`
went back, the next save from that client failed with `:revn-conflict`
and the object the save left became an orphan the storage GC cannot
revisit once it is frozen.
The read now runs in a transaction of its own (`get-file` calls
`get-file*` inside `db/tx-run!` and accepts a cfg map only), so the
`file` row lock covers the re-check and the write. `get-file*` reads
the row once (`sql:get-file` plus `resolve-applied-migrations`) and
completes it with `attach-file-data`; only when a migration is pending
it reads the file again under `FOR UPDATE`, re-checks `need-migration?`
on that fresh map and persists it (`migrate-file`). The common case
stays at one row read and one migrations resolution, the migration case
keeps a single decoded copy in the heap, and a row that changed in
between is not overwritten. A migration found on the unlocked read is
applied in memory only, for the next read to persist.
`migrate-file` returns the reattached map when it persists, with the
`:storage-object-id` of the row the write left, and the realized map
when it does not; `:realize?` is what returns the read fully plain.
`fdata/upsert!` answers `{:written ... :storage-object-id ...}` instead
of a boolean, `update-file!` propagates it and the migrating read puts
it back in the map it returns, so a caller that persists again touches
the current object instead of the replaced one. `insert-file!` keeps
nil: an INSERT never replaces a persisted object.
The write also compares the stored pointer with the caller's value
(`IS NOT DISTINCT FROM`) and raises `:stale-file-data-pointer` when it
affects no row. Every writer holds the `file` row lock, so a pointer
that no longer matches is a protocol violation, not a race to retry.
The expected pointer is a separate argument because the storage branch
replaces `:storage-object-id` before the upsert, and the check lives in
`upsert-in-database`: the `legacy-db` branches do not use that SQL and
their count can legitimately be zero.
The save, snapshot and restore-backup reads take `FOR UPDATE` instead
of `FOR SHARE`: they inherit the lock, and the migration branch asks
for the exclusive one in the same transaction, so a shared lock that
upgrades deadlocks two concurrent readers. A data read rejects
`:lock-for-share?` for the same reason; only row-only reads accept it.
`sql:get-file-without-data` is a plain select now, not a CTE, which
makes locking legal without loading the data. Realization has a single
owner: `attach-file-data` only resolves and decodes, and `get-file*`
realizes once per path. The applied migrations no longer travel in the
`get-file-with-summary` response, cache hit and cache miss alike.
`embed-assets` reads each library without `:realize?`, so its
`:components` is an unloaded pointer map; `absorb-assets` walks it with
`vals`, `PointerMap.seq` calls `load!` and that raised with no loader
bound, so every export with `:merge-libraries` of a library with
components failed. The absorb runs under the library's own loader now.
The in-place binfile import (`import-file-and-overwrite*`,
`::overwrite`, `::bfc/file-id`) had no product caller and cannot carry
the previous `storage_object_id`, so overwriting a storage-backed row
would leave its object unreclaimed. `import-files!` rejects a cfg with
`::bfc/file-id` (`:not-supported`) and the helpers only that path used
go with it (`invalidate-thumbnails`, `*reference-file*`, the
`::overwrite` id branch of `index-object` and of `save-file!`, the
`on-conflict-do-nothing?` flag of the media insert).
Tests: a new `file-migration-read-test` pins the single read, the
locked re-read, the concurrent save interleaving, the double migration
and the read-only path; the upsert tests cover the return shape and the
stale pointer; the SQL recorder they share lives in
`backend-tests.helpers`, and the save, snapshot, restore and summary
sequences follow the new reads. The backend memories document the read
contract and the pointer.
AI-assisted-by: deepseek-v4.1-flash
* 🐛 Harden the get-file guard and skip its read-only transaction
The cfg guard of `get-file` was an `assert`, which `app.config` elides
when `:backend-asserts` is off, so a deployed binary accepted a bare
pool or connection while the docstring promised it raises. It throws
`IllegalArgumentException` now, the class the option guards of the same
namespace already use, and the existing test expects it instead of
`AssertionError`.
A read with `:read-only?` applies the pending migrations in memory and
writes nothing, so the transaction the migration write needs only costs
two round trips there. `get-file` runs it with `db/run!` instead: with
a cfg that carries a connection it still joins the caller's
transaction, and only the reads that would have opened their own skip
it. A lock option needs a read-write transaction from the caller,
because PostgreSQL rejects row locks in a read-only one; the docstring
states that.
Tests: the guard test expects the new class; a cfg marked
`::db/read-only` reads a pending migration without persisting it (the
config flag that fills that key is legacy, but the key is set by hand);
and a read-only read is pinned as not opening a transaction while the
same call without the option does. The backend memory documents the
split and the guard.
AI-assisted-by: deepseek-v4.1-flash
* 🐛 Remove svg extension when creating a component from a svg file
* 🐛 Normalize root names to always match the component name
* ✨ Remove any extension ignoring case, not only .svg
* 🔧 Add some fixes and checks
* 🔧 Improve rename detection
A failed render-wasm call traps without popping its shadow-stack frames.
Each failure leaked stack until it overwrote static data, and from then on
every call trapped as wasm-critical. h/call now saves the stack pointer
before the call and restores it on failure.
Closes#12977
Related to #12906
AI-assisted-by: claude-opus-5-5
* 🐛 Fix font upload crash and preserve variant metadata
opentype.js 2.0.0 returns no fontSubfamily name for fonts whose
Macintosh name table only holds custom records. That made
parse-font-weight/parse-font-style call re-seq on nil and crash the
upload.
A nil or blank variant is now treated as no match in
app.common.media, defaulting to weight 400 and style normal. The
variant parsing also accepts oblique as italic, and the new
parse-font-family helper derives the family from a filename by
stripping the known weight/style tokens, separating a style token
glued to a weight token ("Roboto-BoldItalic" -> "Roboto").
first-nonblank-string returns the first nonblank candidate, so a
blank preferredFamily/preferredSubfamily no longer shadows a valid
fontFamily/fontSubfamily.
app.main.data.fonts gains resolve-font-metadata and
prepare-font-entry. opentype.js values win per field; any missing
value falls back to the filename-derived family, weight and style.
The parsed and woff2 branches share the same helper, so metadata is
preserved instead of collapsing every face to weight 400, and
fonts with missing vertical-metric tables upload with no height
warning instead of failing.
process-upload reports per-file failures (unknown signature,
unparseable file, preparation error) with a single bad-font
notification instead of silently dropping them, and never errors
for per-file failures.
The upload now sends :variant-name to create-font-variant, which
accepts it as an optional param (ships in 2.20), so the foundry
name survives the upload.
Closes#10436
AI-assisted-by: deepseek-v4.1-flash
AI-assisted-by: muse-spark-1.3-contributor
Signed-off-by: CHOI JUNSOO <junsoo1172@gmail.com>
* 🐛 Parse glued font style tokens in the font parsers
EvaMarco's review on #11951: parse-font-family and
parse-font-weight handled glued tokens like "BoldItalic", but
parse-font-style did not. For names such as
"opensans-bolditalic-webfont.woff2" the three parsers disagreed:
the italic face resolved to style normal and collided with the
bold-only file on the (family, weight, style) join key. One
binary silently overwrote the other during upload, losing one
font face without any warning.
A new private normalize-font-name helper is applied by all three
parsers. It splits a style token glued to a weight token, now
with a lookahead boundary so "LightItalic_v2" splits too, and
strips a trailing technical suffix (-VF or _v<number>).
parse-font-family, parse-font-weight and parse-font-style now
interpret the name consistently.
Tests cover the three cases from the review, the bold-only
control case, and an end-to-end process-upload check that a bold
and a bold-italic WOFF2 survive as two variants.
Relates to #10436 and PR #11951.
AI-assisted-by: inkling-2-small-free
* Remove duplicate frontend test registration
---------
Signed-off-by: CHOI JUNSOO <junsoo1172@gmail.com>
* 🎉 Add UI for toggling bullet lists (under config flag)
* 🎉 Store list-style-type as a paragraph attribute
The bullet list toggle now saves list-style-type ("none" or "disc") on
the selected paragraphs and reads it back. A paragraph without the attr
counts as "none", so a bullet paragraph next to an older one shows as
mixed.
WASM keeps the value on each paragraph: the paragraph bytes grow from
12 to 16, Enter keeps it on the new paragraph, and the v3 editor reports
it in the current styles, so the sidebar shows it while editing too.
Refs #12630
AI-assisted-by: claude-opus-5-5
* ✨ Draw bullets for list paragraphs
A paragraph with list-style-type "disc" now gets a solid round bullet,
sized as Chrome draws a CSS disc marker (measured with Source Sans Pro):
a radius of 0.16 font sizes, centered 0.32 font sizes above the first
baseline, with a gap of 0.33 font sizes before the text. The bullet
hangs outside the text box, before its start edge (after it for RTL),
like CSS list-style-position: outside, so the text layout, caret and
selection stay as they were.
The bullet takes the paint of its line's first style run, so fills,
strokes, shadows and masks apply to it as they do to the glyphs. An
empty item, which has no run, borrows the previous item's paint. The
cached paint, the text extrect and text-to-path share one walker over
the paragraphs, so they always agree: the extrect grows to hold the
bullets, like shadows and blur, and flattening a text keeps them.
Refs #12630
AI-assisted-by: claude-opus-5-5
* ✨ Show the list style in Inspect
Text paragraphs with list-style-type "disc" now show a "List Style" row
in the Inspect Styles and Computed views, after Text Transform. Its copy
action gives "list-style-type: disc;". Text without a list shows no row.
The list item check lives in common as list-paragraph?, so every place
that needs it shares one rule.
Refs #12630
AI-assisted-by: claude-opus-5-5
* ✨ Generate list HTML and CSS for bullet paragraphs
In the Inspect code, neighbouring paragraphs with list-style-type
"disc" become one <ul> with an <li> each, keeping the paragraph's class
and dir. Other paragraphs stay <p>. Only code mode groups them; text
rendering still uses <p>.
The CSS sets list-style-type on the items and resets the list's margin
and padding, so the browser hangs the bullet before the text as the
canvas does. Paragraph rules set font-size: 0, so each item also gets a
::marker rule with its first span's font size and text color. A
::marker can't paint a gradient, so a gradient fill gives its first
stop.
Refs #12630
AI-assisted-by: claude-opus-5-5
* ✨ Add unit tests for the token name schema
Cover make-token-name-schema, the schema that rejects a token name whose
path clashes with an existing token or group. The lower-level
token-name-path-exists? already had a test; the schema that composes it
did not.
The cases pin an exact duplicate, a token inside a token, a token name
that collides with a group, an existing child path, valid siblings, the
fail-closed nil tree, invalid syntax and the length limit. This is the
unit-test base for reducing the 15 group-name-clash e2e cases to two.
AI-assisted-by: deepseek-v4.1-flash
* ♻️ Remove duplicate token tests and fix the spread step
The three top-level tests duplicated the ones inside the creation
describe; only the indentation differed. Keep the originals.
The shadow token test labelled "Valid spread" filled the blur field, so
the token under test ended with blur 0 and the spread never changed.
Fill spread instead and expect the referenced token to resolve with
Blur: 4.
AI-assisted-by: deepseek-v4.1-flash
* ♻️ Reduce the token group-name-clash e2e matrix to two tests
The fifteen cases repeated the same setup and only changed the token
type and value. The rule is now covered by the token name schema unit
test, so keep two representatives: thirteen simple types in one test
with a test.step each, and the two composite forms (Typography and
Shadow) in a second test.
The helper no longer re-runs the file setup, so each test loads the
page once: 15 page loads become 2 and the block drops from 433 s to
about 21 s.
AI-assisted-by: deepseek-v4.1-flash
* ♻️ Cover the numeric token creation types in one test
The nine creation tests repeated the same validation ladder and only
changed the token label and placeholder. Run the ladder once in the
border radius step, keep the reference token creation, and create the
other eight types with a test.step each.
The new helpers split the flow: open the form, assert the ladder, save
a token. The e2e drops from nine page loads to one.
AI-assisted-by: deepseek-v4.1-flash
* ♻️ Reduce the remapping e2e spec to three tests
Keep one remapping flow per token shape: box shadow with alias
references, border radius with a value change, and a group rename with
undo. The two cancel paths lose their own page loads and become steps
inside the shadow test: close keeps the original name, and "don't
remap" renames the token and breaks the reference.
The deleted typography and shadow applied-shape cases exercise the same
data path, which the remapping unit tests already cover.
AI-assisted-by: deepseek-v4.1-flash
* ♻️ Shrink the tooltip and numeric-revert e2e tests
The tooltip test asserted five message shapes over eleven layer
selections. Keep the three that have no unit test: a deleted token, a
reference outside the active sets, and a token in an inactive set. Each
is one step with a pill label and its tooltip text.
Three of its five messages were already covered by the token pill unit
test, including both resolved-value cases.
Keep one numeric-revert test and drop the fixed sleeps: await the value
with toHaveValue instead. Division by zero and the min/max clamp are
unit-tested in the numeric input tests.
AI-assisted-by: deepseek-v4.1-flash
* 🐛 Preserve HSL values while editing lightness
Keep HSL editing components through black, white, and grayscale.
Use the same draft for numeric inputs, sliders, and gradient previews.
Strip the draft before emitting color data and reset it on other edits.
Closes#12615
Signed-off-by: makesomethingshit <junsoo1172@gmail.com>
AI-assisted-by: GPT-6
* 🐛 Keep HSL drafts consistent across gradient and ramp edits
Match RGB components to HEX so gradient echoes preserve HSL drafts.
Discard stale HSL values when editing Ramp hue, saturation or value,
while preserving them for opacity-only changes.
Cover grayscale gradient echoes and saturation restoration.
Signed-off-by: makesomethingshit <junsoo1172@gmail.com>
AI-assisted-by: GPT-6
---------
Signed-off-by: makesomethingshit <junsoo1172@gmail.com>
The 0033-normalize-main-instance-names migration walks the shape tree
following the :shapes vectors. When a parent listed a child id missing
from the container :objects map (a dangling reference), the traversal
stored the nil child into :objects, creating a nil key that made
check-file-data reject the whole file: get-file answered 500 "invalid
file data" and the file could not be opened or migrated further.
Fix it in two parts:
- 0033 returns the container untouched when a listed child does not
exist, so the traversal never stores a nil shape.
- The append-only migration 0035-fix-dangling-children prunes from
every :shapes vector, in pages and components, the ids absent from
the owning container :objects map, logging each fix at trace level.
It mirrors 0032-fix-children-not-in-parent, which repairs the
opposite inconsistency.
Both parts are needed: 0033 runs before the repair migration, so
without the guard the same file fails again in the same migrate call.
Add regression tests for the traversal guard and for the repair
migration, and record in mem:clojure/logging the convention that
:hints start lowercase.
Closes#12936
AI-assisted-by: qwen3.8-flash
When components whose name matches the variant container are moved into
it, removing the container name leaves an empty value. Combining several
components named the same produced "Value 1" for the first variant and
empty values for the rest, so variants shared the same combination.
Assign the first unused "Value N" to the last property when all computed
values are blank, tracking values given within the same batch.
Combine moves the remaining mains into the container one at a time, in
the combine order, so the generated values follow that order instead of
the layer order.
Closes#12878
* ⚡ Send objects-map pages to the worker without a second encode
When switching between pages on a large file e.g Material Design Icons, there's
some UI freeze from :index/initialize encoding object map shapes to send to the
worker. An objects map already stores the encoded JSON for each shape, and
`encode-str` was wrapping those strings in a second layer of JSON on the main
thread.
We can just transmit the already encoded JSON in a js/Map alongisde the
payload, similar to :blob, so structured clone just copies the strings verbaitm
without re-encoding. To do this this PR adds two functions to ObjectsMap that
converts to and from a js/Map.
The "Main components" page of the Material Design Icons has ~30k objects, so
this reduces the freeze when switching pages from 306ms -> 189ms (1.63x) on a
release build.
AI-assisted-by: claude-sonnet-5-5
* Split property tests out into new deftest
---------
Signed-off-by: Luke Lau <luke_lau@icloud.com>
Co-authored-by: Andrey Antukh <niwi@niwi.nz>
* ✨ Cache the transit reader when decoding strings
On the cljs side, we create a new transit reader on every call to decode-str.
This turns out to have a fair bit overhead in both transit-cljs and transit-js,
since it has to merge the handlers, call clj->js, allocate a new ReadCache,
etc.
By reusing it the measured decode time goes from ~58us -> ~41us, so about 1.4x
faster. This particularly helps for some operations like whenever a component
is edited and every object needs to be decoded, see
https://github.com/penpot/penpot/issues/9344, although they could probably be
made more efficient in other ways.
The reader usually clears its read cache after a successful read, but whenever
it throws it hangs around. So in this case we need to reset the cache
ourselves, otherwise the next read will be from a clobbered cache.
The tests were written by Claude Sonnet 5.5, and should cover the clobbering
behaviour above.
Signed-off-by: Luke Lau <luke@igalia.com>
* Keep one transit reader per type and rebuild it on add-handlers!
---------
Signed-off-by: Luke Lau <luke@igalia.com>
Co-authored-by: Andrey Antukh <niwi@niwi.nz>
* ⚡ Compute text fonts once per shape when uploading to wasm
* ⚡ Write the shapes batch in place and coerce fills once
* 🐛 Free the upload buffer when a batch record write fails
* 🐛 Normalize stored SVG attribute keys to kebab-case
The v3 reader already emits kebab-case keys, but birth stored the
camelCase React prop names, so import and export never converged:
fill rules painted wrongly and filter defs lost their blur.
Birth now normalizes every stored attribute key to kebab-case
through the `svg-props` whitelist (`attrs->kebab-props`), applied on
all element paths of the svg importer, on the file builder and on
defs. svg-raw `:content` is kebab-ized in spelling only and keeps
unknown keys, because the sidebar reads them. Non-spec attrs Penpot
must keep live in `penpot-extra-attrs`, seeded with `:stroke-style`.
Render boundaries translate back to camelCase for React and for the
wasm props, and the shape schema declares the closed kebab contract
derived from that same whitelist, so birth and validation cannot
drift.
Migration `0035-fix-shape-svg-attrs-and-defs` converts old files
once with the birth transforms, and carries back the svg-viewbox
rect normalization the withdrawn `0021` had. That old id leaves
`available-migrations`, so files recording it migrate again through
this one, and the binfile cleaner drops its own svg repair.
Review fixes included here: `setup-opacity` and the blend mode read
the stored keys, gradient stops accept `:style` as a CSS string or a
map, `inherit-attributes` reads `:style` so a child's inline style
beats the inherited group attr per CSS, and the svg-attrs menu shows
spec spelling (`viewBox`, `class`) through `stored-attr-display-name`.
Tests cover the migration, the closed svg-attr schemas and the
import/export round trip; the svg-content schema test pins its
sample seed, which dealt no svg-raw shape on unlucky runs.
Closes#11947
AI-assisted-by: muse-spark-1.3-contributor
* 🐛 Keep legacy svg node tags and transforms valid in migration 0035
The closed svg node schemas require keyword tags and map attrs, but
stored files hold def and svg-raw content tags as strings, and some
node `:attrs` as nil. Migration 0035 left them as they were, so the
file data check that runs after migrating rejected those files and
they could not be opened. `normalize-node-tree` now turns string tags
into keywords and nil attrs into an empty map, at birth and in 0035.
With keyword tags, the svg defs renderer applies the shape transform
to user-space defs. Legacy `:svg-transform` values are plain maps,
which multiply into NaN matrices. `gmt/coerce` turns them into
matrices, in 0035 and in the binfile cleaner, because every v3 import
decodes the transform as a plain map again. Legacy masks and gradients
now render as designed in the svg renderer.
The binfile test helper calls `export-files!` with the current config
keys.
AI-assisted-by: claude-opus-5-5
---------
A component whose :main-instance-id points to a missing shape fails
validation on its copies with :ref-shape-not-found. A repair scoped to
the copy's page never sees the main's own error, so it detached the
copy, even when the main was still on the component's page.
The repair of :ref-shape-not-found now looks for the component's main
on its page first and points :main-instance-id back to it, which keeps
the copy linked. A copy is still detached when no main is found.
update-file checked the schema of the whole file on every save, which
loads and decodes every page even when the changes touch one. On files
with several pages that was most of the time of a save.
validate-file-schema-affected! checks the file data with the pages index
reduced to the pages the changes modify; the data outside pages is still
checked whole, and a change type it does not know makes it check the
whole file. update-file uses it for the strict and the soft schema check.
A save of 10 commits on a file of 10 pages of 2k shapes goes from ~1.1 s
to ~240 ms; a single page file costs the same.
A save whose changes failed validation was rejected as a whole: autosave
stopped, the edits made after it piled up unsent, and a reload lost
them all, the valid ones included.
The frontend now sends each commit of a save as its own group with its
commit id (:changes-with-metadata). When applying and validating all of
them at once fails, the backend searches by bisection for the first
commit that fails, keeps the commits before it and answers with
:rejected (that commit and every later one). Only the kept changes are
stored, snapshotted and broadcast, so other users never see the rejected
ones. Flat :changes and groups without ids stay all-or-nothing; database
and network errors still fail the whole request. A rejection without an
error type is reported as an internal error.
The frontend rolls the file back to what the server holds: it undoes
every pending commit of that file, applies again the remote changes that
arrived meanwhile, and redoes the kept commits. Undoing only the
rejected commits would write stale values over a peer's later change to
the same attribute. The rejected and later commits leave the queue, the
buffer and the undo history, and a warning that hides after ten seconds
tells the user. The pure bookkeeping lives in
app.main.data.persistence.commits.
The pending notice of a commit the rollback dropped no longer leaves the
save status at pending, which made "Pin this version" hang and fail.
Debug tools drive these cases by hand: debug.commitInvalidChange() makes a
save fail validation, holdSaves() delays the answers, failNextSave() and
failNextRollback() fail them on purpose (backend flag debug-save-faults),
and persistenceState() shows the queue.
* ⚡ Keep only the shape data the worker reads
The worker kept a full copy of the current page for its selection and
snap indexes. It now keeps only the attributes the indexes and the
change replay read, and re-indexes only the shapes that changed.
On a 52k-shape page the worker heap goes from 698 MB to 286 MB.
Closes#12740
AI-assisted-by: claude-opus-5-5
* ♻️ Share the worker re-trim with its replay test
The worker and its replay test now share `retrim-objects`.
AI-assisted-by: claude-opus-5-5