Work with .zip files for bookreader when available

For the offline version of Internet Archive we use pre-scaled files (to about 600 px) because both storage and bandwidth are significant constraints, but for Carl Malamud’s project he’s going to have the zip file locally, so in theory (if I can find a way to do it) we could pull the files out the zip and scale in the NodeJS server.

Start by locating the NodeJS server path used by the offline BookReader and reviewing how pre-scaled files are currently selected. Determine whether locally available .zip files can be read and their contents scaled; the work is done when the BookReader can use those local files without relying on pre-scaled copies.

Written by the indexing model from the issue text.