I'm gonna have a go at this one with some more automation. I used this command on this PDF to get a more easily readable format:

pdftotext -layout -nopgbrk Building_List.pdf - | grep -v -E ".*9/6/2006$" > Building_List.txt

--Brian.Beck 19:10, October 22, 2006 (EDT)

Better yet, I just found this top secret Excel file on the construction web site and put it up on Google Spreadsheets. Should be much easier to parse now. --Brian.Beck 23:05, October 22, 2006 (EDT)

I'm running the address adder script right now. It's practically the same as the building codes script except the data set was obtained like so:

import csv
# csv was exported from Google Spreadsheets link above
buildings = csv.reader(open("Case Building Identification.csv", 'rb'), dialect=csv.excel)
# Exclude blank rows and skip the first 7
buildings = [row for row in buildings[8:] if row]
# Grab all rows before the first Appendix
buildings = it.takewhile(lambda row: not row[2].startswith('APPENDIX'), buildings)
# Filter rows with both name and address, do some formatting
addresses = [(row[2].partition('(')[0].strip().replace('Bldg', 'Building'), row[4]) for \
             row in buildings if len(row) > 4 and row[2] and row[4]]
# This only affects 'Strosacker Aud'
addresses = [(name.replace('Aud', 'Auditorium'), add) for (name, add) in addresses]
addresses = dict(addresses)

--Brian.Beck 18:35, October 23, 2006 (EDT)