code that orchestrates ebook conversion for project gutenberg
EbookConverter manages the creation and update of ebook assets for Project Gutenberg. It uses a postgres database to keep track of both ebook metadata and ebook files. the postgress database is managed by the libgutenberg package.
The cron-rebuild-files.sh script runs as a cron job, rebuilding 2100 books per day, so as to rebuild every book roughly once a month.
ebookconverter talks to the gutenberg database to build a list of ebookmaker jobs. These jobs require some metadata about the book, and a target file to process.
ebookconverter expects source files to be in numbered directories in a 'files' directory. The location of the files directory is given by the FILESDIR config parameter.
Config parameters should be set in a file at /etc/ebookconverter.conf or ~/.ebookconverter
ebookconverter has been tested on Python 3.6.7. It's not expected to run on python 2.7
pipenv install ebookconverter
The following directories should exist: - $PRIVATE/logs - $PRIVATE/logs/json - $PRIVATE/logs/json/backup - $PRIVATE/logs/notifications - $PRIVATE/logs/dopush - $PRIVATE/logs/dopush/backup
you can run these commands either by first entering a pipenv shell
or on a single line using pipenv run <command> <args>
Rebuild one or more books
ebookconverter --range=<start>-<finish> --build=all
ebookconverter --range=<booknumber> --build=all
ebookconverter --range=<booknumber> --build=all --validate
Reload metadata from a workflow json file (use with care, it will overwrite any metadata in the DB)
reload_workflow <booknumber>
Regenerate the csv file
make_csv
Look for any ebooks with changed files in the last X days and then check if any of the previously known files of that ebook have been deleted.
autodelete