Test cases for LCC sorting and parsing

[bmschmidt]

I've implemented some rules for MARC record sorting, but it would be good to have some test cases.

Things like

E1000 .A45 > E 101 .B34,

and so forth. The big questions come up with the arbitrary mixing of Cutter numbers, dates, and decimals.

For example, my reading of the use of table PR4a suggests that a translation of a Shakespeare play may have two sucessive Cutter numbers in a way that breaks the parser. But I'm not immediately sure.