Add notebook.json containers with multi-member pages, fix PDF text
editing (size/bold/drag/double-tap), index SidecarText in search, and
share keyboard page shortcuts plus a PDF scrubber.
Co-authored-by: Cursor <cursoragent@cursor.com>
Redesign the optional FastAPI companion around vault files (manifest /
PUT/GET/DELETE + OCR jobs) instead of legacy strokes_json notes. Wire a
client Server settings panel for health/login. Polish shell UX: l10n for
settings/home/board, sticky-board empty state, and a narrow-screen
diagnostics FAB.
Co-authored-by: Cursor <cursoragent@cursor.com>
Make Surface remote debugging and classroom workflows viable: always-on
structured logs with one-click zip export, a single AppShell chrome,
OOXML PPTX/DOCX annotation without LibreOffice, and a first-class sticky
board. Also drop spike/legacy ink widgets and tighten pen feel
(predictor, PenInfoHistory, page-tile layer).
Co-authored-by: Cursor <cursoragent@cursor.com>
Two-way sync of the vault folder to a user-configured WebDAV server,
so annotations (which travel with the file) sync with the file.
- WebDavSyncService.syncNow: per-file decision — local-only uploads,
remote-only downloads, and when BOTH sides changed since the last
sync it keeps the loser as <file>.conflict-<mtime> on both sides
(last-write-wins by mtime) so no data is ever lost. Creates dirs as
needed; deletes are conservative.
- The decision logic is pure and unit-tested against a fake WebDAV
client; the real client is a thin http adapter (no dio dependency).
- Settings: WebDAV URL / user / password / remote folder, Test
connection, Sync now (with status + last-synced), and an auto-sync
toggle (default OFF).
Real server round-trips are device/server-validated. Credentials are in
SharedPreferences for now (TODO secure-storage). analyze clean, 432 tests.
Search now covers handwriting, the PDF text layer, AND scanned
(rasterized) PDFs.
- PdfTextIndexer runs at import: sums the embedded text layer across
pages; if present it stores that as the document body, otherwise the
PDF is rasterized and its rendered pages are OCR'd in the background.
The result lands in the sidecar `pageText` field (distinct from
`ocrText`, the handwriting OCR). Idempotent (skips a sidecar that
already has pageText); degrades gracefully with no OCR engine.
- pdfrx_page_text_source abstracts text/render so it's testable.
- VaultSearchIndex now harvests title + typed text + handwriting OCR +
PDF pageText, so search finds notes, typed PDFs and scanned PDFs.
analyze clean, 409 tests green.
Phase 6 (final storage phase).
- SidecarRepositoryRegistry tracks every open repo; SidecarFlushObserver
(a WidgetsBindingObserver in main) flushes them all on
inactive/hidden/paused/detached, awaiting each flush — the last
strokes can't be lost on app close, not just on the 800ms timer.
- VaultSearchIndex rebuilds by scanning vault sidecars (the source of
truth) — note titles, OCR text and document names — and search_provider
queries it, so search spans notes + PDFs. Rebuilt on launch / after
import.
The vault file-based storage migration (Phases 0-6) is complete:
annotations travel with the file, picked vault folder, atomic autosave,
one Import-file entry, SQLite migrated to sidecars. analyze clean,
tests green.
Phase 5. On first launch with a valid vault, migrate legacy SQLite
data into vault sidecars so nothing is lost on upgrade.
- SqliteToSidecarMigrator: documents (+ per-page ink, scratch-links
+ scratchpads, bookmarks) -> notebook folder + <file>.badnote.json;
notes (+ strokes) -> notebook.badnote.json. Reuses existing JSON.
- Idempotent (skips already-migrated targets); missing source files
still get their annotations migrated.
- DB relocates to <vault>/.badnote/index.sqlite; the legacy DB is
renamed to .premigration ONLY after a successful pass, so a failed
migration leaves data intact and the run-once flag unset.
- main.dart runs it once, gated on vaultMigrationDone.
Golden migrator tests (seeded legacy DB -> sidecars, idempotent
re-run, legacy preserved). analyze clean, tests green.
Phase 4. Standalone notes move off SQLite into the vault, like the
PDF annotations.
- "Create notebook" makes a vault folder with a notebook.badnote.json
(BadnoteSidecar docType 'notebook' + a title field), opened via
SidecarRepository.
- PenNoteScreen loads/saves its strokes (page 0) + title to that
sidecar instead of the SQLite Note model.
- note_provider lists notes from a vault scan (VaultService.scanNotes
= folders with notebook.badnote.json and no source file); the doc
scan still excludes them. Delete removes the folder.
PDF/slide editors unchanged; pre-existing SQLite notes migrate in
Phase 5. analyze clean, tests green.
Phase 3. Import becomes a single top-level action beside "Create
notebook" and the library is vault-backed.
- VaultService.createNotebook copies a picked file into a fresh
(de-duplicated) notebook folder under the vault; its sidecar lives
beside it, so annotations travel with the file.
- Home screen: one "Import file" action with a multi-extension picker
(pdf / docx / pptx); routes to the editor by extension.
- The document list is now a vault scan (folders with a source file),
not the SQLite documents table — no cache, always correct.
- PPTX soffice detection fix; DOCX convert-on-import is best-effort
and fails gracefully when LibreOffice is unavailable.
analyze clean, tests green.
Phase 0 of the file-based storage plan (docs/plans/
2026-06-24-file-based-storage.md). Foundation only — no editor or
DB change yet.
- VaultService (SharedPreferences): stores the vault root path,
vaultRootValid() = path set AND directory exists.
- VaultSetupScreen: first-run folder picker (file_picker, Windows).
- main.dart gates HomeScreen behind a valid vault, re-prompting if
the saved folder is gone.
- Settings: a Vault section to change the folder.
Editors still use SQLite; later phases move annotations into
per-file sidecars under the vault. analyze clean, tests green.
Replace the rejected standalone sticky-card board with the real
feature: place a link anchor anywhere on a PDF page, tap it to open
split view whose right pane is THAT anchor's own infinite scratchpad
(keyed by anchor id) — like a paper sticky-note tab.
- ScratchLink model + scratch_links table (id, doc, page, nx, ny).
- PDF editor: "place link" tool drops/loads/shows tappable markers;
tap opens SplitViewScreen for that anchor; long-press deletes.
- SplitViewScreen rebuilt on pdfrx (was syncfusion), right scratchpad
keyed by scratchLinkId, new brush palette (was AnnotationToolbar).
- Remove board_screen + its test + the home board entry.
analyze clean, tests green.
Wire the F7 双链 + 无限便利贴 model (Board/LinkGraph) into a
reachable screen. Previously the model existed but had no UI and
no entry point.
board_screen.dart: an infinite InteractiveViewer canvas of
draggable, editable sticky cards. Card text renders [[links]] as
tappable chips that pan to the target card (dangling links styled
apart). A backlinks panel lists "linked from" via backlinksOf.
"Add card" FAB drops a card at the viewport center.
Persistence: a board_cards table (DB v7), one row per card,
debounced 800ms like the ink editors, loaded on open — boards
survive restart. Entry added to the home screen app bar
(dashboard_customize icon).
Ink-on-cards, multi-board management and link autocomplete are
deferred (TODO board-ink / board-multi / board-link-autocomplete).
analyze clean, 285 tests green.
Replace the 2-tool ink system with a data-driven, Krita-style
BrushProfile (lib/editor/engine/brush.dart). Adding a brush is a
const map entry, not render-path branching.
Four presets from the rnote/krita spec:
- fountain pen: quadratic (p^2) pressure, wide dynamic width
- ballpoint: near-constant width (thinning 0.15)
- highlighter: flat width, square caps
- pencil: sqrt(p) pressure, moderate width
Pressure is pre-warped per brush via PressureCurve(gamma) before
perfect_freehand; geometry fields (thinning/streamline/smoothing/
caps) flow through the shared stroke recipe so the PDF overlay and
the note/slide PenCanvas both honor the brush. Brush kind is now
persisted on the stroke model. Picker added to all three toolbars.
Opacity/multiply and pencil grain are carried as data but not yet
composited (TODO brush-opacity / brush-texture); this increment is
width + pressure-curve differentiation. analyze clean, 283 tests.
pdf_service._buildFreehandPdfPath hardcoded its OWN StrokeOptions
(thinning:0.7, streamline:0.5) instead of the shared geometry — the R7
hairline-export divergence. The prior pen-feel commit (streamline 0.5→0.32 on
screen) widened the gap: export still rendered at 0.5.
Extract the ONE perfect_freehand recipe into stroke_geometry.freehandOutlinePoints
(owns thinning/smoothing/streamline/simulatePressure). buildStrokeOutline (screen)
and pdf_service (export, via InkStroke→pfPoints) now both call it, so the
StrokeOptions live in exactly one place and screen↔export can't drift again.
Export now matches screen: thinning 0.85 (kDefaultPenThinning), streamline 0.32.
test/export_geometry_test.dart pins it: buildStrokeOutline traces exactly the
shared outline; default thinning == kDefaultPenThinning; thinning is wired;
empty input is safe.
flutter analyze lib/editor clean; 78/78 tests pass (+4).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
W1 — Custom pen width + pressure sensitivity (Saber-style):
- Root cause of "压感没用": perfect_freehand 1.0.4 IGNORES real stylus pressure
(hardcodes radius=size/2 when simulatePressure=false) — width never tracked pen
force. Upgraded perfect_freehand ^1.0.0 -> ^2.0.0 (honors real pressure); migrated
all 5 getStroke call sites to the 2.x API (PointVector / StrokeOptions / Offset).
- De-hardcoded `thinning` into `kDefaultPenThinning` (0.85), single source shared by
the on-screen painter and the PDF export path; exposed as PenConfig.pressureSensitivity
with a Pressure Sensitivity slider; live-applies via a config listener.
W3 — Native Windows pen plugin (tilt + barrel/eraser buttons):
- windows/runner/pen_channel.{h,cpp}: observe WM_POINTER at the TOP of MessageHandler
(before HandleTopLevelWindowProc, which Flutter uses to consume pen events), read
GetPointerPenInfo penFlags + tilt, stream over EventChannel('badnote/pen'); non-consuming.
- PenInputService: single latched hardware state (no Win32-pointerId<->event.pointer
correlation); graceful no-op off-Windows.
- pen_canvas maps barrel/inverted/eraser through PenConfig.sideButton/eraserEnd
(eraser/undo/toggleTool/pan) and captures tilt into PenPoint.tilt -> EditorPoint.tilt.
W2 — Zoom flicker: page raster isolated in its own RepaintBoundary (safe interim);
definitive crisp-on-zoom fix gated on the on-device root-cause probe (plan M3).
Plans: ralplan-consensus plan at docs/plans/2026-06-22-badnote-pen-polish.md
(Architect APPROVE-WITH-MUST-FIX M1-M4 + Critic ITERATE->APPROVE).
Tests: 58/58 pass incl. shared-thinning invariant + thinning-affects-outline +
tilt-adapter round-trip. flutter analyze clean; linux debug build OK.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Make on-device OCR a pluggable local service so it runs locally on every
platform (not just Windows), aimed at GoodNotes/Notability-class handwriting on
low-power hardware (e.g. Zen2 APU, CPU/iGPU).
- New OcrBackend abstraction (lib/services/ocr/): selector prefers an embedded
ONNX recognition backend, falling back to the OS-native backend (Windows
WinRT), and to a clean no-op when neither is available.
- OnnxRecognitionBackend: flutter_onnxruntime session from a bundled asset,
dart:ui preprocessing (resize to 48px, CHW float32, normalized), pure-Dart CTC
greedy decode. Fully guarded — absent model/dict is a no-op; never throws.
- ocr_engine.dart kept as a thin facade (recognizeImage) delegating to the
selector, so ocr_service.dart is unchanged.
- CtcDecoder unit-tested (6 tests). flutter analyze clean; all tests pass.
- Model is not committed; tool/fetch_ocr_model.sh + assets/models/ocr/README.md
document fetching PP-OCRv4 rec + dict on the dev machine.
- CI: forward HTTPS_PROXY to the Windows build so CMake can fetch the ONNX
Runtime native lib behind the GFW; README documents the system-install
alternative. PP-OCR geometry/blank assumptions documented for on-device tuning.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Bug fixes (Flutter):
- Wrap multi-statement DB writes (insert/update/delete note, deleteDocument,
deletePageData, OCR FTS merge, migrations) in transactions to prevent data
loss on interruption and a read-modify-write FTS race.
- Fix PdfDocument leaks on exception (try/finally dispose) and preserve image
aspect ratio when stamping images onto PDF pages.
- Guard file-picker against empty selection (was .single -> crash).
- Fix eraser ConcurrentModificationError and unmodifiable-list crash on PDF
pages; capture page synchronously on save to stop wrong-page data loss.
- Fix Riverpod DB-not-ready races, broken pull-to-refresh, settings load race,
and search N+1; transform stored annotations on PDF page rotation.
- Normalize pen pressure for devices without a pressure range.
- PPT: single source of truth for slide strokes so ink displays and exports.
UI/UX:
- Material 3 typography, theme-aware colors (dark-mode fixes), hover cursors
and right-click/visible actions on desktop, keyboard shortcuts (undo/redo/
save/find), toolbar overflow handling, friendlier empty states, semantic OCR
status badges, relative timestamps, 1-based page indicators, large-deck PPT
navigation, and a scratchpad-scope label in split view.
Server (optional backend):
- Persist JWT secret (was per-process random), block path traversal in storage,
fix CORS '*'+credentials, add OCR job ownership checks, last-writer-wins sync
guard, constant-time login, and split out heavy OCR deps so the API/tests run
without them.
CI: Gitea workflows for format+analyze+test (Linux, system sqlite) and a
Windows release build; pristine `flutter analyze`, all Flutter and server tests
green.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
