To get started, you should look at the file "log", which shows two
runs.  One is a run on the problem in the file "testProb.pl" and the
other is a run on all the problems in "ten.pl".

At the top of the log file is the banner printed out by my version of
SWI-Prolog.  The lines beginning with "?-" show my input.  My first
input is "[load].", which causes all of the program files to be
loaded.  My next input is "runAllProblems(testProb, _).", which runs
iterative deepening search for the eight-puzzle problem whose initial
state is in the file "testProb.pl".

runAllProblems prints out:

1. The initial state, which has the format where the last digit is the
contents of the last location in the puzzle, and the first digit is
either the contents of the first or the second location depending on
whether there are nine or eight digits in the state respectively
(leading zeroes are not printed so if the blank is in the first
location, the state will print out as 8 digits.

2. After the state is printed out, the number of nodes generated and
expanded for the next to last level are printed out.  These are the
numbers that are important for us.

3. After that is a note of the last level being explored (where the
optimal solution is) and the number of nodes generated and expanded on
that level, which only counts nodes up to the point of discovering the
first optimal solution path.

4. Then a summary of the number of problems solved, the total number
of nodes generated over all the problems, and the average number of
nodes generated per problem.

Items 1 -3 are printed out for each problem, and item 4 is only
printed out at the end.

To change the search to go from having no duplicate state checking at
all, to checking for state loops, you will need to add code to
idsearch to detect and deal with state loops on the path.