If you don’t already read (or browse the table of contents) The Code4Lib Journal, I suggest you start now. Occasionally, there is content that is very relevant to eresources/serials/acquisitions.
The most recent issue contains an article written by Kristina M. Spurgin, the E-Resources Cataloger at University of North Carolina at Chapel Hill. She created a Ruby script to check all the links and look for key words in the HTML or text that indicated if there was a problem with access. Of course, each platform has different ways of indicating whether or not a user from a particular IP range has access to an article/chapter, so each one has it’s own configuration, built into the script. She plans to eventually move them into their own files that can be selectively used by others as appropriate to their collections.
This is still a work in progress, and she notes that it’s not a perfect solution for several reasons, including random errors caused by one of the scripting libraries. That being said, I’m excited to see a potential open source solution to a problem we all have. Automated access checking requires the program to be a smart as an experienced eresource librarian, so it makes sense that a smart, experienced eresource librarian would be writing it.