Libgutenberg

- add media type for `.zip`
- add aliases to metadata parser for `author of introduction` and `author of afterward'.
- fix off by one error in rendering of BCE dates: 1 BCE is represented as 0.
- fix metadata parser not to raise an error only because there's no ebook number - a title suffices.
- ensures no markup in 'language attributes. this was crashing ebookmaker.
- tests added for DCMIType
- improves title truncation in dc.make_pretty_title(). fixes 33

- add handler for 'original publication' in pg header. This field is not parsed.

- handle $c and $v title subfields
- include subfield marker removal in GutenbergGlobals.insert_breaks
- normalize capitalization of "eBook" in an error message

- dc.update_date initialized to datetime.date.min

- libgutenberg is not compatible with SQLAlchemy 2.0; added a version restriction
- in logging tests, clean up the open log file
- fileinfo didn't work properly in non-pg installations, which depended on symlinks in the filesystem. refactored the file path massaging when storing filenames.
- curly quotes in titles are straightened.
- store subtitle in a marc subfield
- pubinfo was not being reconstituted when loaded from the database
- add support for place in pubinfo
- add spaces after tight commas in rendered pubinfo string
- update json support to version 3.03 of workflow
- remove unused DublinCore.PARSEABLE_EXTENSIONS
- added `update_date` property to DublinCore
- removed code that was adding update date to credit.
- strip updates from db credits.

- load_from_pgheader now treats allows any plural metadata keywords. ending 's' of a keyword it stripped so that "Editors" is treated the same way "Editor" is.
- refactored and delinted DublinCore
- added tests for translator handling