The New Sudoku Players' Forum

[denis_berthier]

.
One function I forgot to document in the last version of the Manual is p-different-numbers-out-of-n ?p ?n

This allows e.g. to randomly extract puzzles from a collection:
(p-different-numbers-out-of-n ?p ?n) will produce a list of p different number within [1 n]

Then use: (extract-lines-in-list ?file-in ?file-len ?list ?file-out) , with
?file-in the original file of puzzles
?file-len the length of the previous file
?list the result if the previous function
?file-out the list of extracted puzzles

Combined in one function:
(extract-lines-in-list ?file-in ?file-len (p-different-numbers-out-of-n ?p ?file-len) ?file-out)

If you want to keep the order of the original file, use function p-different-ordered-numbers-out-of-n instead of p-different-numbers-out-of-n.
.

[denis_berthier]

.
Following my result here: http://forum.enjoysudoku.com/ur1-1-again-t37404-152.html, I've made minor changes in SudoRules priorities: they now allowi to raise the priorities of deadly patterns as high as before Singles (i.e. just at the start as the theorem requires in order to be able to apply with certainty), or immediately after Singles. However, the default value remains after S3 (Subsets of size 3), which remains IMO a good compromise between not missing a DP and not having too many guardians in the initial ORk-relations.
.

[denis_berthier]

.
Only minor changes this morning, to prepare the publication of the SudoRules EXPANSION kit.

I've also added the possibility of having external programs and I've included gsf's solver (in the XTERNS directory) as the first of them (It is used by the EXPANSION kit).
The way to use gsf is described in [HCCS2]. As of now, use it as an independent software (after compiling it as detailed in the "how to use" file).
The way to use it from within SudoRules, in particular in the EXPANSION kit, will appear in [HCCS3].

For those who plan to try the EXPANSION kit, if you work on Windows, you will also need to have some equivalent of the Unix "sed" command for deleting the first line of a file.
(the equivalent of: sed '1d' file-in.txt > file-out.txt)
According to ChatGPT, there are several ways of doing it, depending on the extra tools you have on Windows. It seems the simplest way with no extra tool is: more +1 file-in.txt > file-out.txt. Contact me if you have any idea about this.
.