The editing copy of the art that a PDF-compatible .ai carries now holds:
- Transparency: `mode opacity isolate knockout shape Xy` (the blending
mode's number, Knockout Group 0 off, 1 on, 2 neutral), written when it
changes: in a path's paint after its colors, before a guide, and after
a group's end followed by `0 0 Xd` and `6 () XW`. A group without
transparency of its own ends `9 () XW`, and a group's name comes
before its state block. A layer's transparency follows its art: `Xy`,
`0 0 Xd`, `7 () XW`, `LB`.
- Clipping groups as `q`, the art, the clipping path or compound path last
with `h` (or `H`) and `W` before its painting operator, `Q`, as section
5.6.2 of the version 7 specification shows them
(http://mercator.elte.hu/~saman/hu/okt/AI7FileFormat.pdf). A clipping
layer whose clipping object is a path or compound path sets the last
`Lb` flag and writes its clipping path last.
- Opacity masks as an `AI9_Mask` entry in the object's dictionary: the
mask's art between `X=` and `X+`, and Clipping, Inverted, Disabled and
Linked.
Before, clipping groups were written as their art cut to the clipping
path, and transparency and opacity masks were left out. Still left out,
with a warning: the opacity and blending mode of single fills and
strokes, opacity masks on layers (disabled ones too), and opacity masks
nested more than 8 deep. A clipping group or layer whose clipping object
isn't a path or compound path is written as a group of the art cut to
that object, in a layer that doesn't clip, and so are the clipping
groups inside that art. A clipping path's paint that a plain fill and
line can't draw is written as paths inside the clipping, fills under the
art and strokes over it, each side in a group of its own when it has
several, with the clipping path's transparency, lock and hidden state;
the clipping path's opacity mask stays on the clipping path, with a
warning. A clipping group without members is an empty group, and an
empty clipping layer is an empty layer.
Each of these is written in the form VectorCraft's .ai reader reads, and
the tests read the written data back through that reader.
Reading .ai editing data: an `Xy` after a layer's last group or compound
path, followed by the layer's `7 () XW`, is the layer's (a following
`6 () XW` keeps it the group's); groups, compound paths and layers keep
Opacity & Mask Define Knockout Shape.
The `.ai` option `editingData` and docs/file-format-parity.md describe
what the copy now holds.
A PDF-compatible .ai (or a PDF saved with the editing data) now opens
from its layers without its PDF part being imported, and without being
compared with its page, when:
- its text document makes the type of every text object, hidden ones
included, at least one text object shows, and none is drawn by the
appearance of the layer or group it is in;
- its layers hold no unread object, no placed file with a second
placement and no live envelope with drawn art to fall back on, and
no image too large to read;
- it declares no large canvas, and its artboards are its pages' trim
boxes one for one, in points or in the first page's UserUnit.
A file whose layers draw differently from its page then opens from its
layers instead of its PDF part. The bleed comes from the pages' bleed
boxes. The check runs only when no pages are picked, layers and the
editing data are on, the crop box is asked for and type comes in as
text. Files saved without PDF compatibility, files whose PDF part
draws only type or nothing, and files whose PDF part's content can't
be read take the same path when they qualify: their bleed comes from
the pages' bleed boxes, pages in units larger than a point draw them
at their size in points, and the note that the PDF part couldn't be
read is not given, since its content is not read.
Files that don't qualify open as before, and the editing data decoded
and read for the check is reused after the PDF part is imported
(layered_ai_from).
When a file whose PDF part can't be read is scaled up to its size in
points, its text wraps are now laid out again at that size, and the
default graphic styles of a file without graphic styles of its own are
no longer scaled. The same applies on the new path.
A plug-in object that names a brush kind but ends before the brush's
name was dropped without a trace. When it was the file's only brush
object, the document kept the default brushes and showed no note. It is
now counted with the other brushes that can't be read, so the file gets
the Brushes panel note and no default brushes.
A blend's record in the editing data holds its Blend Options spacing.
Word 1 is the spacing: 0 is Smooth Color, 1 is Specified Steps and 2 is
Specified Distance. For Specified Distance, word 2 is the distance in
points as a 32-bit float. The reader took only records whose word 1 is
1, so every Smooth Color and Specified Distance blend came in as its
drawn steps.
A Smooth Color or Specified Distance blend now opens live with that
spacing where VectorCraft lays out as many steps between each pair of
keys as the file draws, in the same places and colors. Otherwise it
keeps its drawn steps with the existing note. The distance is read as
the shortest decimal that reads back as the stored float. The steps are
counted before anything is laid out, and a blend takes from the
import's budget only when its count matches. Specified Steps blends
are read as before, with the steps counted from the drawn ones.
Tests: import::ai::tests::a_blend_keeps_its_spacing,
a_blend_of_three_keys_keeps_its_distance_between_each_pair,
a_blend_whose_spacing_doesnt_lay_out_its_steps_keeps_them,
import::ai::blend::tests::records_say_their_spacing,
a_blend_takes_from_the_budget_only_when_it_is_laid_out,
a_distance_past_the_keys_lays_out_no_steps, and the fuzz case
ai_blend_spacing_never_panics.
A .ai file whose PDF part has a page for each artboard had its layers
compared with the first page only. Art that the layers drew on a later
artboard and its page didn't, or that a later page drew and the layers
didn't, opened from the layers with no warning, and a text object that a
later page doesn't draw stayed.
When the PDF part has a page for every artboard, each page is now
compared with the layers on its artboard, and the comparison fails when
any page differs by more than it allows. Text objects that a later page
doesn't draw are weighed there and left out, unless their layer doesn't
print. Art that differs on a later page is handled as on the first page:
such files now open from their PDF part with the usual warning, where
they opened from their layers before. A PDF part with pages for some of
the artboards, or with its pages cropped to the art, is still compared
on its first page only. Pages are compared until their comparison
drawings take 8,388,608 pixels (about 67 letter-size pages, or 4 when the
PDF part is only images), and at most 128 pages are compared. The type
share in the import's note is now that of the page whose type differs
most.
The layers are drawn once for the pages of the same size and cut into
each page's part. Each page moves by less than half a pixel to start on
a whole pixel, and the column or row the move takes past its artboard
isn't compared. Each page's drawing and result are kept until the layers
change, for the last two documents weighed. Taking a text object out
draws again only the pages within reach of it (every page when type
wraps).
An artboard, or the page of an EPS's editing data, whose sides are within
half a point of its page's is drawn on an area of the page's size, so
both drawings have the same number of pixels. Before, such a file could
open without its layers when the half point changed the rounded size of
its drawing, even when its layers matched the page.
Docs: docs/mcp.md, docs/target-app-parity.md.
When an effect on a layer or a group draws copies of its art, the editing
data holds the type of each copy inside the drawn look. The reader skipped
that type, so the copies came in without text objects, the page's type for
them was given to other groups by painting order, and the import noted that
a text object may be in another group than it was.
The reader now keeps such type as a text slot without a story, in the
copy's group. The page's type that lies over a copy's other art (at least
half of it in the art's box) goes to that copy's slots by painting order,
before the rest is given out. Of overlapping copies, the one that holds
more of it takes it, and the smaller one on a tie. When the page's type
outside the copies is too little for the other text objects, a copy with
more of the page's text objects than slots leaves the extra ones to them.
A slot that gets no type
is removed and not counted as a text object that couldn't be read.
Type beside a copy's art, and every copy's type in a file past the import's
text-object limits or opened with its type as outlines, goes as before. A
text object with an appearance keeps its own type as before.
The comparison of an image-only page draws at four times the comparison
scale and resizes the result to the comparison size. The two sizes round
differently along the shorter side, so a pixel is not 1/scale points
wide. A box at the far edge of a large page moved by up to 0.625/scale
points and could lie outside the object that holds it, so the import
used the PDF part instead of the layers. The boxes now use the size the
raster was drawn at.
When the layers fail the comparison with the file's own page because some objects draw
differently (an object in another color, or one that only one side draws), the import finds
each top-level object whose bounds hold a patch of such pixels (the smallest one that holds it)
and replaces it with the page's art in its bounds, in the same way as an unread object. It then
compares the layers with the page again. When that comparison passes, the layers stay, and a
warning names the layer or sublayer of each replaced object (`<Layer>` for one without a name)
and how many were replaced on it. Like the warning for unread objects, this warning counts as
a loss, so an export or Save As to the file's own path needs `acknowledgeLoss: true`. A
replaced wrap object keeps its wrap options, and area type is wrapped again around the page's
art. Placed files at their other placement are tried too when the first placement can't be
kept.
The file still comes in as its PDF part, with the same reason, when only edges or type differ,
when a difference lies outside every top-level object, when more than half of the top-level
objects (or more than 64) hold one, when such an object is or holds type or holds a wrap object
inside a group, when the page's type of a text object that stays would be copied into a
replacement, when the page's art can't replace an object for the reasons it can't replace an
unread one, when matching the differences to the objects would take too much work, or when the
second comparison still fails. Trusting the editing data keeps the objects as read.
Each artboard's DisplayMark was written as 0. It now holds the center
mark (1), cross hairs (2) and video safe areas (4), as the reader reads
them.
A pattern fill that would place more tiles than a fill places came back
with no tiles and was written as an empty group, so the shape was
missing from the editing copy. Such a shape is now filled gray, as a
pattern that isn't found is, with a warning.
Artboards that together did not fit on the 16,383 pt canvas were
written with %AI24_LargeCanvasScale: 1, so part of them lay outside
the canvas. The editing data now declares a scale of 10 for them and
writes the art, boxes, artboards, strokes and effects at 1/10 of their
size, symbol and pattern art included.
Reading such a .ai, the editing data comes in at its declared scale
when it matches the page that way rather than in the units of the
page's UserUnit, also when the PDF part can't be read, and a file saved
without its PDF part applies the declared scale too.
`eexec` dropped its file operand and skipped the encrypted part of a
font program to `cleartomark` in the program's own file. With
`currentfile eexec` inside a file run with `cvx exec`, the program goes
on in that file, so the skip searched the outer file instead and failed
or skipped the wrong text. Ghostscript's eps2write runs each embedded
Type 1 font program that way, so an EPS with one did not open.
`eexec` now skips in its operand, through the same file reading as the
other data operators: the program's own file or the file being run. A
string operand skips in the file being read, as before.
With one Objects on Path group of two or more objects selected, the
Selection and Objects on Path tools show a square Space widget beyond the
last object, ahead along the path. Dragging it moves the last object
along the path, around its bends, as far as the pointer moves along it,
and the objects between follow evenly while the first one stays. With
Alt (Option on macOS) the objects spread from both ends toward or away
from the center of an open path, or of the first anchor of a closed one.
The distance between the objects shows next to the pointer in the
document's units; the drag is one undo step, and Escape cancels it.
- object.objectsOnPath.options takes spacing (points, or "auto", the
default even spread) and start (points from the first anchor). A
spacing the path has no room for is drawn smaller and kept, so a
longer path brings it back. Older files read as Auto.
- Moving the group, Pivot, Rotate, path edits, copies and saving keep
the Space. Scaling, rotating or shearing the group keeps the first
and last objects where the transform takes them on the new path, live
shapes included; a value that cannot be kept that way scales with the
path's length.
- The Rotate All widget follows the first object when Space moves it.
Behavior: https://helpx.adobe.com/illustrator/desktop/manage-objects/arrange-objects/rearrange-and-edit-objects-on-a-path.html
and https://helpx.adobe.com/illustrator/using/objects-on-path.html.
A filter over `currentfile` waits until it is first read, and it took
its data from whichever file was running by then. One made over the
program's own file and first read inside a file run with `cvx exec`
read that file instead, and one made over a file being run, first read
in a file that one runs, read the inner file. The pending stream now
records the file it was made over (a chained filter inherits it), and
`materialize` decodes from that file.
A file run with `cvx exec` was read whole and run as a separate
program, while `currentfile` stayed the outer EPS. The PLRM has
`currentfile` return the topmost file being run, positioned after the
last token read. Ghostscript's eps2write runs each page's content and
each Type 3 glyph from a SubFileDecode filter, and its `ID` reads the
inline image's samples through `currentfile`, so they were read from
the outer file and then ran as program.
A file being run is now read object by object from its own data, and
while it runs it is `currentfile`. A filter made over it waits to
read, as one over the program's own file does, so it starts where the
program is when the filter is first read.
`known` raised a typecheck for any key that can't be a key here (an
array, a dictionary, a real with a fraction). The PLRM allows any
object but null as a key, and since `put` refuses such keys, no
dictionary holds one: `known` now answers false, and null still raises
the typecheck. Ghostscript's eps2write prolog looks each inline-image
value up in a dictionary of abbreviations, and an image with two
filters has an array there.
The standard encodings were names in systemdict only, and the Encoding
category was empty, so `/StandardEncoding /Encoding findresource`
failed with undefinedresource. The PLRM (3rd ed.) has both as
instances of the Encoding category, the same arrays as the names.
Ghostscript's eps2write prolog takes StandardEncoding that way for a
font whose encoding is a Differences dictionary with no BaseEncoding,
so files with its Type 3 text stopped there.
`gcheck` and `scheck` answered false for every operand. The PLRM (3rd
ed., `gcheck`) says they are true for a simple object, and false only
for a composite one in local VM. The eps2write prolog copies its
initial graphics state into global VM and leaves a value alone only
when `gcheck` is true. So the integers in the `currenthalftone`
dictionary took the copy path, left the stack one item off, and the
next `put` failed with a typecheck in `cp2g`.
Simple objects now answer true. Composites still answer false, since
every composite here is in local VM.
A scaled instance of a symbol with 9-slice guides was written with its art stretched as a whole. The editing copy now places it as the canvas draws it, and leaves the instance out when placing it would pass the symbol art limit.
A group (`u`), clipping group (`q`) or compound path (`*u`) opened after
the last layer and never closed was accepted, and the art inside it was
dropped without a message. The reader now stops with "it ends inside a
group or compound path", as it already does for data that ends inside a
layer. Pattern and symbol definitions, which are read inside a group of
their own, are not affected.
dotCooCoo