This pipeline can be used as a frontend project to create subsets of embedded fonts from a EPUB file. It uses the font tools python libary https://github.com/fonttools/fonttools
Clone the project with:
git clone https://github.com/transpect/epub-font-subsetter --recursive
Usage:
~/epub-font-subsetter $ calabash/calabash.sh xpl/epub-fontsubset.xpl epubfile=path/to/my_epub.epub
Important: Choose your cwd like mentioned above. This pipeline uses an p:exec step to call a bash script, that calls the pyftsubset python script, and unfortunatly choosing the correct cwd is kind of tricky.
The pipeline creates a charset with all used characters for each embedded font and will create a subset called "myfont.otf.subset"
NEW: The main output of this pipeline is an EPUB file with all subsetted fonts embedded, called subset_my_epub.epub. By using the option delete-not-used-font='true' the embedded but not used fonts (no character with this font can be found in the epub) will be deleted.
$ calabash/calabash.sh xpl/epub-fontsubset.xpl epubfile=path/to/my_epub.epub delete-not-used-font=false