Is there a unix command line tool that can analyze font files?

Question

Given a directory of font files (TTF and OTF) I'd like to inspect each font and determine what style (regular, italic, bold, bold-italic) it is. Is there a command line tool for unix flavored operating systems that can do this? Or does anyone know how to extract the metadata from a TTF or OTF font file?

Answer 1

I think you're looking for otfinfo. There doesn't seem to be an option to get at the Subfamily directly, but you could do:

otfinfo --info *.ttf | grep Subfamily

Note that a number of the fonts I looked at use "Oblique" instead of "Italic".

Answer 2

In Linux, if you have .ttf fonts, you most probably also have fontconfig, which comes with the fc-scan utility. You can parse the output for the information you want, or use the badly documented --format option.

For example:

fc-scan --format "%{foundry} : %{family}\n" /usr/share/fonts/truetype/msttcorefonts/arialbd.ttf

The font properties you can print this way are shown here: http://www.freedesktop.org/software/fontconfig/fontconfig-user.html#AEN21

Some properties are listed in multiple languages. For example, %{fullname} may be a list. In that case, %{fullnamelang} will list the languages. If that shows you your language in fourth position in the list, you can use %{fullname[3]} as the format string to print the full name in only that language.

This language stuff being quite inconvenient, I ended up writing a full Perl script to list the info I wanted in only one language:

#!/usr/bin/perl
use strict;
my $VERSION = 0.1;
my $debug = 1;

my @wanted = qw(foundry family fullname style weight slant width spacing file);
my @lang_dependent = qw(family fullname style);
my $lang = "en";

my $separator = ", ";


use File::Basename;
use Data::Dumper; $Data::Dumper::Sortkeys = 1;


my $me = basename $0;
die "Usage: $me FILENAME\n" unless @ARGV;

my $fontfile = shift;

unless (-f $fontfile) {
    die "Bad argument: '$fontfile' is not a file !\n";
}


my $fc_format = join( "\\n", map { "\%{$_}" } @wanted );

my @info = `fc-scan --format "$fc_format" "$fontfile"`;
chomp @info;

my %fontinfo;
@fontinfo{@wanted} = @info;

if ( grep /,/, @fontinfo{ @lang_dependent } ) {
    my $format = join( "\\n", map { "\%{${_}lang}" } @lang_dependent );
    my @langs = `fc-scan --format "$format" "$fontfile"`;

    for my $i (0..$#lang_dependent) {
        my @lang_list = split /,/, $langs[$i];
        my ($pos) = grep { $lang_list[$_] ~~ $lang } 0 .. $#lang_list;
        my @vals = split /,/, $fontinfo{$lang_dependent[$i]};
        $fontinfo{$lang_dependent[$i]} = $vals[$pos];
    }
}

warn Dumper(\%fontinfo), "\n" if $debug;

$fontinfo{'fullname'} ||= $fontinfo{'family'}; # some old fonts don't have a fullname? (WINNT/Fonts/marlett.ttf)

print join($separator, @fontinfo{@wanted}), "\n";