Add TTS functionality to EPUB 3 to EPUB 3 and HTML to EPUB 3

*@josteinaj said (Mar 25, 2015):*

I couldn't find another issue/task for this, but it's in the 2015 workplan so I'm adding it.

Two new scripts:
- EPUB3 to EPUB3 w/TTS
- XHTML to EPUB3 w/TTS
- could be a simple chain of the XHTML to EPUB3 and EPUB3 to EPUB3 w/TTS scripts

A useful feature for the EPUB3 to EPUB3 w/TTS script would be to support full-text partial-audio input. A couple of use-cases:
- Some parts of the book might be considered less important, such as an index, footnotes, or a list of resources. For cost/benefit reasons, these parts can be left to the TTS rather than human narration.
- Students want to receive textbooks as soon as possible, but human narration takes time. One human narrated chapter at a time can be made available to the student, with the remaining chapters using TTS.
- Announcements at the beginning of the book, such as a "this book belongs to Patron Name" statement, legal agreements between libraries and publishers, or reading system specific instructions, can change after the human narration is complete and would be fine to narrate with TTS.

Locate the existing XHTML to EPUB3 and EPUB3 to EPUB3 scripts referenced in the issue and read their entry points and tests. Determine the expected TTS workflow, including partial human-narrated and TTS audio, then verify that both requested conversions have defined behavior and coverage.

Written by the indexing model from the issue text.