Yea, what would happen is that the extracted abstract would go into a databse or a separate text file. during normal viewing the abstract would be retrieved, but if the link contained the override then the full article would be retrieved.
Another approach that might work, thou harder to code, would be to attempt to validate the link to the orginal article, if that failed then skip the abstract and get the full version. I might play around a bit.