hathitables demonstrates how HathiTrust collections can be shared as Linked Data friendly CSV a.k.a. CSV on the Web.
More context for this work can be found at MITH's page for the HathiTrust Workset Creation for Scholarly Analysis project.
If you want to get CSV for a given HathiTrust collection you can use hathitables on the command line:
% hathitables.py 1761339300 > 1761339300.csv
And here's how to write the metadata where a CSVW client can find it:
% hathitables.py --metadata 1761339300 > 1761339300.csv-metadata.json
If you want you can also use hathitables programatically from Python:
import hathitables
collection = hathitables.Collection('1761339300')
collection.write_csv(open("1761339300.csv", "w"))
collection.write_metadata(open("1761339300.csv-metadata.json", "w"))