Drops: libraries, presets and plug-ins open on a drop, as File › Open opens them

With a document open, the desktop app ran File › Place on a swatch or
graphic style library, a Libraries panel file, a presets file or a
plug-in dropped on it. A desktop drop has no position, and without one
the app opened only documents and ran File › Place on every other file.
On the web, the app also ran File › Place on these files when they were
dropped on the canvas. File › Place can't read them, and the drop
failed with "Couldn't place X: ... not a format VectorCraft reads".

VectorcraftApp::drop_target now returns Open for these files, with or
without a drop position, and they open as File › Open opens them:
plug-ins are installed, presets are imported, swatch and graphic style
libraries open in the library panel, and a Libraries panel file is
added to the Libraries panel as a new library. A dropped library,
presets file or plug-in that has a path is added to Open Recent Files
when it opens, as File › Open adds it. io::opens_as returns the kind of
each of these files from its extension, and drop_target and
io::open_bytes both call it. The rules from #472 still apply to
documents, pictures and text. On the web the browser gives no file name
before the drop, so the app still outlines the canvas while such a file
is dragged over it.

Tests: drop_target returns Open for each of the eleven extensions, with
no position and with a position on the canvas. A .gpl dropped through
the desktop drop path loads in the library panel and is added to Open
Recent Files, and a .vclibrary dropped the same way becomes a new
library in the Libraries panel.

Docs: docs/control-protocol.md (file.place), docs/development.md
(Drag-and-drop), docs/target-app-parity.md (Evidence, UI).

docs/control-protocol.md

@@ -69,7 +69,7 @@ the two can still win that race.
 | `engine.execute` `ui.savePdfDialog` | `{path?, preset?, range?, …document.exportPdf options}` | opens the Save PDF dialog (`file.export.pdf` with no params opens it too). Its fields are the `document.exportPdf` options: set `compatibility` with `ui.dialog.set {field: "compatibility", value: "1.5"}`, a whole section with `{field: "compression", value: {…}}`, the artboards with `__allArtboards: false` and `range`, the visible section with `__section` (General, Compression, Marks and Bleeds, Output, Advanced, Security, Summary). `ui.dialog.confirm` writes `path` (else asks for one), returns `{path, bytes, warnings}` and opens the file when `viewAfterSaving` is set |
 | `file.save` / `file.saveAs` / `file.saveCopy` / `file.saveAsTemplate` (via `engine.execute`) | `{path?, format?, options?, svg?}` | Save writes the document's own file in its own format (a document opened from or saved as SVG/PDF saves as that again; the result's `warnings` say what the format loses). With `path` the file is written at once (Save As or a Copy to an SVG path without SVG options first opens SVG Options, below). Without one (never saved, a converted older file, Save As, a Copy, a Template) the app shows a save panel with one file type per save format (`vectorcraft`, `template`, `pdf`, `svg`, `svgz`), then the format's options: SVG opens the `svgOptions` dialog (below; returns `{"pending": "svgOptions", "path"}`), PDF the `savePdf` dialog (its `__save` field names the save command); `ui.dialog.confirm` finishes the save. On the web a `saveOptions` dialog (fields `path`, `format`) first names the file and picks the format. `file.formatOptions {format?}` reads the options a save would use. `file.revert` opens a `confirm` dialog (`{"pending": "confirm"}`); `ui.dialog.confirm` reloads the saved file in the same tab |
 | SVG Options (`engine.execute` of `file.export.svg`, `file.saveAs`, `file.saveCopy`) | `{}` / `{path?, svg?: {…}}` | `file.export.svg` without params, and Save As or Save a Copy to an `.svg`/`.svgz` path without SVG options, open the `svgOptions` dialog (`ui.dialog.set` its fields: `styling`, `outlineText`, `images`, `objectIds`, `decimals`, `minify`, `responsive`, `fewerTspans`, `metadata`, `preserveEditing`, `useArtboards`, `allArtboards`, `range`, `showCode`, `profile` (`svg11`/`tiny12`), `encoding` (`utf8`/`utf16`/`latin1`), `embedFonts`; `ui.dialog.confirm` exports or saves and remembers the choices). With `svg: {…}` (the `document.formats` SVG options) they run directly. `app.save {path?, svg?}` takes the same options; a document saved as SVG saves as SVG again with the same options |
-| `engine.execute` `file.place` | `{paths?}` / `{path \| name+dataBase64, …file.place params}` | with no file, picks files and opens the `place` dialog; with `paths`, opens it for them (fields `files`, `link`, `template`, `replace`; `__replace` says whether Replace applies, `__info` describes each file). `ui.dialog.confirm` places one file centred in the view (or in place of the selected object) and loads the place cursor with several (`file.place.queue`). With a file, it runs `file.place` centred in the view. Files dropped on the window with no document open open. On the web, files dropped on the canvas are placed at the drop point (linked; Shift embeds) and files dropped off it (the tab bar) open, as in Illustrator; desktop drops carry no pointer position, so documents (native, SVG, PDF, `.ai`, EPS, DXF, EMF/WMF) open as tabs of their own and pictures and text are placed in the middle of the view. With the cursor loaded (`ui.inspect` → `tool: "place"`, `toolOptions: {count, current, name, names}`), `ui.pointer` clicks and drags place files and `ui.key` ←/→/↑/↓ and Escape cycle and discard them |
+| `engine.execute` `file.place` | `{paths?}` / `{path \| name+dataBase64, …file.place params}` | with no file, picks files and opens the `place` dialog; with `paths`, opens it for them (fields `files`, `link`, `template`, `replace`; `__replace` says whether Replace applies, `__info` describes each file). `ui.dialog.confirm` places one file centred in the view (or in place of the selected object) and loads the place cursor with several (`file.place.queue`). With a file, it runs `file.place` centred in the view. Files dropped on the window with no document open open. On the web, files dropped on the canvas are placed at the drop point (linked; Shift embeds) and files dropped off it (the tab bar) open, as in Illustrator; desktop drops carry no pointer position, so documents (native, SVG, PDF, `.ai`, EPS, DXF, EMF/WMF) open as tabs of their own and pictures and text are placed in the middle of the view. Swatch libraries and color books (`.vcswatches`, `.gpl`, `.ase`, `.acb`), graphic style libraries (`.vcstyles`), Libraries panel files (`.vclibrary`), presets (`.vcflattener`, `.vcpdfpresets`, `.vcprintpresets`, `.vcperspective`) and plug-ins (`.wasm`) open wherever they are dropped, as File › Open opens them. With the cursor loaded (`ui.inspect` → `tool: "place"`, `toolOptions: {count, current, name, names}`), `ui.pointer` clicks and drags place files and `ui.key` ←/→/↑/↓ and Escape cycle and discard them |
 | `app.export` raster options | `{path?, format: png \| jpg \| webp, ppi?, background?, antiAlias?, interlaced?, quality?}` | `ppi` sets the pixel size (72 = one pixel per point) and is stored in the file; `background` is `transparent`, `white`, `black` or `"#rrggbb"`; `antiAlias` is `none`, `art` or `type` (text snapped to pixels); `interlaced` writes an Adam7 PNG. The PNG Options dialog (`pngOptions`; `jpgOptions` / `webpOptions` for the other raster formats) has the same fields: fill them with `ui.dialog.set` and press OK with `ui.dialog.confirm` |
 | `app.export` JPEG options | `{path?, format: jpg, quality?: 0–100, colorModel?: rgb \| cmyk \| gray, method?: baseline \| optimized \| progressive, scans?: 3–5, embedIcc?, imageMap?: none \| client \| server}` | `cmyk` writes ink amounts in the working CMYK space (CMYK colours keep their inks; RGB ones are separated with the colour settings); `embedIcc` (default true) embeds sRGB, the working CMYK profile or a grey profile, generated by the CMS; `imageMap` also writes `<stem>.html` (client-side `<map>`) or `<stem>.map` (NCSA) beside the image for the objects with a URL and an Image Map shape (`attributes.set`), returned in `linked`. The `jpgOptions` dialog has these fields too (`quality` is shown 0–10; a CMYK document defaults to `cmyk`) |
 | `app.export` PNG-8 and GIF | `{path?, format: png8 \| gif, colors?: 2–256, reduction?: perceptual \| selective \| adaptive \| web \| blackWhite \| gray, dither?: none \| diffusion \| pattern \| noise, ditherAmount?: 0–100, transparency?, matte?, interlaced?, …raster options}` | a palette of at most `colors` entries (art with that many colours or fewer keeps them exactly); pixels under half opacity become one transparent entry unless `transparency` is false, partly transparent edges are blended over `matte` (`none` keeps their colour). `png8` writes an indexed `.png` (1–8 bits per pixel); Export for Screens rows take both. The `png8Options` / `gifOptions` dialogs have these fields |

docs/development.md

@@ -27,7 +27,7 @@ How the web shell (`apps/vectorcraft-web/src/web.rs`) differs from desktop:
 
 - **Open** sets `Services::open_async`, which shows `rfd::AsyncFileDialog`. The bytes arrive in `Services::inbox`, which the app drains every frame.
 - **Save / Save As / Export** go through `Services::download`: a Blob, an object URL and a temporary `<a download>`, all created from Rust. There is no save dialog, so the suggested name becomes the download name.
-- **Drag-and-drop:** `WebShell` takes the frame's `dropped_files` before the app sees them, reads each with `DroppedFile::bytes_async` and pushes the bytes into the inbox (files dropped on a canvas go to `Services::place_inbox` with their `place::DropAt`: the document they were dropped on, by uid, and the point; `VectorcraftApp::drop_target` decides from the position the page's drag events give). Desktop drags carry no position (winit 0.30 reports none, and the pointer position egui last had is stale), so the desktop app decides by kind: documents open, pictures and text are placed in the middle of the view. A file that arrives after another document became active still lands in its own document, which becomes active again; if that document closed meanwhile, the file is not placed and the status bar says so. (The app's synchronous drop path is compiled out on wasm32.)
+- **Drag-and-drop:** `WebShell` takes the frame's `dropped_files` before the app sees them, reads each with `DroppedFile::bytes_async` and pushes the bytes into the inbox (files to be placed on a canvas go to `Services::place_inbox` with their `place::DropAt`: the document they were dropped on, by uid, and the point; `VectorcraftApp::drop_target` decides from the position the page's drag events give). Desktop drags carry no position (winit 0.30 reports none, and the pointer position egui last had is stale), so the desktop app decides by kind: documents open, pictures and text are placed in the middle of the view. On the web and the desktop, swatch and graphic style libraries, Libraries panel files, presets and plug-ins open wherever they are dropped, as File › Open opens them (`io::opens_as`). A file that arrives after another document became active still lands in its own document, which becomes active again; if that document closed meanwhile, the file is not placed and the status bar says so. (The app's synchronous drop path is compiled out on wasm32.)
 - **Data Recovery** keeps its copies in `localStorage` (`vectorcraft-recovery/<area>/<name>`, shared by the site's tabs), so tabs can't lock their areas the way desktop apps lock a folder (`RecoveryStore::lock`). Each tab holds its area with a heartbeat it refreshes every minute and, where the browser has Web Locks (secure pages: `https://` or localhost), with a Web Lock named after the area (`apps/vectorcraft-web/src/locks.rs`, through `js_sys::Reflect` since web-sys has them only as an unstable API). The browser releases the lock only when the tab is gone, not while its timers are paused in the background, which a heartbeat can't tell apart (#367). `navigator.locks.query()` is asynchronous, so the shell keeps a snapshot refreshed every 10 seconds (taken once before the first frame); a snapshot older than a minute counts every area as held. An area is offered as a crash's leftovers only when no tab holds its lock and its heartbeat is older than three intervals. Without Web Locks a paused tab can still be taken for gone; it repairs that when it resumes: its next heartbeat (or recovery save) writes again the copies missing from its area, and a heartbeat newer than the one the other tab wrote taking the area over makes the area running again for that tab.
 - **Losing the graphics (#369):** the browser can drop the page's WebGPU device or WebGL context (a GPU reset, a driver update, too many tabs on the GPU; or a script calling `device.destroy()` / `WEBGL_lose_context.loseContext()`). The app goes on without it, so the shell watches for it: wgpu's device-lost callback for WebGPU, the canvas's `webglcontextlost` event for WebGL2. Both report to a `graphics::GraphicsLoss`, and the next frame of `WebShell::logic` handles it:
   1. `VectorcraftApp::graphics_lost` writes Data Recovery copies of the modified documents at once (to browser storage, also when Data Recovery is turned off).

docs/target-app-parity.md

@@ -381,6 +381,7 @@ Moved from the ROADMAP's "Shipped so far" on 2026-10-10. Update it when a featur
   - Native macOS menu bar, vector tool cursors, a ⌘K command palette, middle-button panning with any tool. Native menus explicitly end tracking on command selection, outside window presses and Escape; menu updates wait until tracking ends, and a dismissing press never starts a canvas gesture (M14.8).
   - The macOS menu bar is built from the in-window menus (one model, `native_menu`) and laid out as a Mac app's: the VectorCraft menu holds About, Settings ▸ (each Preferences page, General… ⌘K), Language ▸, Appearance ▸ (UI brightness), Services, Hide VectorCraft ⌃⌘H (⌘H stays Hide Edges), Hide Others, Show All and Quit; Window gains Minimize ⌃⌘M, Zoom and Bring All to Front, and Help the system's search field. Its key equivalents go back to egui as key presses, so user shortcuts, focused text fields and the Type tool's input method behave as in the window, and nothing fires twice; enabled, check marks and labels update in place. `--in-window-menus` or `VECTORCRAFT_IN_WINDOW_MENUS=1` keeps the menus in the window; `ui.inspect` reports `nativeMenuBar`.
   - macOS: Finder opens every format File › Open reads with VectorCraft (double-click, Open With, a drop on the Dock icon, `open -a`), at launch or while it runs; the release and `cargo xtask bundle` apps share one `Info.plist` that declares those document types, owning only VectorCraft's own formats. The DMG opens on a designed Finder window, VectorCraft.app beside the `Applications` link over a background with the app icon (#493).
+  - Swatch and graphic style libraries, Libraries panel files, presets and plug-ins open wherever they are dropped on the window, as File › Open opens them.
   - A canvas context menu: right-click selects the object under the pointer and lists what applies to the selection (Undo/Redo, clipboard, group, isolation, join, Remove Anchor Points when anchors are selected, masks, compound paths, guides, Transform, Arrange, Select, Export Selection), or the view commands on empty canvas.
   - Save / Don't Save / Cancel before closing or quitting with unsaved documents (tabs, Close, Close All, Quit, window close).
   - Four brightness themes, persistent preferences.