HHC 2014 Programming Contest

[Werner]

JHM's routine should work if you replace the IFTs by their IF THEN END counterparts:
(untested)

Code:

\<<
  \<< IF DUP2 MOD NOT THEN 
        IF DUP2 > THEN DROP2 -1 2 END 
        END
  \>>
  \-> p
  \<< 1 MAX 2 p EVAL 1 + p EVAL 2 + p EVAL 2 +
      WHILE DUP2 SQ \>= REPEAT
          p EVAL 4 + p EVAL 2 + p EVAL 4 + p EVAL
      2 + p EVAL 4 + p EVAL 6 + p EVAL 2 + p EVAL 6 +
      END DROP 1 >
   \>>
\>>

Remark that I changed the ultimate test to 1 >
This way it will work for 0 and 1, too. I edited my previous post as well.

Or else, my own simple concoction.

Code:

\<<
  1 MAX 2
  IF DUP2 MOD THEN 1 +
  IF DUP2 MOD THEN 2 +
  IF DUP2 MOD THEN 2 + 
  ELSE SWAP END
  ELSE SWAP END
  ELSE SWAP END
  WHILE DUP2 SQ \>= REPEAT
    IF DUP2 MOD THEN 4 +
    IF DUP2 MOD THEN 2 +
    IF DUP2 MOD THEN 4 +
    IF DUP2 MOD THEN 2 +
    IF DUP2 MOD THEN 4 +
    IF DUP2 MOD THEN 6 +
    IF DUP2 MOD THEN 2 +
    IF DUP2 MOD THEN 6 +
    ELSE SWAP END
    ELSE SWAP END
    ELSE SWAP END
    ELSE SWAP END
    ELSE SWAP END
    ELSE SWAP END
    ELSE SWAP END
    ELSE SWAP END
  END
  \>=
\>>

Yet, to speed up the PDP routine, consider the following:
If yyyymm01 is prime and the day before isn't, then we have performed a lengthy primality testing for nothing.
What we should have is a routine BOTHPRIME? that tests two numbers for divisibility by ever increasing factors at the same time - If both are prime, that will gain us nothing, but if one or both are not, then we will gain a lot of time. I think.
How many times does it happen that yyyymm01 is prime and the day before isn't?
To be continued..

Werner