* ✨ Stabilize fragmented file data across snapshots and GC
Snapshots, automatic and user, now store pointer-based rows that share
fragments with the live file instead of duplicating page content.
`get-file-snapshot` resolves those pointers on the server so version
previews keep rendering plain data, and `restore!` reuses the shared
rows. Export still materializes (the package carries no fragments) and
import regenerates them.
`file-gc` treats the main file and every unpurged snapshot as liveness
roots, reads the file unrealized, and reattaches the original pointer
wherever the cleaned value is unchanged, so a pass over an untouched
file writes no fragments and a pass over an edited file keeps the ids
of the pages it did not touch. Snapshot creation and the restore
backup take the file row `FOR SHARE`, which serializes them with the
pass and closes the race that let a new snapshot reference fragments
the pass was about to mark.
All backend pointer bindings go through the `fdata/with-pointers`,
`with-tracked`, `with-tracked-pointers` and `with-untracked` macros.
The loader reads fragments a previous pass marked until `objects-gc`
purges them, and the frontend fails loudly with retry instead of
dropping an unresolved fragment.
Review fixes in this branch: the automatic snapshot attaches before
pending migrations, `realize-pointers` is the single verb,
`reattach-unchanged` takes the identity fast path, the restore guard
takes its revn from the backup read, and `PointerMap.withMeta`
generates a fresh id when metadata changes.
Tests cover both snapshot writers (red without the lock), snapshot
sharing and pin release through the real delete and purge paths,
reattach against orphan marking, preview after offload, the pointer
binding contracts and the with-meta id change.
Closes#12410
AI-assisted-by: muse-spark-1.3-contributor
AI-assisted-by: deepseek-v4.1-flash
♻️ Pin no-churn behavior for untouched pages
Pin the behavior the churn verification measured: a save keeps the
fragment id of every page it does not touch, with a page of eleven keys
(PersistentHashMap), a 46-shape objects map and a nine-page index (hash
map). Also pin the mixed case: a data-level objects-map flip rewrites
the old pointer pages once and the next save keeps the new ids.
The memory records the identity rule behind it: re-assoc with the same
value returns the same map in array and hash maps alike.
Closes#12410
AI-assisted-by: deepseek-v4.1-flash
⚡ Stop loading untouched pages when enabling objects-map
`enable-objects-map` wrapped every page on every save, and reading a
page's objects forced its PointerMap to load the fragment. A save of
one page in a seven-page file therefore made seven fragment reads.
Skip unloaded pointers: their fragment was already wrapped by the
encode-file that wrote it, so loading it only to wrap it again is a
no-op. The save now loads only the pages it edits. A page with a plain
fragment (a data-level feature flip) keeps its plain fragment until it
is first loaded or edited; a migration that enables the feature must
rewrite the data itself.
Cover the change with load-count tests for save and restore, update the
flip test to the new contract, and drop the stale restore comment.
Closes#12410
AI-assisted-by: deepseek-v4.1-flash
⚡ Keep page fragment ids through pending migrations
Applying a pending migration on read realized the file, ran the
migration and persisted it again; the persist wrapped every plain page
in a new PointerMap, so every page got a new fragment id even when the
migration did not change it. That leaves the fragment cache cold and
rewrites every fragment on the first read after a deploy.
Capture the original pointers before realizing, reattach the unchanged
positions after migrating (same value comparison file-gc already uses),
persist the reattached file and return the realized one: the read
contract is plain data. Read-only connections skip the persist and the
reattach. A migration that adds objects-map skips the reattach so the
pages get wrapped.
Move the three helpers to fdata so file-gc and the migration path share
them, and cover the path with a real 0031 migration test: the unchanged
page keeps its id, the changed page gets a new one, the caller gets
realized data and a second read does not migrate again.
Closes#12410
AI-assisted-by: deepseek-v4.1-flash
* 🐛 Realize library pointers with each library's own loader
Fragment rows are keyed by file id, and every caller binds the
fragment loader of the file it works on. get-resolved-file-libraries
returned libraries with their pages still as pointers, so the first
read of a library page under fdata/pointer-map asked the consuming
file for the fragment and failed with :fragment-not-found. That
broke migrations and file validation for every file using a
pointer-mapped library.
The weak-map loader now realizes each library's pointers with the
library's own id. The test fails without the change.
AI-assisted-by: claude-opus-5-5
Signed-off-by: Álvaro Tejero Cantero <alvorithm@teje.ro>
* 🐛 Restore the pointer-map require the repair path still needs
The branch dropped `app.util.pointer-map` from the namespace while the
auto-repair path (added to develop before this branch) still rebinds
`pmap/*tracked*` around validation and repair in
`repair-preexisting-damage` and `process-changes-and-validate`. The
namespace did not compile, so the whole backend failed to load.
Those toggles do not fit the fdata macros: `with-untracked` binds nil
and `with-tracked` creates a fresh atom, while the repair path has to
turn tracking off and restore the existing atom (and read it before the
repair starts). Restore the require and keep the bindings explicit.
Also pin the feature-off path with a test: with `fdata/pointer-map` off
(the default), a save creates no fragments, file-gc writes nothing and
the snapshot preview still returns plain data. Drop a stray `\n` char
literal left at the end of one deftest.
Update the memories with the exception to the macro rule, the snapshot
FOR SHARE locking and the pre-migration automatic snapshot.
Closes#12410
AI-assisted-by: deepseek-v4.1-flash
* 🐛 Validate debug files with their own pointer loader
The `/dbg/actions` file validation passed the file id where
`get-resolved-file-libraries` expects the file map, so the library map
was seeded with a nil key. It also validated a file read without
realizing it, so the first page lookup of a pointer-mapped file raised
"load is not supported when *load-fn* is not bind" and the endpoint
could not validate exactly the files the feature touches.
Read the file realized (validation walks every page and component
anyway) and build the library map from the file. The library loader
already resolves each library's pointers with its own id, so the
realized file and its libraries validate as plain data. Covered by a
handler-level test that fails before the change.
AI-assisted-by: deepseek-v4.1-flash
* ♻️ Bind the snapshot loader through the fdata helper
`reduce-snapshots` hand-rolled the same `binding` that
`fdata/with-pointers` builds (it also loaded the function's own
pointer loader before entering the reduction). Use the macro so the
rule holds there too and the namespace no longer needs
`app.util.pointer-map`; the loader is identical (same cfg and file-id)
and snapshot walking behaves the same.
AI-assisted-by: deepseek-v4.1-flash
* 📚 Note the pointer-binding exceptions and snapshot ordering
Record what the feature review settled:
- Two direct `binding`s stay on purpose: the fine-grained `*tracked*`
toggles of `app.rpc.commands.files-update` (validation/repair turn
tracking off and restore the existing atom) and
`binfile.common/encode-file` (fresh tracked set inheriting the
current one). Atom-scoped macros would only be worth it if the set
grows.
- The automatic snapshot is taken before the pending migration, and
the pre-migration window is narrow: the `get-file` RPC migrates on
read and its ETag hashes the available migrations, so a post-deploy
read cannot be a 304 from the pre-deploy state. The old-format
snapshot only appears when the save is the first request to touch
the file after the deploy, and `get-file-snapshot` previews it as
stored.
AI-assisted-by: deepseek-v4.1-flash
* 🐛 Apply pending migrations when previewing a version
Automatic snapshots are attached before the save applies pending file
migrations, so a version can store the old format. The preview RPC
handed that data to the UI as stored, and the viewport renders it with
the current code.
Realize the snapshot with the live file id (fragments are keyed by
file, not by snapshot), normalize its migrations to a set (the decoder
keeps the pgarray as a vector, and `set/difference` would compare
indexes) and apply the pending migrations with the pure
`fmg/migrate-file` and the live libraries. Nothing is persisted: the
stored version keeps its data and restoring it re-migrates on the next
read. The response now carries the snapshot's migrations.
Use `:features` the way `get-file` does for `fdata/objects-map`: a
client that does not declare it gets plain objects. The historical
feature set is not checked against the enabled one on purpose: a
version may predate a feature change.
Cover the preview of a snapshot with a pending migration and the
objects realization; both pin that the stored row is untouched. Update
the file-data memory with the preview contract and the vector/set
trap.
AI-assisted-by: deepseek-v4.1-flash
* 🐛 Decode snapshot migrations as an ordered set
`file-snapshots/decode-snapshot` handed the `file_change.migrations`
pgarray as a vector. `set/difference` falls back to index lookups when
the first collection is shorter than the second, so the pending set was
right only by accident: a stored subset happens to work, while a
stored set larger than `available-migrations` (a snapshot taken by a
newer build) marks every available migration as pending and re-applies
it to already migrated data.
Decode it as a `d/ordered-set`, the type the rest of the code and
`schema:file` expect. This also covers `restore!` and the snapshot
list, and it lets `get-file-snapshot` drop the `(into #{} ...)`
normalization it was carrying. Re-wrap the union in `fmg/migrate` so
the invariant holds even when a caller hands a plain set.
A snapshot with NULL `migrations` predates SQL migration 0140 and
keeps the format version of its time. Pin that legacy path: the
preview regenerates the applied set from that version, migrates in
memory and leaves the stored row untouched. Update the file-data
memory with the invariant and the legacy path.
AI-assisted-by: deepseek-v4.1-flash
* 🐛 Fix flaky waiting-gauge assertion in climit test
The throwable-while-waiting test asserted the exact list of
waiting-gauge publications, but the component sampler republishes
the same count when its first tick lands inside the recorded
window under load. Assert only the last value, which is the
contract the helper documents.
AI-assisted-by: muse-spark-1.3
---------
Signed-off-by: Álvaro Tejero Cantero <alvorithm@teje.ro>
Co-authored-by: Álvaro Tejero Cantero <alvorithm@teje.ro>
* ✨ Rework the import dialog
The import wizard paints a Check/Upload/Import stepper derived from the
dialog status, plus a phase title, a description and a progress bar per
phase, and its rows follow the new design: folder icon, file name and a
size subtitle. While checking, the dialog shows no rows and no spinner,
because the phase header already carries the message.
The wizard writes the job milestone once, in the header
(`jp/current-milestone-text`), since every file of a container reports
the same one. A row keeps its name and its size, says "Ready" once its
container finished uploading, and shows the error text when it fails;
the button that drops a file only appears while the wizard still offers
its import step. The `jobs.queued` label now reads "Waiting job slot".
The bar measures the whole import: determinate while the containers are
uploaded, weighted by their size (`imp/upload-pct`), and animated
otherwise, because the stages of a job are not comparable. The phase
title counts the files, and `imp/imported-count` counts them done with
the job's own counter, so the number moves while a container rebuilds
its files one by one.
The analysis reads the entries of a version 3 package in one pass: a
file shows what it occupies inside its container, its own entries plus
the storage objects its media and thumbnails reference, while what the
container weighs stays in `:container-size`, the size the upload bar is
weighed on. The ready step waits for the end of the analysis
(`imp/ready-to-import?`), not for every known file, so Continue never
leaves out the files that arrive later.
Tests cover the file sizes, the upload percentage, the done counter,
the wizard step and the milestone in course. The change also adds
`-XX:NativeMemoryTracking=summary` to the backend devenv so `jcmd`
shows native memory, and documents the wizard, the promise call styles
(`mem:clojure/promesa`) and two CSS pitfalls: a logical `inline-size`
wins over an inline physical `width`, and Rumext leaves a plain
`:style` map raw, so inline styles leave CLJS as `#js` objects.
AI-assisted-by: glm-5.3-flash
AI-assisted-by: deepseek-v4.1-flash
* ✨ Rework the export dialog
Every exported file is its own job, so each row reports its own state:
folder icon, name and a size-plus-`.penpot` subtitle; waiting for a
worker; running with the milestone of the job and an animated bar; or
exported with a tick. A failed row paints red, takes the close icon and
shows the hint of the job, or a generic label, through
`files-export-modal.export-error`.
Success, failure and cancellation replace each other, so a row never
reports two outcomes at once. The ids of the messages are canonicalized
to string before they match a row, because they cross serialization
boundaries. The export result now carries `:size`, so the subtitle says
the format until the job produces the artifact and the size once it
exists; a transport failure builds a synthetic error, so a nil
`ex-data` no longer leaves the row hanging.
The header counts the files and says whether the export runs or ended;
since every file has its own job, it paints no global bar.
Closing changes too: the Close button is gone. While files are in
flight the footer offers Cancel export and the X cancels as well, which
marks the loading rows cancelled and leaves them in place, so their
outcome can be read; once every row reached a terminal state the footer
offers Done, which dismisses the dialog.
The retry button of the finished design is left out, because the client
has no retry mechanism yet. Styles come along: the rows scroll while
the header and the footer stay put, and the new phase header, row and
bar follow the tokens of the import wizard. The title key becomes
`files-export-modal.title`, and the new keys cover the header, the row
states and the buttons.
AI-assisted-by: muse-spark-1.3
* ♻️ Inline the binfile job adapters into their task handlers
The wrappers in app.binfile.jobs only forwarded to the core, so each
task handler now composes its own run: frozen cfg, active check,
progress bridge and format dispatch. export-cfg freezes the whole
export (file-ids, export-type, output) and v3/export-files! takes it
as its single argument; ::bfc/ids becomes ::bfc/file-ids. The bridge
keeps one job: bfj/run forwards the cfg to its callback, and
check-active answers with it to thread the run.
AI-assisted-by: muse-spark-1.3
* 🐛 Migrate binfile tests to the single-arg export-files!
The refactor made v3/export-files! take a single cfg carrying
::bfc/file-ids and ::bfc/output, but six calls in binfile_test
(added by the tokens handling change) still used the old two-arg
form with ::bfc/ids. Move them to the new form, as the rest of
the file already does.
AI-assisted-by: muse-spark-1.3
* 🐛 Fix export artifact expectations to include file size
The export flow now reports the artifact size, but the stale
expectations dropped it and failed the frontend suite. Include the
size in both assertions so they match the emitted message.
AI-assisted-by: muse-spark-1.3-contributor
* 🐛 Migrate remaining export-files! test callers to the single-arg form
rpc_binfile_test and offload_file_data_test still passed the output apart
and used ::bfc/ids; move both into the cfg as ::bfc/file-ids/::bfc/output.
* 🐛 Hide the downloads hint when no file was exported
After Cancel or when every job failed, the export header said
"0 files exported · Check your downloads folder."; drop the hint then.
* ♻️ Read the object of a container entry once
The index of a container asked twice for the object of the same zip
entry: once to test it and once as the key of the update. Resolve it
once, in the branch that owns it, so the entries of a file never pay
for a question that does not apply to them.
Addresses a review note on #12192.
AI-assisted-by: deepseek-v4.1-flash
* ♻️ Never size an import row with its container's weight
A row of a version 3 package took the size of the container whenever the
container held no entry for its file, while the comment beside it
promised the opposite. An unknown size now stays nil, which the row
already understands: it paints no subtitle instead of a number that
belongs to the whole package.
The test carries a file the manifest lists and the container does not
hold, and the memory records what the row does then.
Addresses a review note on #12192.
AI-assisted-by: deepseek-v4.1-flash
* 🐛 Retry pool stats read until snapshot is consistent
Hikari exposes each pool counter through a separate MXBean call, so a
borrow or return between the reads can leave one snapshot off by one.
The test read once and failed in CI with total 2 vs active + idle 3.
Read up to 20 times with a 10ms pause until total equals active +
idle, then assert on the stable snapshot. The invariant stays strict;
only the transient race is tolerated.
AI-assisted-by: muse-spark-1.3-contributor
---------
Co-authored-by: Elena Torro <elenatorro@gmail.com>
Add a per-process availability flag (app.health) that /readyz
reports as 503 OFFLINE while offline, so the load balancer
drains the instance and in-flight requests finish untouched.
Operator surface is app.srepl.main/set-offline and
app.srepl.main/set-online; a restart comes back online.
Covered by backend-tests.health-test; HA docs updated.
AI-assisted-by: muse-spark-1.3-contributor
* ✨ Enforce idle and absolute session expiration
Sessions now expire on two server-side conditions: an idle window
(PENPOT_AUTH_TOKEN_COOKIE_MAX_AGE, default 7d) and an absolute cap
from creation (PENPOT_AUTH_TOKEN_COOKIE_MAX_AGE_ABSOLUTE, default
30d, enforced by the token :exp claim). The daily session-gc task
deletes rows that exceed either window, so idle sessions can no
longer be replayed and active sessions are not deleted at the idle
window.
Also remove the legacy v1 HTTP sessions: the http_session table and
the string-id / :ver 0 token code paths are gone. Any v1 cookie now
requires a fresh login.
Document the session expiration configuration in the technical guide
and add a backend memory describing the token, renewal and GC model.
Closes#11646
AI-assisted-by: deepseek-v4.1-flash
* 🐛 Address session-expiration review findings F1-F4
Fix the unreadable test (a stray paren broke whole-suite
discovery), enforce idle expiration on every request in
wrap-authz, fail boot fast when the absolute cap sits below
the idle window, and align config defaults with the memory
rule while fixing its migration number and stale reference.
Closes#11646
AI-assisted-by: muse-spark-1.3-contributor
* 🐛 Bound decompressed size of JSON entries on binfile v3 import
Every JSON/text zip entry (manifest, files, pages, shapes, colors, components, typographies, tokens, plugin-data) was decompressed without any size limit, letting a small .penpot archive exhaust the backend heap (GHSA-qcw7-v626-g6cf). Only binary storage blobs were guarded. Reuse the existing size-limiting-stream guard on the text path: 20 MiB cap per entry, 200 MiB cumulative budget per import job, plus a cheap declared-size pre-check. Both limits are configurable and wired through the binfile, management and debug entry points. Adds zip-bomb regression tests for the file entry, the synchronous manifest read and the cumulative budget. Closes#11606
AI-assisted-by: muse-spark-1.3-contributor
* ♻️ Uniform binfile import limits behind init-limits
Move the binfile import limits to a single source of truth in app.binfile.common (default-* vars) and drop the duplicated entries from config/default; env overrides keep working through the schema. Resolve all limits once per job with init-limits (::max-size, ::total-max, ::current-size, ::max-object-size, ::max-zip-entries) instead of rebuilding the map per zip entry. Thread cfg as the first arg through the v3 readers, collapse read-plain-entry into read-entry, and give size-limiting-stream a single explicit-counter arity. v1 keeps using the compiled default (mechanical rename only). No behavior change. AI-assisted-by: muse-spark-1.3-contributor
* ♻️ Rename binfile limits to text-entry/binary-entry terms
Use text-entry/binary-entry vocabulary consistently across config keys, bfc input keys, default-* vars and the limits resolved by setup-limits (::max-text-entry-size, ::max-text-total-size, ::current-text-size, ::max-binary-entry-size). Rename init-limits to setup-limits. No behavior change. AI-assisted-by: muse-spark-1.3-contributor
* ♻️ Rename cumulative text counter and document binary limit
Rename ::current-text-size to ::accumulated-total-text-size for clarity and expand the default-max-binary-entry-size comment to match the other limit vars. No behavior change. AI-assisted-by: muse-spark-1.3-contributor
* ♻️ Harden binfile guards and prove budget accumulation
Add a regression test that only passes when text bytes accumulate across entries (budget between largest entry and summed total; verified red against a per-entry atom). Include the entry name in streaming-guard errors, count skipped bytes against the budget with a direct unit test, and forward all four limit keys in get-manifest. No behavior change. AI-assisted-by: muse-spark-1.3-contributor
* ♻️ Thread cfg through get-manifest
get-manifest now takes the caller cfg and resolves limits with setup-limits like the import job itself, instead of building a single-use mini-cfg from cf/get. No behavior change. AI-assisted-by: muse-spark-1.3-contributor
* 🐛 Add configurable limits for ZIP entry count and object size in v3 import
Add binfile-import-max-zip-entries (default 500,000) and
binfile-import-max-object-size (default 100 MiB) config entries.
Both are configurable via PENPOT_BINFILE_IMPORT_MAX_ZIP_ENTRIES and
PENPOT_BINFILE_IMPORT_MAX_OBJECT_SIZE env vars.
Entry count is checked before processing begins. Per-object size is
checked after each storage object content is resolved.
AI-assisted-by: mimo-v2.5-pro
* 🐛 Enforce actual decompressed byte limits on v3 import
The previous object-size check trusted the ZIP entry header's declared
size (ZipEntry.getSize()), which a malicious zip-bomb can forge. The
check would pass, then the full decompressed payload would be read
anyway during hashing and storage persistence.
Add size-limiting-stream, a FilterInputStream wrapper that counts
actual bytes read and raises :validation :max-file-size-reached when
the configured limit is exceeded. Wire it into zip-entry-storage-content
so both the hash calculation and storage write paths are bounded by
real decompressed bytes, not declared header size.
Also wire import limits into management.clj (clone-template) and
debug.clj (import-handler + clone path) for defense-in-depth, and
add a test that exercises the object-size limit with a real storage
object in the exported ZIP.
AI-assisted-by: mimo-v2.5-pro
* ✨ Add export job model, store and scheduler to exporter
* ✨ Render wasm exports on pooled worker threads
* ✨ Add export job REST API to exporter
* ✨ Use export job API and allow cancelling wasm exports
* 🔧 Show export jobs in the internal debug panel
* 🔧 Pass flags and export job settings to the exporter container
* 📚 Document the exporter job API and its redis layout
* 🎉 Basic lbug connection for ingestion
* ✨ Add Penpot-to-Ladybug graph ingest vertical slice
* ✨ Use embedded Ladybug Java API instead of CLI
* ♻️ Share Ladybug connection across ingest and stats
* ✨ Validate graph ingest projections with Malli
* ✨ Project nested shapes recursively into the graph
* ⚡ Load graph ingest via Ladybug COPY bulk import
* 🐛 Fix graph COPY ingest for multiline text names
* ✨ Add Ladybug graph export to debug UI
* ✨ Add debug graph console for in-memory Cypher queries
* ✨ Add live file-change feed to debug graph console
* ✨ Incrementally sync debug graph from Penpot file changes
* ✨ Handle mov-objects in debug graph sync
* 🐛 Fix batch delete sync and keep graph console feed alive
* ♻️ Derive graph node schema from Malli registry
* ✨ Add G6 graph view to debug graph console
POC per work/g6/plan.md. New /dbg/actions/graph-data exports the in-memory Ladybug session as plain JSON (per-table node queries + multi-table IsChildOf match, row cap 100k with truncation flag). Console page renders it with AntV G6 v5 (jsDelivr CDN, antv-dagre BT layout, color+glyph per node table, validated palette) and refetches debounced on live :file-change messages.
Signed-off-by: Álvaro Tejero Cantero <alvorithm@teje.ro>
* 🐛 Fix list-column CSV ingest and serialize graph session access
COPY failed on any file with container shapes: list-typed DDL columns (shapes UUID[], points STRING[], strokes JSON[], ...) were JSON-encoded in staging CSVs, which Ladybug's list parser rejects. Write Kuzu list literals instead, typed per column. Also: value->clj no longer crashes on LIST/STRUCT values (binding lacks value_get_value support; fall back to string), and the debug session Connection is now guarded by a per-session lock — it was shared unsynchronized between the msgbus sync loop and HTTP query/export handlers, and one lost DETACH DELETE was observed under concurrent refetch load.
Signed-off-by: Álvaro Tejero Cantero <alvorithm@teje.ro>
* ✨ Split graph console in two columns; add file tree and fullscreen
Graph view moves to its own sticky right column (overrides .widget max-width). New /dbg/actions/graph-files endpoint lists teams -> projects -> files for the profile; the console renders it as a collapsible tree where clicking a file loads it. Maximize button fullscreens the graph panel and resizes G6 on fullscreenchange.
Signed-off-by: Álvaro Tejero Cantero <alvorithm@teje.ro>
* ♻️ Replace fullscreen with in-page expand for graph view
Fullscreen API took over the whole output and broke window-manager splits (and is denied in some environments). The Expand button now toggles a fixed-position overlay covering the page while keeping browser chrome; Esc restores. Column positioning moved from inline style to the stylesheet so the expanded class can override it.
Signed-off-by: Álvaro Tejero Cantero <alvorithm@teje.ro>
* ✨ Fold containers as collapsible combos in graph view
Non-empty containers (Page, Frame, Group, Boolean, SVGRaw) render as nested G6 rect combos holding their own node plus direct children; Document stays a plain node. Double-click folds/expands (collapse-expand behavior); collapsed combos show a member count and re-route child edges. Fold state is read back from getComboData and re-marked on every refetch, so it survives live redraws. Layout gains sortByCombo to keep same-rank nodes grouped by box.
Signed-off-by: Álvaro Tejero Cantero <alvorithm@teje.ro>
* ⚡ Fix graph view freeze on large files; add fold toggle and root rule
Root cause of the tab freeze on ~1700-node files was G6's default entrance animation: measured 1700 nodes at >2 min animated vs 1.5 s with animation: false. Secondary cost was antv-dagre (~7 s at that size); since IsChildOf is a tree, an O(n) tidy layout (depth = rank, post-order leaf slots, parents centered) computed client-side replaces it and renders the same file in ~1.4 s. A guard skips auto-render above 4000 nodes with an explicit Render-anyway button, so opening the console with a huge session loaded stays responsive.
Folding is now switchable ('fold containers' checkbox, persisted in localStorage) and generalized: any node with children folds except the IsChildOf root of the loaded graph, so Documents (and later Projects/Teams) fold automatically once they gain a parent node.
Signed-off-by: Álvaro Tejero Cantero <alvorithm@teje.ro>
* ✨ Add layout dropdown to graph view
Adds a layout <select> next to the fold toggle, populated from the LAYOUTS map in the template: 'tree' (the O(n) preset layout, default) plus 13 G6 layouts (antv-dagre, dagre, circular, concentric, radial, grid, force, d3-force, force-atlas2, fruchterman, mds, combo-combined, random), all smoke-tested against combo data on this UMD build. Layout and fold toggle are independent; switching layouts recreates the graph instance (cheap with animation off); both choices persist in localStorage. antv-dagre stays available for when non-tree edges arrive.
Signed-off-by: Álvaro Tejero Cantero <alvorithm@teje.ro>
* ✨ Add query-result subgraph, ws auto-reconnect, adaptive animation
The Cypher result pane now offers 'Show result in graph view': any UUID found in any result cell selects the matching nodes in the cached export and the view renders the induced subgraph (edges kept when both endpoints match); 'Show full graph' resets. No graph reconstruction from the query result is needed.
The notifications websocket reconnects automatically (3 s retry) and resubscribes + refetches on reopen, so backend restarts no longer permanently kill the live feed; a lost session now reports 'no graph session (backend restarted?) - reload a file' instead of a bare 404.
Animation is size-adaptive: graphs (or filtered subgraphs) up to 100 nodes render animated for didactics, larger ones stay animation-free; crossing the threshold recreates the instance like a layout switch.
Signed-off-by: Álvaro Tejero Cantero <alvorithm@teje.ro>
* ✨ Add graph toolbar, animate toggle, filter columns, repaint skip
Graph view gains an on-canvas G6 toolbar (auto-fit, expand, restore - the fullscreen icons drive the existing in-page expand), an 'animate' checkbox that disables animation unconditionally when off (persisted, adaptive <=100-node rule applies only when on), and a ResizeObserver on the canvas so the panel follows window/flex resizes without touching the user's viewport. Preset tree positions are now only injected for the built-in tree layout, removing the tree-then-layout flash on animated re-renders under G6 layouts. Refetches skip the repaint when the display projection (nodes, edges, truncated) is byte-identical, so attribute-only change bursts no longer repaint.
Console: default query returns s/t name+label over all edges plus filter_src_id/filter_tgt_id columns; filter_* columns are hidden from the results table (client and server render) but still feed the 'Show result in graph view' id harvest, keeping the table legible while the graph filter stays available. The query text persists in localStorage across page reloads (restored only over the default, never over a server-rendered query). Legend shows colored Unicode glyphs matching node shapes instead of squares with textual annotations. Load/Unload buttons share one row (HTML5 form attribute), and the loaded file name links to the Penpot workspace via the legacy /#/workspace/<project-id>/<file-id> route resolved client-side from the files-tree payload.
Signed-off-by: Álvaro Tejero Cantero <alvorithm@teje.ro>
* 🐛 Fix runaway graph panel growth and blank canvas; drop Expand button
Root cause of 'graph flashes on load then disappears' plus unbounded horizontal growth of the graph panel: fieldsets default to min-inline-size: min-content, so #graph-view-panel sized to its content, and the new ResizeObserver->setSize path closed a feedback loop (setSize -> slightly wider G6 canvas -> wider fieldset -> wider .dashboard flex column -> observer fires) that grew the page ~10px per frame and wiped the painted canvas on every step. Fix severs the feedback path: #graph-view-panel gets min-inline-size: 0, #graph-canvas gets overflow: hidden, and the page section gets flex: 1 1 0 with min-width: 0 so column widths are viewport-driven, never content-driven. This also fixes the original narrow-window scrollbars defect for real. The observer stays (guarded by a current-size comparison) because G6's autoResize is inert on this UMD build (verified: window resizes left the canvas size untouched); the inert autoResize flag is dropped. Legend items now join with spaces so the nowrap spans can wrap between entries.
Also removes the header Expand button - the toolbar's expand/exit icons cover it, Esc still restores.
Verified against the running devenv with a logged-in profile and variants_simple loaded: graph renders and persists, widths stable over multiple seconds at 1400px and 1000px viewports with no horizontal overflow, canvas follows both window shrink and grow, toolbar expand gives a full-page canvas and Esc restores.
Signed-off-by: Álvaro Tejero Cantero <alvorithm@teje.ro>
* ✨ Make the default query self-explanatory; link the Cypher docs
The default query is now multi-line with // comments that explain the filter_* column convention in place (Kuzu accepts comments and blank lines mid-statement; verified against an in-memory database through the console query path). The query fieldset is retitled 'LadybugDB Cypher' with the Cypher word linking to https://docs.ladybugdb.com/cypher/.
Signed-off-by: Álvaro Tejero Cantero <alvorithm@teje.ro>
* 📎 Fix linter issues
* ✨ Add Component nodes and IsInstanceOf edges
* 🐛 Fix memory leak
* ✨ Style Component nodes and IsInstanceOf edges in graph console
Slice-3 export sends edges with a rel field. Derive tree ranking, combo derivation and fold-ability from IsChildOf only; draw other rels as overlay edges with per-rel styles (EDGE_STYLES: IsInstanceOf violet dashed, matching the new Component diamond in NODE_STYLES). Legend now lists only displayed node tables and rels, re-rendered per redraw; help text trimmed to essentials.
Signed-off-by: Álvaro Tejero Cantero <alvorithm@teje.ro>
* ✨ Add graph diff marks with step fade to graph console
Each display-changing refetch is a step: added nodes/edges get a green halo, removed ones stay as ghosts with a dashed crimson halo (nodes, fading opacity) or thicker crimson stroke (edges), re-entering layout and combos through their ghost IsChildOf edges. Marks fade linearly and drop after N steps; N is the new "fade" number input (localStorage, 0 = off). Dash + fade carry the added/removed distinction under red-green CVD (#40c057/#c2255c, deutan dE 17.4); diff is vs the previous display step, not arbitrary revisions.
Signed-off-by: Álvaro Tejero Cantero <alvorithm@teje.ro>
* 💄 Reserve chroma for changes in graph console diff mode
Monochrome entity scheme: all node tables share one slate hue, lightness separates within-glyph siblings (validated, worst pair dE 17.5), SVGRaw becomes the hollow hexagon, both rels go grey with dash as the only separator. Diff marks now own all color: thick green/crimson stroke ring (dashed for removals) plus a larger, subtler halo; the legend gains +/- entries while marks are live. Two additions to guide the eye: a brief DOM-overlay pulse on age-0 elements (independent of the G6 animation gate) and a "fold unchanged" toggle that collapses every combo not on an ancestor path of a changed element.
Signed-off-by: Álvaro Tejero Cantero <alvorithm@teje.ro>
* 🐛 Expand folded combos that gain changed elements
setData merges datum props by id on a live G6 instance, so omitting style.collapsed retained a previous true: with "fold unchanged" on, a change inside a folded combo pulsed but never expanded it. Write the boolean explicitly both ways.
Signed-off-by: Álvaro Tejero Cantero <alvorithm@teje.ro>
* ✨ Label edge rels with compact unicode symbols
Dash variants alone cannot carry the growing rel roster: EDGE_STYLES entries gain a sym rendered as a small mid-edge label with a white backing (IsInstanceOf = "∈"; IsChildOf stays unlabeled as the background structure), and the legend shows the symbol. Convention from the abacus viewer EDGE_SYM dict.
Signed-off-by: Álvaro Tejero Cantero <alvorithm@teje.ro>
* ✨ Add node inspector panel to graph console
Clicking a node fetches its full attribute row (MATCH (n:`Table` {id: uuid(...)}) RETURN n.*) through the query endpoint and renders non-null attrs into a panel under the canvas (count of empty attrs noted). Panel over tooltip: projected tables carry ~80 columns, and the panel persists for reading without obstructing the graph. Table/id are validated before Cypher interpolation; the listener is re-attached on every instance recreation.
Signed-off-by: Álvaro Tejero Cantero <alvorithm@teje.ro>
* 💄 Graph console QoL round
"Show result in graph view" moves into an actions bar above the results table; results scroll inside a 45vh container (client and server render paths); the Loaded-session fieldset gains a live "Graph size" line that stays fresh through skipped repaints; IsInstanceOf mid-edge label becomes the spelled-out rel name (∈ read as membership, not derivation) with the legend falling back to the dash-arrow for long syms.
Signed-off-by: Álvaro Tejero Cantero <alvorithm@teje.ro>
* ✨ Sync Component library changes into the Ladybug graph
* 💄 Polish graph console session panel and edge labels
Loaded-session fieldset: graph size gains a resident-memory estimate (fit to graph_sizes.md: ~1.1 MiB floor + ~5.4 KiB/node) with per-table counts on hover, replacing the load-time Projection stats; loaded-at compacts to local HH:MM with the full instant on hover. Edge rel labels drop to 7 px and lose the dashed stroke — the text label alone carries rel identity.
Signed-off-by: Álvaro Tejero Cantero <alvorithm@teje.ro>
* ✨ Add PNG export and hover tooltips to graph console
Toolbar gains an export item: graph.toDataURL({mode: "overall"}) downloads the whole laid-out graph as graph-<revn>.png — page-chrome-free captures, also the fast path for agents debugging the console. A hover tooltip (table, label, id) backs the reduced/absent labels on dense layouts.
Signed-off-by: Álvaro Tejero Cantero <alvorithm@teje.ro>
* 💄 Prune graph console layout roster and tune overlap
Remove grid, random, force, fruchterman, force-atlas2 (nothing over the kept set) and mds (stress layout degenerates to spokes on tree distances, no collision term to tune). Parameterize the keepers against node overlap — concentric/radial get preventOverlap+nodeSize, d3-force a collide radius — and shrink node labels to 7 px on those layouts (DENSE_LABEL_LAYOUTS), verified against variants_simple (72 nodes).
Signed-off-by: Álvaro Tejero Cantero <alvorithm@teje.ro>
* ✨ Highlight clicked node neighborhood in graph console
click-select behavior with degree 1: the clicked element keeps a black ring, direct neighbors stay full-strength, everything else dims to 0.2 opacity (inactive state); clicking empty canvas clears. Works on edges too (selects both endpoints) and composes with the node inspector on the same click.
Signed-off-by: Álvaro Tejero Cantero <alvorithm@teje.ro>
* ✨ Add overview mode: fold containers at or beyond a depth
"fold >= depth" number input (root = 0, empty = off, localStorage): every combo whose container sits at that IsChildOf depth or deeper collapses, giving a top-of-file overview (e.g. 2 folds the containers hanging from a Page). Composes with fold-unchanged — depth folds first, changed ancestor paths are then drilled open. Derived fold state overrides manual folds while active.
Signed-off-by: Álvaro Tejero Cantero <alvorithm@teje.ro>
* 🐛 Restore fold-containers as the combo master gate
Since fold-unchanged and depth folding arrived, withCombos ORed them in, so unchecking "fold containers" could no longer remove the combo boxes. The checkbox is the gate again; the derived fold rules are dormant without it.
Signed-off-by: Álvaro Tejero Cantero <alvorithm@teje.ro>
* 💄 Console UI polish round
Merge the load form and files tree into one "Load graph from Penpot" box (tree first, uuid + Load/Unload in a row); Loaded session carries HH:MM in its legend; the Live changes box stays hidden until the first change arrives; query fieldset reads "Query graph (LadybugDB Cypher)" with the link covering both terms. Drop the hover tooltips (distracting, useless zoomed out) and the resident-size estimate (per-table counts stay on hover); every toggle gets a "When set/checked ..." title. Depth fold: 0 now expands every container (no more hunting for max depth). Node inspector: two-column flow, structured or long values folded behind the file-tree disclosure triangle.
Signed-off-by: Álvaro Tejero Cantero <alvorithm@teje.ro>
* ✨ Report actual graph memory from the buffer manager
graph-data gains bm-bytes (CALL bm_info() -> [mem_limit mem_usage], nil-safe, under the session lock); the session panel shows it as MiB behind the node/edge counts — real resident memory replacing the removed estimate.
Signed-off-by: Álvaro Tejero Cantero <alvorithm@teje.ro>
* 💄 Console control-bar and session-panel rework
Left column narrowed 440->330 px (uuid input flexes). Control bar reordered: layout first, then animate and fade (narrow inputs), then the fold set; "fold containers" renamed "foldable containers" (on = foldable, not folded). Load becomes Reload once a session exists (same operation as the removed Full-reload button — load-session! on the current id; tooltip explains the fallback role) with Unload beside it. Session panel: revisions on one line ("ingested at N · graph now M", hover explains the difference), duplicate uuid after the file name dropped. Tried and rejected: fishbone (no positions on graph data) and compact-box (G6 tree layouts walk parent->child, IsChildOf points child->parent).
Signed-off-by: Álvaro Tejero Cantero <alvorithm@teje.ro>
* ✨ Legend entries toggle node-table visibility
Clicking a legend entry hides/shows that table across the view (struck-through while hidden, kept listed for re-enabling; pure client-side id filter through filteredGraphData, edges drop with their endpoints, ghosts respect it). Also: setting fold >= depth above 0 now switches foldable containers on — a positive depth was silently inert without combos.
Signed-off-by: Álvaro Tejero Cantero <alvorithm@teje.ro>
* ✨ Enable the edge-bundling plugin
Signed-off-by: Álvaro Tejero Cantero <alvorithm@teje.ro>
* 💄 Session breadcrumb, changelog colors, spacing
File line becomes team › project › file (clickable) with the resident-memory figure beside it (moved up from the graph-size line; breadcrumb resolves from the files-tree payload, so files outside the profiles teams show plain). add-obj/del-obj in Live changes wear the canvas diff colors. Paragraph margins tightened above Feed; left column 330→350 px.
Signed-off-by: Álvaro Tejero Cantero <alvorithm@teje.ro>
* 🐛 Guard renders against heavy graphs; add ?safe escape hatch
A heavy file could freeze the tab on load-and-render despite the animation gate: the render guard counted nodes only, and the edge-bundling plugin is iteration-heavy in edges. Guard now also trips on edges (8000), edge bundling only activates at <= 300 edges, and /dbg/graph?safe disables auto-render entirely (counts + "Render anyway"), so a page that hung can always be re-entered with the session intact.
Signed-off-by: Álvaro Tejero Cantero <alvorithm@teje.ro>
* 🔥 Remove the edge-bundling plugin
Bundled edges render unsmooth and ugly on this build; the gating constant goes with it.
Signed-off-by: Álvaro Tejero Cantero <alvorithm@teje.ro>
* 💄 One row per operation in the Live changes table
Columns revn | op | id: the revn repeats across a batch, the op wears the canvas diff colors (shape/attrs detail on hover), and the id column shows the uuid last group with the full uuid on hover, or N/A for ops without a subject id (e.g. mov-objects).
Signed-off-by: Álvaro Tejero Cantero <alvorithm@teje.ro>
* 🐛 Use app.system/system in the graph ingest helper
develop renamed app.main/system to app.system/system and dropped the
app.main require while this branch was away. Rebasing replays the old
call, so clj-kondo reports an unresolved namespace and the ns will not
load.
* ✨ Put the graph subsystem behind a flag, off by default (#11075)
`app.graph.ladybug` imports `com.ladybugdb.*` at namespace load. Two
namespaces reach the subsystem and both required it at the top level:
`app.http.debug`, which registers the `/dbg` routes, and
`app.srepl.main`, which loads with the REPL server. Every backend built
from this branch therefore linked the Ladybug native library into the
JVM at boot, whether or not a graph was ever used.
Add a `:graph` flag to `varia`, deliberately absent from `default` so
that a released Penpot ships with the subsystem off. Both require sites
now resolve `app.graph.*` at call time, so with the flag off no
`com.ladybugdb` class is loaded. The nine `/dbg` graph routes are
registered only when the flag is on, and 404 otherwise. The `/dbg` admin
gate is untouched: the flag decides which routes exist, not who may
reach them. When the flag is on, route init requires the subsystem
eagerly, so a missing or unusable native library fails the boot rather
than the first console request.
No tracked file turns the flag on. `backend/scripts/_env` leaves it out,
so a devenv boots with the subsystem off exactly as a released build
does, and `docker/images/docker-compose.yaml`, the self-hosting
distribution, is untouched. Whoever works on the graph turns it on for
one checkout through the gitignored `backend/scripts/_env.local`, which
every backend and exporter dev script sources right after `_env`.
Verified with `-verbose:class` over a boot's namespace load plus
`ig/init-key ::routes`: 9 `com.ladybugdb` classes before this change
with no flag set, 0 after it with the flag off, 9 with `enable-graph`.
* ⬆️ Take Ladybug 0.19.1
`com.ladybugdb/lbug` moves from 0.18.0 to 0.19.1, the current release on
Maven Central. The engine fixes a SIGSEGV on an unwrapped parameter and
moves parameter coercion out of JNI, so shipping 0.18.0 would land a
native library into `develop` with a known crash already fixed upstream.
Nothing else changes. This branch has no `app.graph.arrow`, so the
top-level Arrow field-name backticking that 0.19.x retires does not
exist here and there is no workaround to remove alongside the bump.
AI-assisted-by: mixed models
* 📎 Pin the graph console's G6 bundle to an exact version
The console loaded `@antv/g6@5` from jsDelivr, a floating major range,
so the JavaScript served into the page could change without a Penpot
release. Pin it to 5.1.1, the version the range resolves to today.
Where the dependency finally belongs is an open question for review:
vendored into `backend/resources`, declared in `frontend/package.json`
if the console moves out of `/dbg`, or left on the CDN. Pinning removes
the floating-code problem without pre-empting that decision.
AI-assisted-by: mixed models
* ✨ Add graph provenance, column naming and two transforms
A projected graph is a cache of one file at one revision, built by one
schema, and nothing in it said so. `GraphMeta` records the file, the
revision, the schema version and the producer, and is written last, so
its presence also marks the build complete and its contents say whether
a cached database is still worth opening.
- `graph/meta.clj`: the `GraphMeta` table and its writer.
- `graph/schema/contract.clj`: one place that maps a Penpot key to its
graph column. The rule is snake_case of the key; every exception, be
it a rename, a drop or a type override, is recorded there with its
reason, so a divergence is a diff to review rather than a silent
rename.
- `graph/project/document.clj`: `page-id` and the inherited
`component-id` are written during the tree walk, which already knows
both, rather than by a post-ingest statement. `graph/sync.clj` does
the same on the incremental path, so a live-synced graph matches a
rebuild.
- `graph/project/transforms.clj`: a registry, so adding a derived-link
pass is one entry. Adds `RefersTo` (from `shape-ref`) and
`FillsSwapSlot` (from `swap-slot-*` entries in `touched`, then
stripped as `ctk/normal-touched-groups` does).
- `graph/debug.clj`, `graph/stats.clj`: enumerate relationship tables
from the catalog instead of naming them, so the console's graph view
and the ingest counts pick up new edge types without being told.
- `graph/debug.clj`, `http/debug.clj`: `graph-export` gains
`source=session`, which snapshots the live in-memory console graph
through EXPORT/IMPORT DATABASE. Live sync moves that graph away from a
fresh projection, and taking it away to query elsewhere is the point
of asking for it.
AI-assisted-by: mixed models
* 🐛 Write graph values Ladybug's CSV reader cannot carry through Cypher
Three parity failures against beadpot's suite, all one cause: the bulk loader
put compound and multi-line values into CSV, where Ladybug parses a field's
*contents* as a literal with no escape mechanism at all. Verified against
0.18: a comma inside a list element ends the element, quotes are kept as part
of the value rather than delimiting it, and the parallel reader rejects
quoted newlines outright.
So a value now goes through CSV only if it cannot be misread there — UUIDs,
numbers, booleans, single-line strings, and lists of those. Everything else
(MAP, STRUCT, STRING[]/JSON[], any string containing a newline) is written
after the COPY by one Cypher statement per row, where `app.graph.ladybug`
escapes properly. Parquet removes the distinction entirely and is still the
right destination (masterplan P0 T1); this is what CSV can honestly do.
Consequences beyond the encoding:
- `touched` entries reached the graph as `:swap-slot-…`, keywords stringified
with their colon, so `LinkSwapSlots` matched nothing. Keywords now render
through `name`.
- Shape names lost their newlines to a flattening step that existed only to
keep the CSV writer happy. They are preserved.
- `applied_tokens` keys are rendered camelCase, the form Penpot's own JSON
encoder produces and the one beadpot's `AppliedTokenKey` holds — a MAP
column's keys are values, not schema, so they are not snake_cased.
- `link-component-instances!` keys on `component-file`, not `component-id`
alone. The projection denormalizes `component-id` down the shape tree, after
which it no longer tells an instance head from a shape inside one, and the
transform linked every descendant frame; `ctk/instance-of?` requires both
keys anyway. IsInstanceOf on the variants fixture: 78 -> 60, matching
beadpot exactly.
`app.graph.schema.nodes/format-column-value` is now the single place that
knows a column's type and its contract details, used by the bulk loader and
the incremental sync alike so the two cannot disagree about a value's shape.
* ✨ Type graph columns as tightly as Ladybug allows
Ladybug is schema-first and strongly typed: a property key gets its type
at table-creation time and there is no widening later. That makes the
Malli to Ladybug mapping the whole of the graph's typing, and it was
leaving a lot on the table: a transform stored as `STRING`, a rect as
`JSON`, a set of feature flags as a single `STRING`. A column typed
`DOUBLE[4]` is four numbers a consumer reads as a tensor row; the same
value as JSON is text somebody has to parse and trust.
`app.graph.schema.types` now maps, in order: scalars; Penpot value types
whose layout is fixed even though Malli only sees a map or a string
(`::gmt/matrix` to `DOUBLE[6]`, `::gpt/point` to `DOUBLE[2]`,
`::grc/rect` to `DOUBLE[4]`, `::clr/hex-color` to `UINT32`); then
structure, with collections to `T[]`, `:map-of` to `MAP(k, v)`, and a
closed map of scalars to a `STRUCT`. JSON is the fallback of last
resort, for schemas that genuinely admit more than one shape.
Two defects fell out. `::sm/set` was unmapped, so `features` and
`migrations` were single strings rather than `STRING[]`, and
`::sm/one-of`, how Penpot spells a closed set of keywords, was unmapped
too, so `blend-mode`, `grow-type`, the constraints and every `layout-*`
were mistyped.
A tight column is only worth having if the writer fills it in that
shape, so `app.graph.schema.values` shapes a value for its type: a
matrix record into six doubles, a hex colour into a packed integer, a
map into a struct's fields. Both writers go through it, so the bulk load
and the incremental sync cannot disagree. What that required:
- STRUCT field names must be backticked in the DDL *and* in every
literal, because a grid cell has a field named `column`. The catalog
reports them bare.
- A struct literal's type is its field list, so every declared field
must appear, and an absent one needs `cast(NULL, '<type>')`. A bare
NULL is typed STRING and changes the struct's type.
- `STRUCT(…)[]` starts with `STRUCT(` but is a list, so the list check
comes first.
- Nested lists cannot be rendered with `str`: Clojure's `[1 2]` is
space-separated and Ladybug reads it as a one-element array.
Three more corrections in the same area:
- `project-attrs` used truthiness where it meant `some?`, so `opacity 0`
and `blocked false` projected as absent.
- Set-valued columns are written sorted. A set has no order, so the
column varied between builds of the same file, which is precisely what
stops two builds being diffable.
- An empty collection is written as `[]` rather than skipped. A shape
with no fills has none; NULL would say "unknown".
Renamed the `kuzu-*` helpers to `ladybug-*`: Kùzu is deprecated and
Ladybug substitutes it, so a name bearing the engine should bear this
one. The one remaining mention cites the upstream issue Ladybug
inherits.
AI-assisted-by: mixed models
* ✨ Add the file-level graph columns and tighten the svg ones
Split out of "🐛 Declare the shape attributes stored files carry",
which is now #11125 and carries only its `common/` half. This commit is
the graph's own side of that change, and it stays on this branch.
`app.graph.schema.contract` pins `svg_viewbox` to `DOUBLE[4]` and
`svg_transform` to `DOUBLE[6]`. The shape schema types both `:map` on
purpose, because legacy files hold them as plain maps rather than as
`::grc/rect` and `::gmt/matrix` records, and a tighter *schema* would
reject those files. A tighter *column* costs nothing, since
`app.graph.schema.values/coerce` reads either form.
`app.graph.schema.nodes` declares four file-level attributes as
projection `:extra` rather than in `ctf/schema:file`: `:options`,
`:backend`, `:comment-thread-seqn`, and `:ignore-sync-until`. Declaring
them in the file schema breaks saving, measured at 185 failures, because
`app.binfile.common/update-file!` derives its UPDATE column list from a
file map's keys and the `file` table has no `backend` column, that value
being synthesized on read. An `:extra` is local to the graph and cannot
reach a write.
`app.graph.project.document` lifts `:options` out of `:data` before the
blob is dropped, so a consumer reads file-level configuration without
opening the blob.
AI-assisted-by: mixed models
* ✨ Add the Arrow prerequisites for in-memory bulk load
lbug pulls arrow-memory-core and arrow-vector but no allocation-manager implementation, so RootAllocator cannot be constructed; arrow-memory-netty 18.2.0 matches the arrow-vector lbug already brings and pulls only netty-buffer, netty-common, jackson and slf4j-api, all of which the backend already has.
--add-opens=java.base/java.nio=ALL-UNNAMED is the second half: without it MemoryUtil's static initializer dies with an InaccessibleObjectException that surfaces as an unhelpful NoClassDefFoundError from anything touching RootAllocator. It has to be present at JVM start, hence all three places. Note app.main/restart will not pick it up — it restarts integrant inside the same JVM, so the process must be restarted.
Worth a reviewer's attention: this is a JVM-wide flag added for one subsystem. It is the standard Arrow requirement and grants nothing beyond reflective access to java.nio, but it strengthens the case for putting the whole graph subsystem behind a feature flag.
* ✨ Bulk load through in-memory Arrow; delete the CSV loader
app.graph.arrow stages rows as Arrow VectorSchemaRoots and COPYs from them. No file is written at any point and no value is rendered as text for the engine to re-parse, so the defect class that produced three of this branch's four backend defects cannot recur.
app.graph.bulk is deleted whole. csv-representable?, defer-to-cypher?, multiline?, fixup-statements, ladybug-literal, ladybug-list-element, ladybug-list-cell and staging-dir go with it, along with the post-COPY Cypher pass that emitted one SET per row.
Measured before deciding: the fixup pass was ~77% execution, 16-22% parse and 6-7% round-trip, and prepared statements could not have recovered any of it — every fixup row carries a MAP column and Ladybug binds scalars only. So this replaces rather than optimizes. Marginal ingest 4.0 -> 1.21 ms/shape; ~25 s extrapolated at 20k shapes against the ~2 min the CSV path projected. Size unchanged.
Four engine facts the implementation rests on, each verified against 0.18.2 with a standalone probe:
- An Arrow table is not a COPY source identifier but is a MATCH-able node label.
- A MAP vector's entries child must be a non-nullable struct, and MapVector.getWriter promotes it to a sparse union, so map vectors are built from an explicit Field and filled child-first.
- Ladybug names a staged table's columns and struct fields from the Arrow field names and quotes none of them, so anything needing quotes must arrive quoted — hence cypher-property-key, not column-name, names the Arrow fields.
- createArrowRelTable cannot resolve endpoints against a UUID-keyed node table under any encoding, so edges stage as a node table and the COPY subquery joins them.
values/coerce is reused unchanged, so the Arrow and Cypher writers cannot disagree about a value's shape; nodes/column-map-key-fn is extracted so they cannot disagree about a MAP's key spelling either.
Verified with pytest --graph-origin=penpot-only unchanged at 225/38/1 and --graph-origin=penpot unchanged at 258 passed / 2 pre-existing failures, both baselines re-established against a reverted backend rather than assumed; with bp graph diff between a CSV-built and an Arrow-built graph reporting "Graphs agree"; and with an adversarial round-trip carrying a quote, a backslash, a newline, a CRLF and a tab through STRING, STRING[] elements and MAP values.
The diff was necessary, not belt-and-braces: both parity suites passed an earlier revision of this change that was writing EDN into every JSON column, because beadpot's assertions never parse those columns. It also showed Arrow correcting a CSV defect — an empty Component.path was being stored as NULL, because Ladybug's CSV reader cannot distinguish an empty field from an absent one.
* ✨ Add a prepared-statement surface to the graph connection
`app.graph.ladybug` could only run Cypher as text. Every value the sync
path writes is therefore concatenated into the statement, and nothing can
ask the engine whether a statement is even valid without running it.
Add the four functions that close both gaps. `prepare-on-connection!`
parses and binds without executing. `execute-prepared!` binds a parameter
map and runs it. `exec-prepared-on-connection!` prepares every statement
in a batch before executing any of them, so a parse or bind failure
aborts before the first mutation. `validate-on-connection!` returns
`{:ok? :error :read-only?}` instead of raising, which is what a gate
wants.
`->param-value` is the only `Value` constructor on the write path. It is
unconditional: on lbug 0.18.2 an unwrapped parameter does not raise, it
SIGSEGVs the JVM inside `lbug_value_clone`. Parameters are scalars only,
because the JNI `Value` constructor takes no list or map, so `MAP`,
`STRUCT` and `T[]` columns stay literal-rendered and the `:else` branch
raises rather than crashing.
Two departures from the design, both closing a JNI-handle leak on the
error path: `prepare-on-connection!` closes the failed
`PreparedStatement` before raising, and `execute-prepared!` closes every
`Value` it built, including the ones built before a later parameter was
rejected.
`as-statement` accepts a bare string, so the sync builders can convert to
bound parameters one family at a time rather than in one commit.
AI-assisted-by: mixed models
* 🐛 Write the document revision to the column that exists
`set-document-revision-statement` emitted `SET d.revn`, but the column is
`revision`: the beadpot contract renames `:revn` and the DDL has followed
it since. The statement is the last one in every sync batch, so each
batch raised after its mutations had already committed, and the session's
in-memory index stayed frozen at its load-time revision.
Name the column through `nodes/cypher-property-key` rather than spelling
it, so the DDL and the statement cannot disagree again.
Found by the binder gate in the next commit, on its first run.
AI-assisted-by: mixed models
* ✨ Gate every sync statement template through the binder
Nothing checked that the eleven Cypher templates `app.graph.sync` emits
still bind against the DDL the schema registry generates. A renamed
column, a dropped table or a reserved word emitted unquoted surfaced only
when a live session ran the statement, and by then the batch's earlier
mutations had committed.
`backend-tests.graph-binder-gate-test` opens a `:memory:` database,
creates the live schema on it, and *prepares* one instance of each
template without executing any of them. 14 tests, 51 assertions: the
eleven templates, label coverage over all twelve registered node tables,
and two assertions on the gate itself, that a `RETURN` reads as read-only
and a `SET` does not, and that an unbindable statement is reported rather
than thrown.
It was not green on HEAD: it caught `set-document-revision-statement`
writing a column that no longer exists, fixed in the previous commit. Red
on both injected templates tried.
No `:jvm-opts` change: CI's `-M:dev:test` already carries the native
access flags the engine needs.
AI-assisted-by: mixed models
* 🐛 Let the engine quote the Arrow field names it interpolates
`node-batch` named every top-level Arrow field with backticks, so that a
column whose name is a reserved word (`Page.index`, `Document.options`)
survived the DDL Ladybug generates for a staged table. The engine now
quotes those identifiers itself, and it does not collapse a doubled
backtick, so a pre-quoted name reaches the parser as ``index`` and
`createArrowTable` fails outright:
Parser exception: mismatched input '``' expecting PRIMARY
Name the fields with `column-name`. The `COPY` projection is Cypher
rather than DDL and keeps its own backticks through
`cypher-property-key`, and STRUCT member names keep theirs too: those
come out of `LogicalType::toString()`, which the DDL builder does not
touch, so an unquoted member called `column` still fails to parse.
Measured with `probes/arrow/probe25.clj` against lbug 0.19.1: a plain
top-level reserved word loads and reads back, a pre-quoted one fails to
parse, a plain STRUCT member fails to parse, and a pre-quoted one loads
and reads back.
Also re-dates the engine facts in the `app.graph.arrow` docstring to the
version they were checked against, drops the SIGSEGV note from
`->param-value` now that `Connection.execute` rejects an unwrapped
parameter, and removes two references to the CSV loader.
AI-assisted-by: mixed models
* 📚 State what the graph schema does, not what it mirrors
The graph namespaces explained themselves by citing a separate project
whose Python pipeline reads the graphs this backend writes. A reader of
this repository does not have that project and should not need it, and a
docstring that justifies a choice by pointing elsewhere cannot be checked
here.
Every claim survives; only the framing changes. Column names and types
are Penpot's own decision, recorded with the reason for each divergence
from the snake_case default. The transform registry describes the edges
it materializes. The denormalizations in `app.graph.project.document`
are justified by the walk already holding both answers.
Three corrections fall out of the rewrite:
- `app.graph.schema.contract` claimed a test, `graph_contract_test`,
that walks a checked-in schema manifest and fails on any divergence.
No such test exists. The paragraph is gone.
- `app.graph.project.document` pointed at
`app.graph.meta/projection-transforms`, which does not exist.
- `app.graph.project.transforms/registry` claimed its entries were "in
application order" while `apply-transforms!` reduced over the literal
vector. The three registered transforms read disjoint columns, so the
order is not load-bearing. The docstring now says so, and the one real
ordering constraint is stated where it applies: `link-swap-slots!`
strips `swap-slot-*` entries from `touched`, so anything reading
`touched` has to run before it.
`contract/pending-beadpot-columns` becomes `contract/unprojected-keys`.
It is referenced nowhere else.
AI-assisted-by: mixed models
* ✨ Refuse a mutating query from the graph console
`debug/query-session!` ran whatever it was handed against the session
connection. A session graph is a projection of a file, rebuilt from that
file by Reload, so a mutation from the console produces a graph no
rebuild reproduces and no query result explains.
Bind the statement against the live schema first. A statement that does
not bind reports the binder's own message and executes nothing, which
also turns a misspelt table or property into an immediate error instead
of an empty result. A statement that binds runs only when the engine's
own read/write analysis calls it read-only.
The console's query box is labelled read-only. Load, Reload, Unload and
live sync are unaffected: they are separate handlers and do not go
through this path.
AI-assisted-by: mixed models
* 🐛 Keep a synced graph equal to a rebuilt one
Cold projection and incremental sync are two implementations of one
mapping and nothing checked that they agree. They did not.
`backend-tests.graph-sync-parity-test` projects a file into one
`:memory:` database, applies a change list to that database and the same
list to the file data, projects the result into a second database, and
diffs the two down to the row and the column. It found four
disagreements, each fixed here.
**Sibling order was inverted.** A container's stored `:shapes` list runs
bottom to top and `IsChildOf.position` numbers children in Penpot
z-order, so appending to the list means taking position 0 and pushing
every sibling up. Sync instead handed each new child the next free
number, so any container edited live carried its children in the
opposite order to a rebuild, and a delete left a gap where a rebuild
renumbers densely. `insert-position` and `renumber-siblings` put the two
paths on the same rule for `:add-obj`, `:mov-objects` and `:del-obj`,
including a block move and `:after-shape`.
**A moved shape kept its old parent.** `:mov-objects` moved the edge and
left the shape's own `parent_id` and `frame_id` columns pointing at the
container it came from. Both now follow, and `frame_id` follows through
the whole subtree the shape carries, as
`app.common.files.changes` does for `:mov-objects`. A top-level shape's
column holds `uuid/zero`, the page's root frame, while its edge points
at the Page.
**A container's `shapes` column went stale.** Nothing maintained it
after an add, a move or a delete. It is now rebuilt from the sibling
order on every change that touches a container.
**Pages came out backwards.** `projection-data` reversed `:pages` before
numbering them, which is right for child shapes and wrong for pages:
`:pages` is the tab order and has no second ordering to undo. `Page.index`
and the page's `IsChildOf.position` are now that order.
One defect the test does not reach, fixed on the way past:
`index-add-shape!` accepted `:component-ctx` and dropped it, so a shape
added under an instance head added in the same session inherited no
`component-id`.
AI-assisted-by: mixed models
* 🐛 Build a synced page node the way the projection does
`apply-add-page` sent the new Page node through `nodes/validate-node`,
which checks a map against the registry schema and returns it unchanged.
Every other node on both write paths goes through
`nodes/project-attrs`, which also selects the projected keys and is the
single place a column-level rule can live. A rule added there reached a
rebuilt page and not a synced one.
AI-assisted-by: mixed models
* 🐛 Let the graph view's query filter follow the graph
"Show result in graph view" froze the set of node ids the query returned
and filtered every later repaint against it. Live sync creates ids the
set has never seen, so a shape created while a filter was on could not
appear in the view at any point, and clicking "Show full graph" was the
only way to see it. A node the query would no longer match stayed.
Keep the query beside the ids and re-run it whenever the graph repaints,
which is only when the projection actually changed. A failed re-run
keeps the ids in hand and says so on the status line rather than passing
a stale view off as current.
`idsInResult` and `presentIds` are extracted from the two places that
scraped UUIDs out of a result.
Verified in the devenv: with a filter showing 108 of 276 nodes, a
`:file-change` adding a Frame published on the session's msgbus topic
took the view to 109 of 277, with the new node carrying its added mark,
and no interaction.
AI-assisted-by: mixed models
* ♻️ Rename app.graph.project to app.graph.projection
`project` is a Penpot noun: a team holds projects and a project holds
files, and the graph will carry a `Project` node table. A namespace
called `app.graph.project.document` therefore reads as "the graph of a
Penpot project" and means the opposite.
`projection` is the word the rest of the subsystem already uses for the
operation: `projection-data`, `load-projection!`, `:projection` in the
ingest report, and `app.graph.schema.projection`.
Pure rename. Both namespaces and every alias move; nothing else changes.
AI-assisted-by: mixed models
* 📎 Apply the project formatter to the graph namespaces
`cljfmt check src/ test/` is a step of the Backend workflow and these two
files did not pass it: an import block sorted the way a human reads it
rather than the way the formatter sorts it, and a `cond` in
`format-typed-value` indented one column short.
Formatter output only. No semantic change.
AI-assisted-by: mixed models
* 📚 Document graph experiment architecture
Add Serena memory coverage for the embedded Ladybug graph subsystem.\nDocument projection, incremental sync, console data flow, tests, and operational risks.\n\nAI-assisted-by: gpt-5.6-luna
---------
Signed-off-by: Álvaro Tejero Cantero <alvorithm@teje.ro>
Co-authored-by: Alejandro Alonso <alejandroalonsofernandez@gmail.com>
Co-authored-by: Andrey Antukh <niwi@niwi.nz>
Implement RPC methods for querying server error reports with pagination
and filtering. Add CLI tool (tools/error-reports.mjs) for convenient
access with table and JSON output formats. Extract profile-id from audit
events and logging context for better error categorization. Build
improved HREF using request path when available.
AI-assisted-by: qwen3.7-plus
Mainly make it receive the whol cfg/system instead only props. This
makes the api more flexible for a future extending without the need
to change the api again.
This upgrade also includes complete elimination of use spec
from the backend codebase, completing the long running migration
to fully use malli for validation and decoding.
Relevant changes:
- Add the ability to create migration in both directions, defaulting
to identity if not provided
- Move the version attribute to file table column for to make it more
accessible (previously it was on data blob)
- Reduce db update operations on file-update rpc method
This commit separates the email cleaning mechanism to a separated
function, and enables a proper cleaning of `mailto:` prefix, usually
found on invitations because users just copy and paste from external
source.
Previously the file processing was implemented 3 times using similar
approaches bug each own with its own bugs. This PR unifies the
loging to a single implementation used by the 3 operations.
