Add Corcoran patterns to regex for build_etd_pdf_hash task

[kerchner]

Whereas the main PDF files for GW ETDs have a pattern like: ..._gwu_0075A_12345.pdf

Corcoran PDFs have another pattern: ..._corcoran_0501M_12345.pdf

Expand the regex to account for these. A sufficient regex should be: pattern = /\w*_(?:gwu|corcoran)_(?:0075|0501)[A-Z]_\d+\.pdf/