This was a project I worked on with two professors from Texas A&M University who were editors for the Variorum Edition of the Poety of John Donne (see http://www.iupress.indiana.edu/index.php?cPath=1037_3130_3171 and http://donnevariorum.tamu.edu). They wanted to determine if there were sets or sequences of poems that stayed together through the early years of distribution of Donne's poems, when they were primarily distributed by hand-copied manuscripts.
I wrote some simple python scripts to parse the sequences into memory, then look for any groups that stayed together over time. It ended up that very few sets, and even fewer sequences stayed consistent over time, something that had never been determined before.