|laziness, impatience, and hubris|
Re^3: Sorting by dependenciesby LanX (Canon)
|on Apr 27, 2013 at 21:33 UTC||Need Help??|
> Is it possible to combine different types of rules in this algorithm?
Well not A > B is equivalent to B >= A.
But the B == A part - i.e. B and A belong to the same phase - must be ignored, otherwise one risks that A comes after B in the unsorted output of a phase.
So translating to B > A should be safe, w/o loss of solutions.
Just modify the parser to translate different rules accordingly.
( addicted to the Perl Programming Language)