This script is for scanning font files and listing their unicode coverage by unicode blocks. Currently truetype fonts (TTF), opentype fonts (OTF) and truetype collections (TTC) are supported.
It is inspired from the versatile FontForge suite (actual Unicode characters covered by font is listed under Element → Font Info), as well as ttfcoverage website.
Note that due to certain bug in Font::TTF this script may not be able to read all fonts embedded inside TTCs. This is fixed in Font::TTF 1.04.
font-coverage.pl [option…] FONT_FILE…
Invoking font-coverage.pl -h produces a list of options.
Print statistics of each font separately
# font-coverage.pl a.ttf dir/b.otf moredir/c.ttc
Print combined statistics
# font-coverage.pl -c a.ttf dir/b.otf moredir/c.ttc
Using Unicode 5.2.0 as basis, generate CSV output, plus demonstrating other options
# font-coverage.pl -i -s -z -u 5.2.0 a.ttf
Running the script on Musica truetype font produces output like:
..............
Geometric Shapes (U+25A0-U+25FF) => 96 / 1 / 0
Miscellaneous Symbols (U+2600-U+26FF) => 256 / 7 / 0
Byzantine Musical Symbols (U+1D000-U+1D0FF) => 246 / 246 / 0
Musical Symbols (U+1D100-U+1D1FF) => 220 / 220 / 11
Ancient Greek Musical Notation (U+1D200-U+1D24F) => 70 / 70 / 0
Supplementary Private Use Area-A (U+F0000-U+FFFFF) => 0 / 0 / 58
Numbers appearing in output represents, in order:
- Total number of code points assigned for specific Unicode range
- Number of glyphs assigned in unicode standard for that range
- Number of glyphs not assigned in unicode standard for that range
So the output snippet above means all Music symbol related unicode ranges are 100% covered (though not for other ranges), and there are 11 extra glyphs in Musical Symbols not used by current version of Unicode.
Note: all code points in Control Chars, Surrogates and Private Use Areas are treated as unassigned.
The -u option allows one to compare the font against alternative versions of Unicode (by default 12.1.0 is used, as of 2019 ). Include files for latest update of all major Unicode versions (2.1 onwards) have been pre-generated, but if one somehow wants to compare against other versions, it is possible to invoke gen-include-file.pl to generate the ones you need. Please refer to the script itself for detail.
- user options to show/hide PUA
combined count of all fonts(done)- points in No_block & non-unicode glyphs not handled
- well defined codepoint ignore list
- cross platform support
- scan all fonts used by system to get an idea about the system’s unicode coverage in general, like what BabelMap does