epub-embed-fonts.py

#!/usr/bin/env python
# -*- coding: utf-8 -*-

import os
import sys
import zipfile

from BeautifulSoup import BeautifulSoup, Tag # Better for HTML
from lxml import etree # Better for XML

CSS = """
@font-face {
    font-family: 'Droid Serif', serif, sans-serif;
    font-style: normal; font-weight: normal;
    src: url(DroidSerif-Regular.ttf)
}
@font-face {
    font-family: 'Droid Serif', serif, sans-serif;
    font-style: normal; font-weight: bold;
    src: url(DroidSerif-Bold.ttf)
}
@font-face {
    font-family: 'Droid Serif', serif, sans-serif;
    font-style: italic; font-weight: normal;
    src: url(DroidSerif-Italic.ttf)
}
@font-face {
    font-family: 'Droid Serif', serif, sans-serif;
    font-style: italic; font-weight: bold;
    src: url(DroidSerif-BoldItalic.ttf)
}
@font-face {
    font-family: "DroidSansMono", monospace;
    font-style: normal; font-weight: normal;
    src: url(DroidSansMono.ttf)
}
@font-face {
    font-family: "DroidSansMono", monospace;
    font-style: italic; font-weight: normal;
    src: url(DroidSansMono.ttf)
}
@font-face {
    font-family: "DroidSansMono", monospace;
    font-style: normal; font-weight: bold;
    src: url(DroidSansMono.ttf)
}
@font-face {
    font-family: "DroidSansMono", monospace;
    font-style: italic; font-weight: bold;
    src: url(DroidSansMono.ttf)
}
body { font-family: 'Droid Serif'; word-wrap: break-word; }
pre, code { font-family: 'DroidSansMono'; font-size: 85%; }
pre { white-space: pre-wrap; }
"""

FONTS = [
    'DroidSerif-Regular.ttf',
    'DroidSerif-Bold.ttf',
    'DroidSerif-Italic.ttf',
    'DroidSerif-BoldItalic.ttf',
    'DroidSansMono.ttf',
]
FONTS_PATH = '/usr/share/fonts/google-droid/'

OPF = '{http://www.idpf.org/2007/opf}'


if len(sys.argv) != 2:
    exit(1)

filename = sys.argv[1]

os.rename(filename, filename+'.bak')
old = zipfile.ZipFile(filename+'.bak')
new = zipfile.ZipFile(filename, 'w', zipfile.ZIP_DEFLATED)

for name in old.namelist():
    if name == 'content.opf': ### Add items to manifest
        doc = etree.parse(old.open(name))
        manifest = doc.find(OPF+'manifest')
        manifest.append(etree.Element(OPF+'item',
            {'id': 'fonts', 'href': 'fonts.css', 'media-type': 'text/css'}))
        for font in FONTS:
            manifest.append(etree.Element(OPF+'item',
                {'id': font[0:-4], 'href': font,
                 'media-type': 'application/octet-stream'}))
        new.writestr(name,
            etree.tostring(doc, xml_declaration=True, encoding='UTF-8'))
    elif name.endswith('.html'): ### Add links to CSS
        print name
        html = BeautifulSoup(old.open(name))
        head = html.find('head')
        link = Tag(html, 'link')
        link['rel'] = 'stylesheet'
        link['type'] = 'text/css'
        link['href'] = 'fonts.css'
        head.insert(0, link)
        # Fix common problems
        for tr in html.findAll('tr'):
            if tr.parent.name != 'table':
                tr.name = 'div'
        for td in html.findAll('td'):
            if td.parent.name != 'tr':
                td.name = 'div'
        # Special fixes for Slovo
        if str(html.head.title.string).find(' Slovo |') != -1:
            # Clean up title
            html.find('h1').extract()
            h = html.find('h4')
            h.string = h.string[0:-142]
            # Remove sidebar
            html.body.div.contents[3].extract()
        new.writestr(name, str(html))
    else: ### Copy everything else
        info = old.getinfo(name)
        new.writestr(info, old.open(info).read())

# Add files
new.writestr('fonts.css', CSS)
for font in FONTS:
    new.write(FONTS_PATH + font, font)

old.close()
new.close()