allsorts tools
1.0.0
Allsorts is a font parser, shaping engine, and subsetter for OpenType, WOFF, and WOFF2 written entirely in Rust. This repository contains tools that were developed to debug and test Allsorts and provide examples of its use.
Note: These tools are for demonstration, reference, and debugging purposes. You should not rely on them for production workflows.
Available tools:
bitmaps — dump bitmaps from bitmap fontscmap — print character to glyph mappingsdump — dump font informationhas-table — check if a font has a particular tableinstance — create a static instance of a font from a variable fontlayout-features — print a list of a font's GSUB and GPOS featuresshape — apply shaping to glyphs from a fontspecimen — generate a HTML font speciment for a fontsubset — subset a fontvalidate — parse the supplied font, reporting any failuresvariations — list the variation axes of a variable fontview — generate SVGs from glyphsbitmapsThe bitmaps tool extracts bitmaps from fonts containing glyph bitmaps in
either the EBLC/EBDT or CBLC/CBDT tables.
-o is the path to the directory to write the bitmaps to. It will be created
if it does not exist.The images are written out as PNGs in a sub-directory for each strike (size).
The format is {ppem_x}x{ppem_y}@{bit_depth}, the files are named
{glyph_id}.png:
terminus
├── 12x12@1
│ ├── 0.png
│ ├── 1.png
│ ├── 2.png
│ ├── 3.png
│ ├── 4.png
│ ├── 5.png
│ ├── 6.png
│ ├── 7.png
⋮ ⋮
├── 14x14@1
│ ├── 0.png
⋮ ⋮
└── 32x32@1
⋮ ⋮
allsorts bitmaps -o noto-color-emoji NotoColorEmoji.ttf
cmapThe cmap tool chooses a preferred cmap sub-table and dumps the character to
glyph index entries. If the encoding of the table is Unicode then the characters
are printed along with the code point, otherwise just the numeric value of the
character is printed.
-f, --font specifies the path to the font file.-i, --index is index of the font to dump (for TTC, WOFF2) (default: 0).$ allsorts cmap --font profontn.otb
cmap sub-table encoding: Unicode
'' U+0000 -> 0
'' U+0001 -> 1
'' U+0002 -> 2
⋮
'?' U+003F -> 63
'@' U+0040 -> 64
'A' U+0041 -> 65
'B' U+0042 -> 66
⋮
'»' U+00BB -> 187
'¼' U+00BC -> 188
'½' U+00BD -> 189
'¾' U+00BE -> 190
'¿' U+00BF -> 191
'À' U+00C0 -> 192
'Á' U+00C1 -> 193
'Â' U+00C2 -> 194
'Ã' U+00C3 -> 195
'Ä' U+00C4 -> 196
⋮
dumpThe dump tool prints or extracts information from a font file.
allsorts dump path/to/font prints out information about the font.
--name includes the metadata contained in the name table in the output.-c can be used to print information about a CFF font or table not
wrapped in a TrueType or OpenType container.-t extracts the named table from the supplied font. The output should be
redirected to a file. E.g. allsorts dump -t glyf > glyf.bin-g prints information about a specific glyph in a font.-l prints out all offsets in the loca table in the font.$ allsorts dump noto-subset.otd | head
TTF
- version: 0x4f54544f
- num_tables: 9
CFF (checksum: 0x625ba831, offset: 156, length: 166505)
OS/2 (checksum: 0x9f6306c8, offset: 166664, length: 96)
cmap (checksum: 0x131b2742, offset: 166760, length: 274)
head (checksum: 0x09e560e8, offset: 167036, length: 54)
hhea (checksum: 0x0c1109cf, offset: 167092, length: 36)
hmtx (checksum: 0x1b9b0310, offset: 167128, length: 52)
maxp (checksum: 0x000d5000, offset: 167180, length: 6)
name (checksum: 0x1f3037ad, offset: 167188, length: 418)
post (checksum: 0xff860032, offset: 167608, length: 32)
- CFF:
- version: 1.0
- name: NotoSansJP-Regular
- num glyphs: 13
- charset: Custom
- variant: CID
has-tableThe has-table tool checks if the supplied font file contains the table passed
via the -t argument. If the font contains the table it exits with status
success (0), if the font does not contain the table it exits with status 1.
This tool is handy combined with find, to locate fonts that have the desired table.
-t, --table TABLE table to check for-i, --index INDEX index of the font to check (for TTC, WOFF2) (default: 0)-p, --print-file print the path to the font if it contains the table.-v, --invert-match select fonts that don't have the given tableIn this example, we search the current directory for files ending in ttf,
otf, or otc and check to see if they contain an EBLC table. If the table
is found the path to the font is printed.
find . -regextype posix-extended -type f -iregex '.*.(ttf|otf|otc)$' -exec allsorts has-table -t EBLC -p {} ;
instanceThe instance tool applies a set of values (tuple) to the variation axes of a
variable font to produce a static, non-variable font with those settings.
-t, --tuple is a comma separated list of values one for each variation axis
of the font. The variations tool will list the axes, their order, and limits.-o, --output is the path to the output font.In this example the font has two axes: UNDO and UNDS. We supply a value of
500 for each one and write the output font to UnderlineTest.ttf.
allsorts instance --tuple 500,500 UnderlineTest-VF.ttf -o UnderlineTest.ttf
layout-featuresPrints an indented list of a font's GSUB and GPOS features.
$ layout-features fonts/devanagari/AnnapurnaSIL-Regular.ttf
Table: GSUB
Script: DFLT
Language: default
Feature: aalt
Lookups: 56
Feature: abvs
Lookups: 27,28,29,30
Feature: akhn
Lookups: 4
Feature: blwf
Lookups: 9
# additional output omitted
shapeThe shape tool shapes the supplied text according to the supplied font, language, and
script. It prints out the glyphs before and after shaping.
-f, --font PATH path to font file-i, --index INDEX index of the font to shape (for TTC, WOFF2) (default: 0)-s, --script SCRIPT script to shape-l, --lang LANG language to shape--vertical vertical layout, default is horizontal$ shape -f fonts/devanagari/AnnapurnaSIL-Regular.ttf -s deva -l HIN 'शब्दों और वाक्यों की तरह'
# output omitted
specimenThe specimen tool generates a HTML font specimen sheet containing sample text
set in the font as well as information about the font and its supported features.
-i, --index INDEX index of the font to subset (for TTC, WOFF2) (default: 0)--sample-text TEXT sample text to use in the font specimen$ allsorts specimen ../allsorts/tests/fonts/bengali/Lohit-Bengali.ttf
subsetThe subset tool takes a source font and some text and writes a new version of
the source font only containing the glyphs required for the supplied text.
-t, --text TEXT subset the font to include glyphs from TEXT-a, --all include all glyphs in the subset font-i, --index INDEX index of the font to subset (for TTC, WOFF2) (default: 0)$ allsorts subset -t 'This a subsetting test' NotoSansJP-Regular.otf noto-subset.otf
Number of glyphs in new font: 13
validateThe validate tool attempts to parse all the glyphs (or various DICTs in the
case of CFF) in the supplied font. It reports any errors encountered but is
otherwise silent. This command was useful for bulk testing Allsorts against a
large repertoire of real world fonts.
$ allsorts validate ../allsorts/tests/fonts/bengali/Lohit-Bengali.ttf
$ fd '.(ttf|otf|ttc)$' /usr/share/fonts | sort | parallel --bar allsorts validate {}
variationsThe variations tool lists information about a variable font. The information
includes:
This example prints variation information for the font at
../text-rendering-tests/fonts/TestHVARTwo.ttf.
$ allsorts variations ../text-rendering-tests/fonts/TestHVARTwo.ttf
Axes: (2)
- wght = min: 0, max: 1000, default: 0
- cntr = min: 0, max: 100, default: 0
Instances:
Subfamily: ExtraLight
PostScript Name: TestFont-ExtraLight
Coordinates: [0.0, 0.0]
Subfamily: Light
PostScript Name: TestFont-Light
Coordinates: [150.0, 0.0]
Subfamily: Regular
PostScript Name: TestFont-Regular
Coordinates: [394.0, 0.0]
Subfamily: Semibold
PostScript Name: TestFont-Semibold
Coordinates: [600.0, 0.0]
Subfamily: Bold
PostScript Name: TestFont-Bold
Coordinates: [824.0, 0.0]
Subfamily: Black
PostScript Name: TestFont-Black
Coordinates: [1000.0, 0.0]
Subfamily: Black Medium Contrast
PostScript Name: TestFont-BlackMediumContrast
Coordinates: [1000.0, 50.0]
Subfamily: Black High Contrast
PostScript Name: TestFont-BlackHighContrast
Coordinates: [1000.0, 100.0]
viewThe view tool shapes the supplied text or list of codepoints according to the
supplied font, language, and script. Then, it generates an SVG of the glyphs.
-f, --font PATH path to font file-s, --script SCRIPT script to shape-l, --lang LANG language to shape--mark-origin mark the origin of each glyph with a cross-hair--margin num or top,right,bottom,left specify a margin to be added to the edge of the SVG--fg-colour rrggbbaa set the fill colour of the glyphs--bg-colour rrggbbaa set the background colour of the generated SVG--fg-color rrggbbaa alias for --fg-colour--bg-color rrggbbaa alias for --bg-colour-t, --text TEXT text to render-c, --codepoints CODEPOINTS comma-separated list of codepoints (as hexadecimal numbers) to render-i, --indices GLYPH_INDICES comma-separated list of glyph indices to render-F, --features FEATURES comma-separated list of OpenType features to enable (note: only enables these features)$ view -f fonts/devanagari/NotoSerifDevanagari-Regular.ttf -s deva -t 'खि'
# output omitted
$ allsorts view -f fonts/devanagari/NotoSerifDevanagari-Regular.ttf -s deva -c '916,93f'
# output omitted
In this example, the OpenType pres feature is enabled, which allows glyph 30
to be replaced by its special presentation form (glyph 547).
$ view -f fonts/devanagari/NotoSerifDevanagari-Regular.ttf -s deva --features pres -i '30,54'
# output omitted
Minimum Supported Rust Version: 1.70.0
To build the tools ensure you have Rust installed.
cargo build --releasecargo install --path .There is an AUR package for allsorts-tools:
git clone https://aur.archlinux.org/allsorts-tools.git
cd allsorts-tools
makepkg -si
Contributions are welcome, please refer to the Allsorts contributing guide for more details.
We aim to uphold the Rust community standards:
We are committed to providing a friendly, safe and welcoming environment for all, regardless of gender, sexual orientation, disability, ethnicity, religion, or similar personal characteristic.
We follow the Rust code of conduct.
Allsorts and these tools are distributed under the terms of the Apache License (Version 2.0).
See LICENSE for details.
