rcbroughton wrote:
SudokuSolver strictly numbers grids from top left to bottom right (except for Clueless Grids, which are always last).
IMHO this is the best numbering scheme ... as it is the one I also use in TUSP
Having the same numbering scheme in all helpers/solvers would be preferable (at least for the users).
Jean-Christophe & Richard:As far as I can determine neither JSudoku nor SudokuSolver checks if the codes for overlapping nonets are identical in all code strings.
What I do in TUSP is:
For every cell with a mismatch I produce a warning message of the form
"Different values for this cell in the codes for the 9x9 grids X and Y.
In grid X it [has the value A / is empty]. In grid Y it [has the value B / is empty]."
and highlight the cell in an ugly red colour.
When done reading the code I put TUSP in "Forced Clue Mode", i.e. only givens can be entered until the user has entered a given / empty cell in all cells that have a mismatch.
Alternatively the user can clear the grid or paste in another puzzle.
_________________
Quis custodiet ipsos custodes?
Normal: [D Y-m-d, G:i] PM->email: [D, d M Y H:i:s]