File > Open and File > Place read BMP files named .dib or .rle (run-length
encoded bitmaps included), Photoshop files named .pdd and PNG files named
.pns, in any letter case. The Open and Place dialogs list them in "All
readable files", "All placeable files" and their format's own filter, and
Glass's Load Texture lists them too. Linked files with these names update
and relink, the Links panel shows their format, and Edit Original opens
them.
The extensions are read-only: exports still write .bmp, .psd and .png, and
an export to a path with one of the new extensions is refused as before.
The desktop packaging does not associate them with the app.
Found while testing crates/engine/src/cmd/fileio/zip.rs:26 (store, deflate) vs :178 (Archive::open)
Co-Authored-By: Claude Haiku 5.5 <noreply@anthropic.com>
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
Found while testing crates/engine/src/cmd/fileio/naming.rs:103
Co-Authored-By: Claude Haiku 5.5 <noreply@anthropic.com>
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
Export for Screens compared names and folders by upper-casing each
character, which turned dotless ı into I. A row with the folder ı then
wrote into the folder i, and an artboard named ı was numbered as a
second i. NTFS keeps ı apart from i and I, and so does Unicode case
folding outside Turkic languages, so fold now leaves ı as it is.
Fixes#2763. An object whose editing-data dictionary holds the text wrap
options keeps them (offset in points, 6 pt by default, clamped to
±1000 pt; Invert Wrap), and the objects after the text in a version 7
wraparound group (`*w` … `*W`) are wrap objects at the default options.
Area type below a wrap object in its layer flows around it before the
layers are compared with the page, and threaded type flows on into its
next frames. A Large Canvas file's offset is scaled with its art.
The wrap resolver moves into `vectorcraft_doc::text::resolve_wrap` for
the import and the edit-time refresh, and builds each wrap object's
outline once per layer. A file whose wrap shapes would copy more than
2,000,000 anchors is not read from its editing data.
Docs: docs/file-format-parity.md.
The status bar said "Opened with 1 note(s)", "Saved <path> with 1 note(s)",
"Discarded 1 recovered document(s)" and "Restored 1 recovered document(s)",
and the Global Edit options showed "Result(s): 1". They now use the singular
for 1 and the plural otherwise, in English through i18n::msg_n and in every
catalog through @msgplural rows (the Results line through an @plural row).
The engine's warnings for linked images and placed documents output as their
previews, and its errors for a page past a placed file's last page or
artboard, also name their counts in the singular or the plural.
The catalogs drop the rows no message reads any more: the plain "(s)" rows
of the changed messages and the @msg row for "Packaged {n} file(s){note}".
Follows #5002.
A curve rejected for having more than 256 segments now counts the most points a curve may have against the document's pressure points, so linking it from many strokes reads it a bounded number of times. Once the points are used up, later strokes import as plain strokes with a warning.
A version 7 text object has no arbitrary-angle character rotation: its
character attributes include baseline shift (Ts), scaling (Tz) and
tracking (Tt), Tv only turns Roman characters 90 degrees in vertical
text, and the positioning operators such as Tm are ignored on reading.
Type with Character Rotation was written as editable text with the
rotation left out, so rotated characters opened upright.
Such type is now written as its glyphs' outlines, the way type 8-bit text
can't hold already is, with the warning "type with rotated characters is
written as outlines, which keep the rotation, and is no longer editable
text". The outlines keep each character's rotation, baseline shift and
scale. Type without rotated characters, and type turned a whole number of
times, is still written as text.
Docs: docs/file-format-parity.md, docs/mcp.md, docs/target-app-parity.md.
- Each segment of a pressure curve is also sampled where its width stops
rising or falling, so a dip between the regular samples of a long
curve is kept. The segment's values are scaled into -1..1 before that
is solved, so huge values in a file can't overflow it.
- A curve counts as full width only within 1e-9 of it, so a curve just
under full width keeps its profile.
- The pressure points of all strokes in one read are limited to about a
million, as for characters and table cells. Curves at full width count
too, and once the limit is reached no more curves are read; the
strokes past it are imported as plain strokes with a warning.
An Affinity stroke shaped with the Width tool stores its pressure curve
(LDeP: position along the path against width). The importer warned and
drew the stroke at full width all along.
- The pressure curve is read and sampled into the stroke's width
profile (2 to 8 points per segment, 256 segments at most, widths up
to the full weight). A curve at full width all along stays a plain
stroke; one that can't be read is warned about.
- Brush strokes keep their own warning.
- The pinned Affinity 3 signatures for the five corpus files with such
strokes are updated.
File > Place reads a Word 97-2003 document (a compound file named .doc,
or one that holds a WordDocument stream) and sets its main text as area
type, as it does for .rtf and .docx files. The text comes from the
piece table, in 16-bit Unicode or compressed 8-bit pieces. Formatting
is resolved from the style sheet's default font, the paragraph style,
the character style and the direct character and paragraph properties,
and keeps the same properties the .docx reader maps.
Table cells are separated by tabs and each row is a paragraph, and a
field places as its result. Hidden and deleted text, pictures, headers,
footers, notes, comments and text boxes are left out, and the reply's
warnings name each kind of content left out or changed once. Word 6.0
and Word 95 documents, encrypted documents and documents last saved
with fast save are refused, and so are structures that lie outside
their streams.
The compound file is read with the cfb crate (MIT), which adds one
package to Cargo.lock.
A save in the native, template or .ai formats with separateArtboards writes an
artboard file even when the document has one artboard, so an existing file with
that name now gives the data set the next free suffix too.
When a step can write one file per artboard (Save in the native, template or
.ai formats, or Export in SVG, EPS, the raster formats, DXF, EMF, WMF or .ai)
and the document has more than one artboard, an existing file with one of the
stem's artboard file names now gives the data set the next free -2, -3 suffix,
as the stem's own file does. Names a data set's outputs will take are also
reserved for the later data sets of the same batch.
Destination and Override Save/Export paths reject only the current platform's
path separators, so a folder batch writes `a\b.svg` into the destination
folder on systems where that is one file name.
The naming choice reaches batch_datasets as a field of the run, and the short
Batch entry in docs/mcp.md is back to main's text; the longer Batch section
names datasetFileNaming and when an existing output gets a suffix.
The `-2`, `-3` suffix for a data set output that already exists is now
looked up only when the destination folder and the output path pass the
read and write access checks, so a call cannot learn which files exist in
a folder it may not use.
The lookup uses only extensions of formats the app knows, without path
separators or `..`, and only paths whose parent is the destination folder.
A step whose `format` is unknown keeps its own path instead of being
pointed into the destination folder. A data set or document named `.` or
`..` gets the `Untitled` stem with its three-digit number.
The Batch dialog has a File Name choice for a Data Sets source, starting
with File + Number (`source001`), as well as File + Data Set Name
(`source_Alpha`) and Data Set Name (`Alpha`). `action.batch` takes it as
`datasetFileNaming`; a call without it keeps File + Data Set Name, and a
folder source ignores it. An unknown value is an error.
A data set output whose name already exists in the destination folder
now takes the next free `-2`, `-3` suffix instead of replacing the file,
and a Windows device name under Data Set Name gets a trailing `-`.
File › Open and File › Place read JP2 files and raw JPEG 2000
codestreams (.jpf, .jpx, .jp2, .j2k, .j2c, .jpc) through hayro-jpeg2000,
and store the pixels as PNG. Gray, RGB, CMYK and ICC-tagged images, with
or without alpha, are converted to RGBA. Several codestreams, JPX
composition or layers, other color spaces, inconsistent or oversize
palettes, inconsistent channel definitions, oversize tile grids and
images past the pixel allowance are refused with a message before the
pixels are decoded.
The capture and display resolution boxes size the image: the box that
comes last sets each axis, and an axis without a usable value uses
72 ppi.
IccProfile::to_srgb_into converts a pixel with the cached transform and
no heap allocation.
SVG export ignored an image's colorize paint and wrote the image's own
black, white and gray pixels. EPS export flattened every colorized image
into a process-color raster, so a spot color lost its plate. Both now
write the image in the fill color, with the coverage the canvas and PDF
export use (darkness times alpha), so white and transparent pixels take
no color.
- SVG writes a coverage PNG as a mask over a rectangle in the color,
shared between repeated images while each keeps its own color. Linked
mode writes the coverage PNG as a file, and SVG Tiny 1.2 writes a PNG
in the color. A colorized image whose pixels are missing, or whose
grayscale pixels can't be decoded, is left out with a warning.
- EPS writes fully covered or uncovered pixels as an imagemask in the
color's process or Separation ink (Registration on every plate), and a
grayscale image as ink samples scaled by coverage. Level 3 masks its
uncovered pixels out and Level 2 writes them white; gray midtones are
written over white, with the existing warnings.
- EPS flattening no longer counts a solid colorize paint on a grayscale
or 1-bit image as transparency by itself. Opacity, blending modes,
opacity masks, effects and overlapping transparent art still flatten
it, and so does a colorize paint the writer can't draw.
The svg crate's tests turn on resvg's raster-images feature to render
the embedded PNGs.
Docs: docs/mcp.md, docs/target-app-parity.md.
EPS export expanded the pattern fills whose tile images need flattening and
then flattened the result with flatten_document, which expanded them again
with a new object limit. PostScript printing did the same, on the layers
shown rather than the layers that print. EPS export now flattens the
expanded document without expanding it again (flatten_expanded), and
PostScript printing leaves the expansion to the flattener, which runs on
the layers that print, and only checks for such pattern fills to warn.
Flattening counted every colorized image as transparency, in the art and
in pattern tiles. A new flattener option, keeps_colorize, stops counting
an image colorized with a solid paint whose pixels are grayscale or 1-bit,
for a format whose writer draws such images in their ink. Opacity,
blending modes, opacity masks and effects still flatten them, and pattern
tiles are still expanded for those. The option is off, so exports flatten
as before; the EPS writer that draws colorized images turns it on.
The EPS pattern tile tests now check the exported colors, and a tile
group with opacity for the warning.
PDF 1.3, PDF/X-1a, PDF/X-3, EPS and PostScript printing flatten
transparency with `flatten_document`, which only looked at art in the
layers. An image with see-through pixels, a colorize paint, opacity or
a blending mode drawn only through a pattern fill stayed in the pattern
definition, so the flattener returned early and the writer drew the tile
opaque and uncolorized (PDF 1.3 export failed on the transparency left).
`flatten_document` now first expands pattern fills whose tile image
needs flattening into their tiles, as legacy .ai save does. See-through
tile pixels count only where the caller counts see-through images (PDF
1.3); EPS level 3 already masks them out. The EPS and PostScript print
transparency warnings now see those tile images too.
An Affinity text object whose children are nested in it (an image dragged
into artistic text, for example) opened as a plain group, so the image
was drawn over the text. The text now becomes a clip group whose
clipping path is the type, as a shape with children already does: the
nested art shows only inside the characters, with the characters' fill
behind it and their stroke over it.
Alter Paths for Appearance wrote every stroke as its filled outline, so a
1 pt circle became two circles and a hatch, and an open arc became a
closed outline and a hatch: two cut lines where a CAM program expects
one. The option now outlines only the strokes a line can't draw: the ones
Preserve Appearance outlines, strokes wider than the widest lineweight
(2.11 mm, to the hundredth), and every stroke in R13 and R14, which have
no lineweights. Other strokes stay CIRCLE, ARC, SPLINE and LWPOLYLINE
entities with their lineweights; R12 keeps them as POLYLINEs with their
width.
A dashed stroke whose offset starts part way into its pattern is now
outlined with Preserve Appearance and with Alter Paths: a linetype always
starts at the start of its pattern, so the dashes were written in the
wrong place.
An open path with a single straight segment was written as a two-vertex
LWPOLYLINE (POLYLINE before R14). It is now a LINE entity from its start
to its end point. A stroke with a weight in R12, R13 and R14 keeps its
polyline, since a LINE has no width. Closed paths and paths with more
than one segment are unchanged.
Part of #1299.
Saving a legacy .ai with transparent images flattened expands pattern
fills whose tile image is see-through or has opacity or a blending mode,
so the tile images flatten with the art under them. A pattern whose tile
is an opaque grayscale image with a colorize paint was not expanded, so
the writer drew the source gray pixels over white. A colorize paint now
counts too.
Follow-ups to the review of #4697 and #4698:
- With textAs: outlines, a .ai saved without PDF compatibility, read
from its editing data alone, has its type outlined too, as an empty
page or a text-only placeholder already did.
- Threaded type read from the editing data holds its story in its first
frame: it now flows through its frames before it is outlined, so each
frame's outlines show the text it holds, and the outlined frames
leave their threads.
- Each outline group keeps its type's opacity, blend mode, isolation,
opacity mask and effects (the same carry-over the flattener uses).
- A gradient-mesh corner (/N) holding anything but numbers is unread,
instead of being read with the numbers after it shifted into its
colour and opacity.
A colorized grayscale or 1-bit image draws its white pixels transparent and is written to
PDF with the image as a soft mask. Node::shows_transparency did not count it, so a PDF 1.3,
PDF/X-1a or PDF/X-3 export whose only transparency was a colorized image was not flattened
and kept the soft mask. A legacy .ai save also wrote such an image as opaque.
- shows_transparency counts a visible image with a colorize paint.
- The legacy .ai writer flattens or leaves out colorized images like other see-through images.
Two follow-ups to the empty-page fallback:
- A page counts as empty only when nothing on it was skipped: its only
notes are the page's own (art off the artboards, art past the page). A
page whose only content couldn't be read, such as an image that can't
be decoded, keeps its notes and goes through the usual check against
the editing data, instead of being replaced by it unchecked.
- With textAs: outlines, the document read from the editing data alone
has its type outlined. The usual path takes outlines from the page,
but an empty page (or a text-only placeholder, #1387) has none to give.
A PDF-compatible .ai whose PDF part's pages are cropped to its art (each
page the box of all the art, the editing data's bounding box) rather
than to its artboards came in as that PDF part: the editing data's
artboard wasn't the page's size, so its layers were refused. With
several artboards every page held all the art, so the document had a
copy of everything per page.
When the artboard isn't the page's size but the editing data's bounding
box is, the layers are compared with the page on that box and keep
their artboards. The comparison of their art with the page still
decides whether they are used.
An 8-bit PCX file whose image data is followed by the palette marker
(0x0C) and fewer than 768 color bytes was read as gray levels. The reader
now notes where the run-length encoded image data ends, and when the
palette marker follows it without all 768 colors, refuses the file as cut
short. An 8-bit file with nothing after its image data still reads as
gray levels, and a gray level of 12 that falls 769 bytes before the end of
such a file is read as a pixel instead of as the palette marker.
dotCooCoo