#4466 landed `{name}` / `{serial}` / `{date}` file naming templates and `startingSerial` on main while this branch was open. The merge keeps that contract on top of the branch's `batch::Naming`: `{serial}` pads to three digits (`{serial:1..4}` explicit), `{date}` is the UTC `YYYY-MM-DD`, a template without `{ext}` gets the output extension appended, `naming` must be a string of at most 255 bytes, `startingSerial` is 1..4294967295 and the serials of a run must stay within that, and the whole plan is checked before anything is written (`plan_outputs`): no output may replace another input before it is read. `file_cmds/batch_naming.rs` and `process_outputs` are superseded; #4466's tests stay and pass. The Batch dialog starts with a blank template (hint `{name}.{ext}`), `startingSerial` replaces `startSerial` everywhere (engine, lens batch, CLI validation), docs/batch-naming.md describes the merged tokens.
2.9 KiB
Batch output naming
Select a recorded action, then open File > Automate > Batch. Set the source (a folder, with
Include All Subfolders mirroring its tree under the output folder, or the open documents),
the output folder and the format. File Naming is a template for the output file names; leave
it blank to keep the source names ({name}.{ext}). The tokens follow Photoshop's File Naming
fields:
| Token | Value |
|---|---|
{name}, {NAME}, {Name} |
The input's name without its extension: as is, upper case, capitalised |
{serial} |
The input's position plus Starting serial (default 1), padded to three digits |
{serial:1} … {serial:4} |
The same serial padded to that many digits (never truncated: {serial:2} after 99 is 100) |
{letter}, {LETTER} |
The serial as letters: a … z, aa, ab … (upper case with {LETTER}) |
{date} |
The UTC date at the start of the batch, YYYY-MM-DD |
{date:mmddyy}, {date:mmdd}, {date:yyyymmdd}, {date:yymmdd}, {date:yyddmm}, {date:ddmmyy} |
The same date in Photoshop's formats |
{ext}, {EXT} |
The output format's extension, lower or upper case |
Any other text is literal; braces in an input's name are literal too. A template without {ext}
names a stem and gets the extension of the output format, so {name}_{serial} with starting
serial 7 writes photo_007.png and photo_008.png for two PNG inputs named photo.png. Folder
inputs use the listing order (each folder's files by name, then its subfolders), and an explicit
input array keeps its order. A failed input keeps its position, so later names do not shift.
The whole run is planned before anything is written. Unknown tokens, unmatched braces, a
template that is not a string or is longer than 255 bytes, path separators, .., reserved
filename characters, a starting serial outside 1 to 4,294,967,295, and a run whose serials would
pass that are rejected first, with no output folder created. So is a template whose output
would replace another input before that input is read ({serial} over the input folder would
turn 002.png into 001.png); choose another destination or template. Per-file read, action and
save failures still appear in the batch's errors result (and in the error log with Log Errors
To File; Stop For Errors stops at the first and reports the rest as skipped). When two
results share a name, including a case-only difference, the first is kept and the later
collision is reported.
The same options are available on file.automate.batch (and file.scripts.imageProcessor, the
lens-correction batch and the CLI's batch) through the engine command interface:
{
"input": ["first/photo.png", "second/photo.png"],
"output": "exports",
"steps": [["image.adjustments.invert", {}]],
"format": "png",
"naming": "{name}_{serial}",
"startingSerial": 7
}
The CLI's batch --naming <template> --start-serial <n> takes the same template.