Raw script to build a quick and dirty preview from an existing full EPUB. Useful to send out previews of existing ebooks.

previewbuilder.py

#! /usr/bin/env python
 
# by Gabriele Alese <[email protected]> / http://www.alese.it
# Released in Public Domain where applicable: http://creativecommons.org/publicdomain/zero/1.0/
# ***
# REQUIREMENTS:
# this script makes use of pyepub 2.0.9
# download here: https://github.com/gabalese/pyepub
# ***
# USAGE:
# $ python previewbuilder.py [files...]
# output will have the same filename, with "preview_" prepended
 
import sys
try:
    from pyepub import EPUB
except ImportError:
    print "pyepub module not found: see https://github.com/gabalese/pyepub"
    sys.exit(1)
import mimetypes
import os
from StringIO import StringIO
import xml.etree.ElementTree as ET
 
def previewfy(path):
    epub = EPUB(path, "r")
    # if no type="text" is found, provide 20% of content
    num = int(len(epub.info["spine"]) / 100.00 * 20.00)
    items = [x for x in epub.info["spine"][:num]]
    itemslist = []
    for item in items:
        for manifest in epub.info["manifest"]:
            if item["idref"] == manifest["id"]:
                itemslist.append(manifest["href"])
 
    fakefile = StringIO()
    output = EPUB(fakefile, "w", title=epub.info["metadata"]["title"], language=epub.info["metadata"]["language"])
 
    src = []
    for i in itemslist:
        from htmlentitydefs import entitydefs
        parser = ET.XMLParser()
        parser.parser.UseForeignDTD(True)
        parser.entity.update(entitydefs)
        filelike = StringIO(epub.read(os.path.join(epub.root_folder, i)))
        root = ET.parse(filelike, parser)
        map(src.append, [os.path.normpath(os.path.join(os.path.dirname(os.path.join(epub.root_folder, i)), x.attrib["src"]))
                         for x in root.findall(".//*[@src]")] +
                        [os.path.normpath(os.path.join(os.path.dirname(os.path.join(epub.root_folder, i)), x.attrib["href"]))
                         for x in root.findall(".//{http://www.w3.org/1999/xhtml}link[@href]")])
 
    src = list(set(src))  # remove multiple references
 
    # add non-part manifest items
    for i in src:
        output.additem(epub.read(i), i.replace(epub.root_folder+"/", ""), mimetypes.guess_type(i)[0])
 
    # add selected parts
    for i in itemslist:
        output.addpart(epub.read(os.path.join(epub.root_folder, i)), i, "application/xhtml+xml")
 
    output.close()
    output.writetodisk("preview_"+os.path.basename(path))
 
if __name__ == '__main__':
    for i in sys.argv[1:]:
        previewfy(i)

Please note: I know the files generated by this script do not pass epubcheck. Maybe I'll amend that, but don't hold your breath.