A simple library for converting ebooks in XHTML format into EPUB.
xhtml2epub --write-template-dir book_directory
xhtml2epub --input-xhtml book_directory/index.xhtml --output-epub converted_book.epub
When writing or editing ebooks, it is often convenient to store them as a single XHTML file which can be viewed with a web browser and edited using a normal text editor, and then converted to EPUB later for reading on other devices such as a phone or ebook reader. This library provides a simple way to do this conversion.
xhtml2epub requires the XHTML source to follow two conventions to be processed
properly. First, some basic information about the book should be defined in the
form of XML internal entity declarations in the document's DTD. Currently,
xhtml2epub recognizes the following entities as book metadata:
titleauthor(or authors, but as a single string either way)language(such as "en")uid(a unique identifier, such as a UUID (urn:uuid:00112233-4455-6677-8899-aabbccddeeff) or an ISBN (urn:isbn:987123456789X)
All of these are optional, but recommended, especially if you want the resulting EPUB file to conform to the EPUB standard. (Why would you want a book without a title, anyway?)
Secondly, in the XHTML file, any <div> element with an id attribute set is
assumed to be a separate chapter. If the <div> has a title attribute set,
that is used as the title as shown in the book's table of contents. Otherwise,
the title will be auto-detected from the contents of any h1, h2, or h3
heading elements immediately after the opening <div>; or if there are no such
headings, from the id itself.
Chapter <div> elements may be nested to create sub-chapters, typically shown
in ebook readers as hierarchical trees. For example, a book with a body
structure like this:
<div id="chapter-1">...</div>
<div id="part-1">
...
<div id="chapter-2">...</div>
<div id="chapter-3">...</div>
</div>
<div id="chapter-4">...</div>
will show up in the table of contents as something like:
- Chapter 1
- Part 1
- Chapter 2
- Chapter 3
- Chapter 4