Python library to extract 'epub' information

Question

I'm trying to create a epub uploader to iBook in python. I need a python lib to extract book information. Before implementing this by myself I wonder if anyone know a already made python lib that does it.

Answer 1

An .epub file is a zip-encoded file containing a META-INF directory, which contains a file named container.xml, which points to another file usually named Content.opf, which indexes all the other files which make up the e-book (summary based on http://www.jedisaber.com/eBooks/tutorial.asp ; full spec at http://www.idpf.org/2007/opf/opf2.0/download/ )

The following Python code will extract the basic meta-information from an .epub file and return it as a dict.

import zipfile
from lxml import etree

def get_epub_info(fname):
    ns = {
        'n':'urn:oasis:names:tc:opendocument:xmlns:container',
        'pkg':'http://www.idpf.org/2007/opf',
        'dc':'http://purl.org/dc/elements/1.1/'
    }

    # prepare to read from the .epub file
    zip = zipfile.ZipFile(fname)

    # find the contents metafile
    txt = zip.read('META-INF/container.xml')
    tree = etree.fromstring(txt)
    cfname = tree.xpath('n:rootfiles/n:rootfile/@full-path',namespaces=ns)[0]

    # grab the metadata block from the contents metafile
    cf = zip.read(cfname)
    tree = etree.fromstring(cf)
    p = tree.xpath('/pkg:package/pkg:metadata',namespaces=ns)[0]

    # repackage the data
    res = {}
    for s in ['title','language','creator','date','identifier']:
        res[s] = p.xpath('dc:%s/text()'%(s),namespaces=ns)[0]

    return res

Sample output:

{
    'date': '2009-12-26T17:03:31',
    'identifier': '25f96ff0-7004-4bb0-b1f2-d511ca4b2756',
    'creator': 'John Grisham',
    'language': 'UND',
    'title': 'Ford County'
}