Using a library's character or paragraph style, and scrubbing a
numeric field, closed their undo step with Session::end_undo_group,
which closes the open undo group of every document. An undo group open
in another document, such as one begun with edit.undoGroup.begin for
an MCP command_batch with atomic, was closed too, and that document's
next edits became undo steps of their own. When the style failed to
apply, or Escape ended the scrub, the other document's group was
canceled as well and the edits made in it were undone. Inside an undo
group already open in the same document, a scrub closed that group,
Escape undid the edits made in it before the drag, and a style that
failed to apply left the style it had added.
Session::begin_undo_scope and end_undo_scope make the edits in between
one undo step of the active document: the scope opens an undo group
there, or joins the one already open. Ending a scope closes only the
group it opened, in its own document, even when another document
became active in the meantime, and a canceled scope undoes only its
own edits, also inside a group that was already open. library.use and
the scrub use the scope, so Escape inside an open group also takes
back a scrub of the Transform panel's Shear field. edit.undoGroup.end
still closes the open group of every document.
When a cancel removes journal entries numbered below a recording's
start, the entries after them are numbered down. The recording's start
and the journal's hold kept the old number, so Stop Recording left out
as many of the recorded steps as there were entries removed, and the
hold no longer covered them. `Journal::remove` now moves the hold down
with the entries, and Stop Recording reads the start from the hold.
`Journal::replace` puts an entry in place of another and counts it at
its new size. A cancel uses it for the `document.activate` that takes
the place of a canceled group's entry that switched documents, so the
journal's size estimate does not keep the old entry's size.
A cancel cut the journal back to where the first canceled group began.
That also removed the commands run in other documents after it began,
which the cancel does not undo, so a replay of the journal lost them.
Each undo group now keeps the numbers of its own journal entries: the
commands that start and end in its document and the commands that edit
it. A cancel removes only those entries, and the later entries are
numbered down to close the gaps (`Journal::remove`). An entry of a
canceled group that also switched to another document, such as a batch
that draws and then activates another document, is replaced by a
`document.activate` of that document, so the commands after it replay
in the same document.
Session::end_undo_group with cancel cancels the open undo group of
every document. edit.undoGroup.end with cancel calls it, and so do a
field scrub ended with Escape and a library style that fails to apply.
The journal was then cut back to where the group of the last document
in the document list began, not to where the first of the canceled
groups began. When the group that began first belonged to an earlier
document, the journal kept the commands of that group's edits, which
the cancel had undone. After a scrub ended with Escape, which journals
no end command, replaying the journal drew those edits again.
Session::end_undo_group now cuts the journal back to the earliest start
of the groups it cancels.
An EPS or PostScript file that stops on a PostScript error named the
error, such as typecheck or rangecheck, with the operator and the
procedure, but did not say what the name means in any language.
The warning or error now gives the name as PostScript spells it,
followed by VectorCraft's own wording of what it means in the reader:
"a PostScript error (typecheck in `add`; a value has the wrong type for
the operator)". vectorcraft_eps::POSTSCRIPT_ERRORS lists the 14 names
the reader raises other than undefined, with their meanings, and a
test checks that each of those names has one. Each meaning and the
new message shapes have a row in every catalog. i18n::message now
translates placeholder values five levels deep instead of three, which
the status bar's line for an EPS file that can't be opened needs.
Docs: docs/mcp.md, docs/target-app-parity.md.
The reference app's graphs help describes adding totals to a column
design: type in the design holds a percent sign followed by two digits.
VectorCraft drew that type as typed, stretched with the design.
Type in a column design that holds a % followed by two digits now shows
the value of each column or stacked segment in place of the % and the
two digits. The first digit is the least number of places before the
decimal point, with zeros added in front (0 for as many as the value
needs), and the second is the number of places after it. Values are
rounded half away from zero. The type keeps its own size and moves to
where the design's copy puts the point it is set from. A repeating
design shows it once per column, where a whole copy ending at the value
end puts it, and Chop Design does not cut it. A sliding design moves it
with the part of the design on its side of the guide. Type inside a
group that clips or has its own opacity, blend mode, appearance or
opacity mask, or inside a blend, compound shape, envelope, repeat or
intertwine, stays in the design and shows the value in each copy. Type
inside a symbol stays as typed. The legend swatch leaves the totals
out.
Docs: docs/mcp.md, docs/target-app-parity.md.
The Tabs panel named its alignment buttons Left-, Center-, Right- and
Decimal-Justified Tab and labeled the position field X for every text.
The reference app's alignments for vertical text include Top-Justified
Tab and Bottom-Justified Tab, and its position field is Y for vertical
text.
In vertical type the Left-Justified button is now named Top-Justified
Tab and the Right-Justified button is named Bottom-Justified Tab, in
their tool tips and for screen readers, and the position field is
labeled Y. The reference page does not say which buttons these are;
VectorCraft's own mapping follows the line direction: a line of
vertical type runs from top to bottom, so a Left stop aligns the text
to the top and a Right stop aligns it to the bottom. Screen readers now
call every alignment button by its name; before, they read its glyph.
In horizontal type the tool tips and the X label do not change.
Docs: docs/target-app-parity.md.
object.objectsOnPath.attach took a path inside a compound path, blend,
envelope, repeat or Intertwine group, or a clipping path, as the path
and put it into the new Objects on Path group, so it was no longer a
part of its container. As objects, a blend's key path, a repeat's
source art, an envelope's content and a clipping path were put into
the new group the same way.
The Objects on Path tool highlighted a blend's key path or a clipping
path and attached the selection to it.
objects_on_path::held now finds the container that holds an object.
The command refuses such a path or object with a reason, and the tool
does not highlight it or attach to it. A member of a compound path
given as an object still attaches as its whole compound path. The four
new messages have rows in every complete-message catalog.
Docs: docs/mcp.md (Objects on Path).
Since #2664, swatch.new separates an RGB color into CMYK through the
color settings in a CMYK document unless keepModel is given. The Color
Guide, Color Themes and Recolor Artwork send keepModel when they save
colors they show, but the Color panel menu's Create New Swatch, the
Swatches panel and the Gradient panel did not. An RGB object's color
saved from them became a CMYK swatch, and a gradient's RGB stops
became CMYK stops.
These now send keepModel: Create New Swatch, an Alt-click on New
Swatch, the New Swatch dialog for a gradient, a Fill or Stroke proxy
or the Gradient panel's thumbnail dropped on the Swatches panel or
Alt-dropped on a swatch of its kind, and the Gradient panel's Add to
Swatches and Save to Swatches. A new swatch_paint_params returns
paint_params with keepModel: true added for an unlinked color or a
gradient. swatch.new without keepModel still converts RGB input.
Docs: docs/target-app-parity.md.
Object › Live Paint › Merge built the merged group anew and copied
only the bottom Live Paint group's name, opacity and blend mode, so
its lock, visibility, appearance, opacity mask, graphic style link and
other properties were lost. Make with a Live Paint group selected, and
a Live Paint Bucket click on a selected path beside a selected Live
Paint group, made a new group with a new id instead, and every face
took the fill of its paths again.
Merge now starts from the bottom Live Paint group, as Gap Options
does, and replaces only its faces, edges and paths and its Gap
Options. Make, and livePaint.fill with ids when the point is outside
that group's faces, add the other objects to the bottom Live Paint
group among them as Merge does.
Docs: docs/mcp.md, docs/target-app-parity.md.
An edge hidden on its own (for example with Object > Hide) while its
Live Paint group stays visible could still be highlighted, selected or
stroked, and a hidden edge nearer the pointer kept a visible one from
being picked.
lp_part_at now takes the nearest visible edge (visible_edge_near), inside
and outside the faces, for the hover, the Live Paint Selection click and
the Bucket's Shift-click. The Bucket's Shift-click sends nothing where no
edge is highlighted, and livePaint.strokeEdge also skips hidden edges, so
the Bucket strokes the edge it highlights.
With Select Fills off, or with Alt held, the Live Paint Selection tool
highlighted an edge near the pointer even when the pointer was outside
every face, but a click there looked only for a face under the pointer
and cleared the selection. Outside the faces, the highlight came from
the front-most Live Paint group with an edge in reach, not from the
nearest edge, and it included hidden groups. The Live Paint Bucket's
Shift highlight and Shift-click used the same search.
The hover, the Live Paint Selection click and the Bucket's Shift-click
now call one function. Outside every face it picks the edge nearest
the pointer, within 4 screen pixels, among the visible Live Paint
groups. A click selects or strokes the edge that the highlight shows.
This change follows up #3052.
Distribute Objects and Distribute Spacing space the selected ruler
guides across the axis (vertical ones for the horizontal options,
horizontal ones for the vertical options) together with the selected
objects, as items with no size. Guides along the axis stay where they
are. Two or more guides selected without objects are enough, and the
Align panel's Distribute buttons are enabled for them. With Align to
Artboard, the outermost items go to the artboard's edges.
A key guide (select.key {guide}) is one of those items. With a spacing,
it stays put and the other items are spaced from it, as from a key
object with no size; a key guide along the axis leaves everything where
it is. The Align panel sends the Distribute Spacing value with both
rows of Distribute buttons while a guide is the key, and enables the
value for it.
The selection tools' clipping-path outline hit was also used where the
Selection and Direct Selection tools act on what is drawn rather than
select. Near a layer's clipping mask edge, with art drawn over it:
loaded text threaded into the unpainted mask instead of the art, a
double-click on type entered isolation on the mask instead of editing
the type (Selection) or did nothing (Direct Selection), and Cmd/Ctrl+
Shift double-click could not release a guide path there. These now
take the ordinary hit (ToolContext::hit_options); a press or hover
that selects keeps the outline hit.
Tests (each fails with the selection hit options):
select::tests::loaded_text_threads_into_the_art_on_a_layer_masks_edge,
a_double_click_on_type_at_a_layer_masks_edge_edits_it (Selection and
Direct Selection) and a_guide_on_a_layer_masks_edge_is_released, over
the new testutil helpers add_masked_layer and
doc_with_type_on_a_layer_masks_edge. Recorded in
docs/target-app-parity.md.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
A clip group answered a press only strictly inside its clipping path,
and there the art above the path came first in the hit test, so where
the clipped art extends past the mask's edge the clipping path's
segments could not be reached on the canvas.
With the new HitOptions::clip_outlines, which the Selection, Direct
Selection and Group Selection tools set
(ToolContext::selection_hit_options), a press within the selection
tolerance of a clipping path's outline, inside or just outside it,
hits the clipping path (HitKind::Outline) before the art it clips:
Direct Selection takes its segments and the Selection tool the clip
group. It applies to a clip group's clipping path and to a layer's
clipping mask, and only to one a click may hit (may_hit: visible,
editable, within reach). Inside the mask the art is hit as before.
The other tools, which act on what is drawn (the eyedropper, type,
gradients...), leave the option off, and hit_test_visible never
applies it, so they still take the art at the mask's edge rather
than the mask's empty appearance. Select Behind's one-pass walk takes
a layer mask's outline first too, so it lists what the repeated hit
test finds. Object Highlighting and a right-click's selection in the
canvas (selection_hit_at) take what those tools take; dropping a
swatch on art still hits what is drawn.
The reference app reshapes a clipping path with Direct Selection:
https://helpx.adobe.com/illustrator/desktop/manage-objects/edit-objects/edit-clipping-masks.html
Tests in doc: a_clipping_path_is_hit_on_its_outline_before_its_art,
only_the_selection_tools_hit_a_clipping_path_on_its_outline and
select_behind_takes_a_layer_masks_outline_first; the random Select
Behind test now clips layers and sublayers and sets clip_outlines,
and fails without the Select Behind change. In tools,
a_press_on_a_clipping_path_takes_its_segment_over_the_art fails
without the change; select::a_press_on_a_masks_edge_selects_its_clip_group
fails when the Selection tool uses the plain options; the eyedropper
copies the art on the edge. Recorded in docs/ui-parity.md,
docs/gaps.md and docs/target-app-parity.md.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
A second finger, a long touch or a cancelled contact replaced Drag::Art
with Drag::Gesture without ending the live Move interaction, which then
outlived the pointer: later edits began on top of an open interaction.
drop_press now cancels the move for Drag::Art too and clears the panel
payload, as it already did for Drag::Tool.
Test: a_second_finger_drops_a_move_waiting_over_a_panel.
The Selection tool ended its Move interaction as soon as the pointer left
the canvas, so the art jumped back to where it started while the pointer
was over a toolbar or panel. As #2826 describes, the art should wait where
it last moved and the move should continue when the pointer returns.
Keep the interaction live while the pointer is off the canvas: Drag::Art
holds the art where it last moved, and coming back clears the panel-drop
payload and sends a Drag so the move goes on with the pointer. A release
over a panel ends the move before the panels draw (release_over_panel,
called from ui_frame), so a drop that takes art still gets it as it was
and no interaction is left open.
Tests: art_moved_off_the_canvas_waits_and_drops_on_the_panel,
art_moved_off_the_canvas_resumes_when_the_pointer_returns and
a_move_released_over_a_panel_ends_before_the_panels_draw cover the canvas
states; a_move_dragged_onto_a_panel_reads_the_menus checks the full app
waits over a panel and ends on release. cargo xtask ci: all 10 steps pass.
When text.tabs.set refused a new X (past 100,000 pt), the panel kept
the stops it had changed and sorted, so later in the same frame the
ruler painted the refused position and highlighted another stop.
apply_moved now moves and sorts a copy of the stops and keeps it only
when the stops are set.
A tab stop moved past another stop in the Tabs panel, by a drag on the
ruler or by a value typed in X, lost the selection. The panel sorts the
stops by position after the move but kept the selected index, so the
ruler highlighted another stop and the X, Leader and Align On fields
showed and changed that stop. With stops at 30 and 50 pt and the 50 pt
stop dragged to 10 pt, the panel selected the stop at 30 pt.
After a drag or an X value re-sorts the stops, the panel now selects
the moved stop at its new place, found by its position as a click that
adds a stop does. When text.tabs.set refuses the new position (an X
past 100,000 pt), the stops and the selection stay as they were.
Docs: docs/target-app-parity.md.
After "nothing in data set A was applied: no object is bound to a
variable it has a value for" showed in the status bar, choosing a data
set that did apply, or stepping to one, left that message in place.
`run_now` set the status from a command's `warning` and cleared it only
for `file.new`, and the Variables panel test cleared the status by hand
before its last check.
`dataset.select`, `dataset.next` and `dataset.prev` now clear the status
bar when they succeed, and the warning shows when no object is bound to
a variable the new data set has a value for. The test no longer clears
the status by hand and checks all three commands.
Pattern Brush Options' Scale and Art Brush Options' Width were number
fields only. The reference app's Brushes page documents a Scale slider
for pattern brushes and a Width slider for art brushes.
Each row now has a slider beside its field, as the Bristle Brush
Options sliders do. The slider covers the value's whole range, 1% to
10000% for Scale and 1% to 1000% for Width, in whole percents. Screen
readers name the slider as its row and read the field's number and
text.
Docs: docs/target-app-parity.md.
docs/control-protocol.md and the tool.options params text say that 0
turns the Pencil and Paintbrush options Edit selected paths within and
Close paths when ends are within off. A stroke that
started exactly on an end of a selected open path still continued that
path at 0, and a stroke that ended exactly where it started was still
closed at 0, because both checks compared the distance with <= 0.
At 0, a stroke that starts on an end of a selected open path now draws
a new path, as one that starts on the path away from its ends already
did, and a stroke that ends where it started stays open.
Docs: docs/target-app-parity.md.
brush.new and brush.options refused a Hue Shift colorization given
without a key color ("invalid brush: missing field `key`"), and only
Brush Options picked a default. The reference app's Brushes page says
that by default the key color is the most prominent color in the art.
A Hue Shift colorization without a key now takes the solid color that
the art's fills and strokes use most, the first one found on a tie, or
black when the art has no solid color, as Brush Options already does.
Brush definitions read from a document get the same default.
brush.options replaced each field its params named, so an edit that
gave only the colorization method or only an art brush's start guide
was refused because the key or the end guide was missing. It now
merges colorization and an art brush's scale field by field.
Docs: docs/mcp.md, docs/target-app-parity.md.
Document Info had no Brushes category, and document.info refused
category "brushes". The reference app's Document Info panel menu lists
Brushes after Graphic Styles.
Document Info now has a Brushes category after Graphic Styles. For the
whole document it lists the Brushes panel's brushes. With Selection
Only it lists the brushes that the selected objects' strokes use, each
once. document.info gives the names as brushNames, the panel's flyout
offers the category, and the text report that Document Info › Save…
and File › Package write has a BRUSHES section.
Docs: docs/mcp.md, docs/target-app-parity.md.
The reference app's View menu has Show Live Paint Gaps, which
highlights the gaps in the selected Live Paint group based on the gap
options set for that group. VectorCraft had no such item.
View > Show Live Paint Gaps (view.livePaintGaps) turns the highlight
on or off, and the item reads Hide Live Paint Gaps while it is on.
While it is on and edges show, the canvas draws the gaps of each
selected Live Paint group up to its Paint Stops At size, the gaps that
livePaint.info lists. It draws them as the Gap Options preview does,
as lines in the group's Gap Preview Color. The setting works as View >
Hide Text Threads does: one setting covers every document, it is saved
with the interface state, and Hide Edges hides the gaps. It is off by
default.
Docs: docs/mcp.md, docs/target-app-parity.md.
The reference app's Brushes panel menu has Select All Unused, which
selects the brushes that are not used in the document. VectorCraft's
Brushes panel menu had no such item, and no command listed the unused
brushes.
brush.unused (Select All Unused) is a query that lists, in library
order, the brushes no stroke paints with. Strokes in the art and its
opacity masks, in symbols, graph designs, pattern tiles and graphic
styles count, and so do strokes in the art of a brush that is used.
The art of a brush that is not used does not count, so a brush whose
own art is the only art that paints with it is listed, and so is a
brush used only in the art of an unused brush. The Brushes panel
menu's Select All Unused selects these brushes, and Delete Brush and
Duplicate Brush then act on them.
Docs: docs/mcp.md, docs/target-app-parity.md.
Select > Object > Bristle Brush Strokes selected the objects whose
brush name contains "bristle". It missed objects with a bristle brush
of any other name, and it selected objects whose brush is of another
type with "Bristle" in its name.
select.object.bristleBrushStrokes now looks up the brush of each
object's topmost stroke in the document's brush library and selects the
object when the brush is a bristle brush, whatever its name.
Docs: docs/mcp.md, docs/target-app-parity.md.
The Artboards panel menu showed Convert to Artboards disabled whatever
was selected, and choosing it did nothing, although Object › Artboards
› Convert to Artboards runs artboard.convertToArtboards on the
selection.
The panel menu's item now uses the same check as Object › Artboards ›
Convert to Artboards, so it is enabled when something is selected, and
it runs the same command. Each selected object that has an area becomes
a new artboard of its geometric bounds and is removed, in one undo step.
Docs: docs/target-app-parity.md.
The reference app makes a document without opening the New Document
dialog with Ctrl+Alt+N on Windows and Option+Command+N on macOS.
VectorCraft had no default key for it.
file.newWithoutDialog (New Without Dialog, default Cmd+Alt+N) makes a
document with the settings New Document opens on: the first Recent
entry, or the Print category's Letter while Recent is empty, named
after the next untitled document. It fills the fields with the
dialog's own open, runs file.new with the params Create sends, and
does not show the dialog. A dialog that is already open stays open.
Keyboard Shortcuts lists it under File.
Docs: docs/mcp.md, docs/ui-parity.md, docs/target-app-parity.md.
The reference app adds a new fill with Ctrl+/ on Windows and Command+/
on macOS, and it adds a new stroke with Ctrl+Alt+/ on Windows and
Option+Command+/ on macOS. VectorCraft had Add New Fill and Add New
Stroke on the Appearance panel, with no default key for either.
appearance.addFill now has the default Cmd+/ and appearance.addStroke
has Cmd+Alt+/, which are Ctrl+/ and Ctrl+Alt+/ on Windows and Linux.
Like the panel's buttons and menu items, each key runs its command with
no params and adds the item to every selected object. Keyboard Shortcuts
lists both under Window › Appearance with their keys, and press_key
runs them for agents.
Docs: docs/mcp.md, docs/ui-parity.md, docs/target-app-parity.md.
testutil: :cx defaults measurement_labels on, so the aspect-ratio drag assertion must not expect a lone Marquee overlay.
Co-authored-by: Cursor <cursoragent@cursor.com>
Honor Show Measurement Labels on the file Place cursor so sizing drags get W/H readouts like other draw tools.
Co-authored-by: Cursor <cursoragent@cursor.com>
With image mips on (the canvas always sets them), drawing an image
fetched its full decoded pixels first, decoding them again when the
shared image store no longer held them, and only then looked up the mip
it draws. The full pixels are only peeked, so they are the least
recently used entries and the first the store drops over its budget
(1 GiB on macOS and Windows). Once a document's full-size images exceed
it (20 photos of 6000 x 4000 px are 1.9 GB as RGBA), every frame decoded
every image again, though only their small mips were drawn: moving one
image redrew seconds later.
The store now remembers the size each blob's pixels decode to, so the
mip level is found without the pixels, and a mip the store holds is
drawn (and marked used) without fetching or decoding them. Without the
mip, at full size, or with mips off (exports) drawing is unchanged.
Measured on an Apple M4 (load average 15-24 from other work), 20 noise
PNGs of 6000 x 4000 px fitted in a 2880 x 1800 view, one moved 5 pt per
frame: before, 20 decodes and 1.1-3.0 s every frame; after, the first
frame decodes 20 and the later ones none, in 3-5 ms.
The readout was a moving average of 1/dt over any two consecutive frames.
An idle window wakes now and then (a timer, a control request, a repaint
after input) and each wake-up draws two frames about 4 ms apart, the second
one not waiting for vsync because the swapchain has a free image. Every
wake-up pushed the average up by 0.1 x 240 and the idle gap pulled it down
by only 0.1 x 2, so at rest it climbed past the display's refresh rate
(105 fps on a 30 Hz monitor that was drawing 4 frames a second). As soon as
the pointer moved and frames became vsync-paced, it fell back to the real
rate, which looked like a slowdown.
- frame_rate: frames counted over the last second of continuous drawing.
A gap more than 100 ms longer than the previous frame's own work means
the window was idle and starts a new count. The readout changes only
after half a second of drawing, so a wake-up's frames keep the last
rate. Slow frames still read slow, because their gap is their own work.
- chrome: "– fps" until the window has drawn continuously, instead of 0.
Measured on an Apple M4 driving a 30 Hz 4K display: drawing continuously
reads 30; at rest with a wake-up every 0.5 s it stays 30 (before: 45 -> 105).
In a clip group or a clipping layer the bottom child is the clipping path.
Object › Arrange moved the selected children with no regard for that slot:
Send to Back (and Send Backward from just above it) put the art under the
clipping path, so the art became the clipping path, lost its role as art,
and the group drew nothing. Bring Forward / to Front on the clipping path
itself would likewise have made art the clip. Layers-panel moves already
keep the clipping path at the bottom (layer.rs, `index.max(1)`); arrange
didn't.
- object.rs `arrange`: in a parent that clips, the clipping path is left
where it is and the art is arranged within the slots above it (Back goes
to just above the clipping path, Backward stops there).
- tests: Back, Backward, several objects, and Forward/Front on the
clipping path in a clip group (with a render check); Send to Back in a
clipping layer.
Include guide visibility in effect raster cache inputs, skip hidden guide paths in outline members, and clear selected guide paths when hiding guides. Add regression coverage for cached effects and selection overlays.\n\nThe cached-effect regression test is adapted from the overlapping #3264 implementation.\n\nCo-authored-by: Mahtab <smmalam9@gmail.com>
Address CodeRabbit review on #3307: the new `Seq::Erase` arm discarded
the `tool_key` Result, dropping an erase error silently, whereas the
old `route_type_input`/`text_key` path reported it via `app.status`.
Mirror that behaviour so a failed Backspace/Delete still surfaces in
the status bar. Behaviour otherwise unchanged; fmt/clippy clean and the
three IME tests pass.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
dotCooCoo