Algorithm for tracking issues should use difference in position of line

Current algorithm depends only on size of block (SONAR-3072), whereas would be better also take into account difference in position of lines.
As an example:

line 1: A
line 2: B // violation
line 3: B // violation

line 1: C
line 2: A
line 3: B // violation
line 4: B // violation

Current algorithm can create association between version 1 line 2 and version 2 line 4 or version 1 line 2 and version 2 line 3, depending on order of violations in lists, because size of block is the same (3 unmodified lines).