This picks up from an earlier article on trying to quantify the comparison of two stored procedures.
For a more complete explanation, please go to my blog entry here.
Here's a sample of that blog entry that summarizes what this script does:
"I devised an algorithm that would compare blocks of lines between them, starting with the largest possible sequence to compare (the size of this would be the number of lines of the smaller spd), and then compare smaller and smaller sequences until the comparison size was just one line (which is what was compared in Comparing Stored Procedures, Part 1). The trick to this approach is that if a match was found between to large sequences of lines, the subsequences in them would never be compared. Although this certainly helps with performance, the real reason behind this is that the fact that large blocks of code matched is very meaningful. Take this example: the string 'ABCD' matches 'ABDC' more than it matches 'DCBA'. How much "more" it matches is the subject of a later posting, where we establish our scoring algorithm."
Written by Jesse McLain
jesse@jessemclain.com
www.jessemclain.com
www.jessesql.blogspot.com
