A recorded save in the template format writes a copy and leaves the
document's file as it was, so it no longer clears the warning or the
refusals. A recorded Save or Save As with its own path gives a never-saved,
converted or restored document its file before Save and Close, so the
batch runs and warns that every data set saves over that one file.
With a destination folder and Override Save on, a recorded Save or Save As
step writes each data set to its own file in that folder and makes that
file the document's own, so Save and Close then saves each data set to its
own file. The batch no longer warns that every data set saves over one
file in that case, and no longer refuses a document that was never saved,
converted or restored, since the recorded step gives it a file first.
A recorded PDF, EPS, DXF, Save for Web or Save for Office Documents step
that gives a `path` writes the same file for every data set, because the
batch does not point these steps into the destination folder. The one-file
rule now counts them, so the dialog and `action.batch` warn.
Save and Close over the data sets of a document converted from an older
version or restored by Data Recovery is refused with "save the document
with Save As before Save and Close" before a data set is selected. Saving
such a document does not write its file, so the batch reported success
without saving anything. The message has a row in each of the 26 catalogs.
Batch… with a Data Sets source offers None, Save and Close and Folder again.
Save and Close saves the open document over its own file after each data
set. When every data set saves over the same file, the dialog shows a
warning and `action.batch` returns `warning`: with Save and Close, or with
a recorded step that writes a file and that the batch does not point into
the destination folder. A Save step without a path counts when it writes
the document's own file; Save As, Save a Copy, Save as Template and Export
steps count when they give a path; Export for Screens counts when its
folder, given or from its preset, names a folder. Save and Close over the
data sets of a document that was never saved is refused with "the
document has never been saved" before a data set is selected.
Data sets share one open document, so Save and Close saved every data set
over the same file and only the last one survived. action.batch now
returns an error for source datasets with destination saveAndClose before
it selects a data set or runs a step, and the Batch dialog offers None and
Folder for a Data Sets source, moving a Save and Close choice to Folder.
An importer plug-in (kind "importer", permission import.read) lists 1 to 8
file extensions and turns a file's bytes into objects, which open as a new
document whose artboard is their bounds. File › Open lists its extensions,
File › Open With (file.openWith) opens a file with a chosen importer, and
document.open takes importer and params. A built-in format that reads the
file by its content or extension always opens it; among importers listing
the same extension, the one whose id sorts first opens it. An importer reads
at most 128 MiB, and Place still reads only the built-in formats.
Docs: docs/plugins.md, docs/mcp.md, docs/gaps.md, docs/target-app-parity.md.
A folder source's stem is picked once the file is open, so the files a
save or export writes for each artboard count, as they do for data sets.
With the destination folder the same as the source folder, saving
a.vectorcraft with separateArtboards wrote a-Artboard-1.vectorcraft over
the source file of that name; a.vectorcraft is now saved as a-2.
Output stems for a folder batch went through safe_file_name, which turns
`\` and `:` into `-`, so `a\b.svg` on Linux was written as `a-b.svg`.
The stem now only loses trailing periods and spaces and gets `-` after a
Windows device name, and keeps every other character.
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 `-`.
With `includeSubdirs` and the destination folder the same as the source
folder, the Export for Screens folder of `a.svg` was `a`, and its
`Artboard-1.svg` replaced the source `a/Artboard-1.svg` before Batch
opened it. Batch now keeps the names of subfolders of the destination
folder that hold source files, so that export folder is `a-2`.
With the destination folder the same as the source folder, `a..svg` gets
the name `a` and was saved as `a.svg`, replacing the source `a.svg`
before Batch opened it. Batch now keeps the names of the source files in
the destination folder for their own output, so `a..svg` is saved as
`a-2.svg` and `a.svg` is saved over itself.
With the Folder destination, `a.svg` is saved as `destinationFolder/a.svg`.
A later source `a.svg.svg` then got the Export for Screens folder `a.svg`,
which already existed as a file, and its export failed. Batch now numbers a
source's name when its folder or any file it writes in the destination
folder has the name of a file or folder an earlier source wrote, so that
folder is `a.svg-2`.
A source file named `a..svg` gave the folder `a.` and a data set named
`a.` gave `Untitled_a.`. Windows drops a trailing period from a folder
name, so these folders were the same as the ones for `a.svg` and the
data set `a`, and the second export replaced the first one's files.
Batch now passes each name through `safe_file_name` before it numbers
repeated names, so the second folder is `a-2` or `Untitled_a-2`.
A misspelled parameter key got a did-you-mean suggestion, but a
misspelled command id only got "unknown command `select.al`", through
the CLI, MCP and the control channel alike. Session::execute now names
the nearest command id, using the same matching as parameter keys
(ignoring case), and action.batch does the same for an unknown step. An
id with no near match keeps the plain error.
The matching skips candidates whose length differs by more than the
allowed number of edits, so a very long unknown command id or parameter
key is answered at once instead of being measured against every command
id or accepted key.
The new message has a row in every catalog, and docs/mcp.md's
Parameter keys section shows it.
Over a folder, action.batch named each source file's output after the
file without its extension. Two files with the same name, such as a.svg
and sub/a.svg with includeSubdirs, got the same Export for Screens
folder, Override Save path and Folder destination file. The second
file's output replaced the first's, and the batch reported no failure.
Batch now adds -2, -3... to the name of each later file whose name was
already used, compared without case, as it does for data set names. The
numbers follow the order the batch processes the files. With the Folder
destination, a.svg and a.ai in one folder now give a.svg and a-2.ai.
Docs: docs/mcp.md, docs/target-app-parity.md.
With Override Export on, action.batch gave each document.export,
document.exportSelection and document.exportForScreens step a path in
the destination folder. document.exportForScreens reads folder and has
no path parameter, so its files still went to the step's own folder,
or came back as data when the step named none. assets.export was not
overridden at all.
Override Export now gives a document.exportForScreens or assets.export
step the folder destinationFolder/<stem>, in place of the step's own
folder or an export preset's. Each source file gets its own folder,
named after the file without its extension, and each data set gets one
named <document>_<data set>. The export names the files in that folder
as it does outside Batch. Two data sets, or two source files whose
names without the extension differ, no longer write over each other's
files when their artboards or assets have the same names. Other export
steps are unchanged.
Docs: docs/mcp.md, docs/target-app-parity.md.
Export for Screens, Batch over data sets, Save for Web and the
Libraries panel name files after text the person typed: a prefix, a
suffix, a name pattern, the name of an artboard, asset, layer, slice,
data set or library, or the document's title. The prefix, the suffix,
name patterns and Batch's data set names had only /, \ and : replaced,
so on Windows a name holding <, >, ", |, ?, * or a control character
failed to save, and Save for Web without a path named its files after
the document's title as it was. A row folder name could end in a
period, which Windows does not allow, and names such as CON, NUL, COM1
or LPT1, which Windows reserves for devices, were written as they
were.
file_part, which Export for Screens' prefix and suffix and Batch's
data set names use, now writes all of those characters as -, and name
patterns use it in place of their own copy. A new safe_file_name runs
a whole file or folder name through file_part, removes trailing
periods and spaces, and adds - after a device name that is the whole
name or comes before its first period (CON.png becomes CON-.png).
Export for Screens uses it for each file, folder and zip it names, and
keeps item names distinct after it. Save for Web uses it for slice
images and, without a path, for files named after the title. Library
files use it too. This happens on every platform. Other names are
written as before, except where a changed name now equals another
item's name, which then gets a number added.
Docs: docs/mcp.md, docs/target-app-parity.md.
With Override Save or Override Export on, action.batch named each
step's file in the destination folder with the step's format exactly as
written. The save and export commands read the format in any case and
with a leading dot, so a step with "format": ".svg" wrote
"Untitled_a..svg", "SVG" wrote ".SVG", "png8" wrote ".png8" and
"template" wrote ".template".
The file now ends in the format's extension: the extension as written,
in lower case and without the dot, when it is one of the format's
(".JPEG" gives ".jpeg"), and otherwise the format's first extension, the
one Export and Save use for a file name ("png8" gives ".png", "template"
gives ".vctemplate"). An unknown format ("xyz") keeps its spelling, and
the step fails with the command's error as before.
Docs: docs/mcp.md, docs/target-app-parity.md.
Batch over data sets names each output file after the document and the
data set, as the reference app's File + Data Set Name option does. It
joined that name onto the destination folder as it was, and a data set
name can hold any character. A data set named a/b failed to save, and
one named a/../../x wrote its file outside the destination folder. On
Windows, one named c\d also failed to save, and one named e:f left an
empty file.
The data set's name now goes through file_part, which Export for
Screens uses for its file name prefix and a format row's suffix, so
output file names have - in place of /, \ and : from a data set's name.
When two data sets then give the same name, compared without case, the
later ones get -2, -3 and so on, as artboard file names do, through a
distinct_names helper split out of unique_file_names. This covers the
folder destination and Override Save/Export.
Docs: docs/mcp.md, docs/target-app-parity.md.
On main, cargo xtask wasm printed an unused-import warning for json in
crates/engine/src/cmd/fileio/action_batch.rs, and this check reported
it as an error:
RUSTFLAGS=-Dwarnings cargo check -p vectorcraft-engine --target wasm32-unknown-unknown
Every use of json! in action_batch.rs is in code compiled only for
native targets. The wasm32 version of action_batch uses Value only.
The json import now has the same #[cfg(not(target_arch = "wasm32"))]
attribute as the code that uses it. Native builds import the same
names as before.
Play an action on every file in a folder or every data set, with destination
and Override Open/Save/Export. Actions panel and File › Automate › Batch…
open the dialog; new strings in pt-br.
Co-authored-by: Cursor <cursoragent@cursor.com>
dotCooCoo