Paragraph Style Options and New Paragraph Style gain a Tabs section. It
lists the style's stops and edits the selected stop's Alignment, position
(X:, or Y: for vertical type), Leader and Align On character. Add puts a
left stop 36 pt past the furthest one, Delete Tab removes the selected
stop, and Clear All Tabs leaves the style with no stops. OK sorts the
stops by position and keeps the stop edited last when two share one
position. Reset Panel removes the style's tabs.
paraStyle.new, paraStyle.setAttrs and paraStyle.edit check changed tabs
stop by stop: at most 256 stops, and each changed stop's position from 0
to 100000 pt and leader of at most 8 characters. Stops the style already
has are kept as they are. Changed tabs are sorted by position, keeping
the first of stops at one position. text.tabs.set rejects a leader longer
than 8 characters, an Align On value of more than one character and more
than 256 stops. Loading paragraph styles skips a style whose indents,
spacing, hyphenation, justification or tabs are out of range.
Docs: docs/mcp.md, docs/target-app-parity.md.
flat_charged and the property-key path now root the rope itself, so a string held only by
native code (a getter's result for ++o.x, a valueOf result, a toString key) still counts
what it holds when the reservation collects. The per-call roots for these cases are removed.
When the flattened copy of a String object's concatenated string would pass the heap limit
before a collection can run, the indexed read walks the rope to the unit instead, taking a
unit of fuel for each node it visits, so no copy is made.
Inside braces, a line such as `#include$size = 1` was taken for an `#include` line, because only the ASCII letters, digits and `_` after the `#` were compared with the directive names. The name now runs over every character a name may hold, written out or as an escape, before the comparison.
- wrapping a string no longer flattens it, so Object(s).length and new String(s) on a large concatenated string make no copy and cannot collect while a call's operands are off the stack
- reading a String object's character by index, `in` and hasOwnProperty count the flattened copy at the next limit check, and the run ends before the copy when the fuel for it is not left
- unary plus on an object whose valueOf returns a concatenated string, and a property key from toString, keep that string rooted while its copy is reserved
- the unshift comment cites step 6 of section 15.4.4.13
The drag out of the rulers' corner was ended only by the rulers' own
input, which the canvas skips while the rulers are hidden (View › Rulers,
Presentation Mode). Hiding them mid-drag left the drag's snapping guides
on the canvas and its state open until the rulers came back.
The canvas now ends the drag, cancelled, while the rulers are hidden,
through the same helper a release uses.
The zero point dragged out of the box where the rulers meet snaps as a
Pen or shape tool's point does: with Smart Guides onto the anchors,
centres, edges and paths of the art and the artboards or into line with
them, with Snap to Point onto the anchors and along the ruler guides,
else onto whole pixels or the grid. The guides showing what it snapped
to show during the drag and go when it is released or cancelled
(Escape, focus loss, another document). The origin is still set once,
on release (document.setRulerOrigin, one undo step).
The engine snaps it (Session::ruler_origin_drag, with the tools crate's
rulerguide::OriginDrag over PointSnap::drawn), as it does a guide
dragged out of a ruler.
Under --automation-read-root, prefs.get, prefs.list, prefs.set and
prefs.reset returned the folder preferences (fontsFolder, pluginsFolder,
scratchPrimary, scratchSecondary, recoveryFolder, templatesFolder) and
the export presets' folder even when they were outside the read roots.
Such a folder preference is now null, as ui.inspect leaves it out, and
such a preset's folder is left out. An empty folder and the scratch disk
choices stay; without roots the replies are unchanged.
A filled open path wrote its outline as an open polyline, so with fill
hatches left out (and in R12, and in CNC / CAM) the area lost its closing
edge. Fill outlines now close each open subpath that can enclose an area;
strokes of the same path stay open.
A manifest whose `commands` value is not a list, whose entry is not an
object, or whose command `id` or `label` is not a string was refused with
serde's own text ("invalid type: ..., expected command [0] as {id, label,
params?}"), and no catalog had rows for it or for the reader's other
`commands` messages. The reader now gives its own message for each of
these cases, and the complete-message catalogs translate them, together
with "missing field", "duplicate field" and the manifest prefix with its
line and column.
docs/gaps.md, docs/file-format-parity.md and docs/ui-parity.md still
listed the CNC preset as open after #5500 added it. They now name DXF
Options' Preset menu and dxfPreset: "cnc" on document.exportDxf; VCarve
QA stays open. The ui-parity row also no longer lists native CIRCLE/ARC
export as pending.
An approval question that a control request, MCP, injected input, or a
file dialog one of them opened leads to now keeps OK disabled and
refuses every confirm, a person's click included. The dialog says to
choose Object › Plug-ins › Install Plug-in…, and ui.installPlugin answers
approval: "personMustStartInstall". The person's own Install Plug-in…,
File › Open or drop inspects the module again and asks a new question
with its half-second delay.
The origin is fixed when the request is made and goes with a native
file dialog's deferred answer and the web picker's answers, which now
arrive apart from drops and from window.vectorcraft.openFile (always
automation). A question about taking another module's place now opens
in the approval dialog, and automation's replace: true no longer skips
it. A module that needs no approval still installs at once.
Boxing the string receiver of a non-strict function flattens it, and reserving the copy can
collect. The callee, receiver and arguments are already off the stack at that point, so they
stay in the temporary roots until the frame holds them; before, an argument object could be
freed and its cell reused by the wrapper. new String(value) roots the converted string the
same way, so what the rope holds is still counted while its copy is reserved.
Convert To Point Type left Dynamic Text on when it turned Block area
type into point type. Point type has no Dynamic Text, but the flag was
saved, and converting the type back to area type turned Block on again
and fitted the lines to the new frame. The conversion now turns it off;
the point type keeps the sizes and leading Block wrote into the
characters, and the conversion is still one undo step.
- A name may hold \u{...} escapes with any number of hexadecimal digits
up to U+10FFFF. An empty, unclosed, out-of-range or surrogate escape is
not part of a name. Each \u{...} escape read in a name is reported as
syntax newer than ES3, with the escape as written.
- # followed by an escaped name start is one private name, also in the
code of an XML literal, and its escapes are read in the report
(this.#app is #app).
- A # and a name at the start of a line is a private name, not a
directive, when the name starts with an escape, or when the line is
inside braces and the name is not a directive's. A private name whose
decoded name would make the line a directive is reported as written.
- The identifier test against vectorcraft-js also compares braced
escapes.
Docs: docs/mcp.md, docs/target-app-parity.md.
draw_shape's rotation also turns rectangles and ellipses; object.reflect's numeric axis is counter-clockwise from horizontal. A test covers that axis and the Scatter brush's rotation range, which every_turn_param_says_which_way_it_turns can't reach.
BFlat