glyph-vacuum
Measure every streamed glyph or picture under a tree before writing a glyph rule.
Before you write a glyph rule, measure every image embedded in your forms: size, format and how many images each list holds.
How to run it
Full flag list: glyph-vacuum in the CLI reference.
Flags
| Flag | Argument | Meaning | Default |
|---|---|---|---|
--root | DIR | ||
--output | DIR | ||
--append | |||
--db | PATH |
Details
Measure every image your forms stream before you write a glyph rule. It walks the
.dfm / .fmx files under one or more root folders, decodes each Picture.Data /
Glyph.Data blob (wrapper class, format, width, height, bits per pixel, palette) and pairs
it with its count property (NumGlyphs and kin). Use it when you convert a component whose
pictures or glyph strips must survive the conversion and you need to know what is out there
first.
Running it from the CLI
drag-lint glyph-vacuum --root DIR [--root DIR ...] --output DIR [--append] [--db PATH ...]--root is a folder to scan and may be repeated. --output is the folder that receives the
result files; --append adds this run to a measurement already there instead of replacing it.
--db may be repeated.
What you get
In the output folder: instances.tsv (one row per streamed graphic), classes.tsv (one row per
wrapper class), skipped.tsv (blobs it could not decode, with the reason) and an images\
folder with the decoded pictures.
Reaching it in the IDE
No IDE surface -- this is a CLI-only feature. The glyph expressions of a rule book
(G[I/N]) are authored in the ConvRulesEditor; see ConvRulesEditor.
What it needs
The folders to scan. An index is optional (--db).
Where it fits
Run glyph-vacuum first, read classes.tsv to see which wrapper classes and sizes occur, then
write the glyph rule in your book and check it with convert-validate.