A catalog's search rows are now checked together: a word that two rows
add to different groups fails the i18n test, naming the word and both
groups. A word repeated for one group, and a catalog with no search
rows, still pass.
A font can place a glyph by the glyph after it, so the glyphs before a
mismatch are laid out again as a line of their own before they replace
their outlines. The glyph budget covers that second check.
Readings of a shape are capped by those tried, not those kept, so a shape
whose every reading fails stops at the cap. The layout check grows a
window from each line start, doubling while the type draws the outlines,
and starts the next line at a mismatch, so a line split at every glyph
costs about twice its glyphs, not half their square; a glyph budget
leaves anything past it as outlines. A tracked line of 450 glyphs: 259 ms.
The lone-glyph rule looks at lines as the outlines lie, before the layout
check splits them, so tracked one-contour letters (L, I, T, H, E) become
type. A shape keeps every glyph and pose it fits exactly, upright first
(an L, and a turned L half round; an I upright or upside down), and the
line around it chooses; a one-glyph line may turn to another candidate
to take the next glyph.
The whole-outline check flattens the shape once per shape, samples every
corner and points twice the tolerance apart, and asks a flat uniform grid
of segments. With Microsoft YaHei loaded: 300 CJK outlines in 350 ms
(1.9 s before), one line in 25 ms. FontDb::outline reuses
outline_uncached, and the compositing test covers knockout_shape.
Measure effect-bearing Align to Path content after spine placement, retaining
the conservative envelope and covering damage containment with original
software-rendered move and Undo controls.
A transform no longer leaves locked guide paths in place: a guide in a
group moves with its group again, as on main. Keeping them put made
Align, Transform Each, the pixel-grid snap and Objects on Path measure or
place art the transform did not move. Scaling::keep_guides and its
engine wiring and test are removed.
The marquee rule now covers the rest of the selection side, while guides
are locked or hidden:
- the guide rule goes down the walk as an explicit Rules value instead
of a thread-local guard;
- an Enclosing Lasso loop does not take a compound path for its guide
contours;
- an enclosing marquee measures a group without its guide paths
(Node::without_guides);
- Direct Selection's marquee takes no anchor or crossed segment of a
guide path.
Docs: the Guides area row, its evidence entry and the M1 row describe
the marquee behaviour; status lines and revision rows say a guide in a
group still moves with it.