From mboxrd@z Thu Jan  1 00:00:00 1970
X-Spam-Checker-Version: SpamAssassin 3.4.4 (2020-01-24) on polar.synack.me
X-Spam-Level: 
X-Spam-Status: No, score=-1.9 required=5.0 tests=BAYES_00,FREEMAIL_FROM
	autolearn=unavailable autolearn_force=no version=3.4.4
X-Received: by 10.36.0.21 with SMTP id 21mr29689697ita.13.1514751287091;
        Sun, 31 Dec 2017 12:14:47 -0800 (PST)
X-Received: by 10.157.90.22 with SMTP id v22mr1703489oth.12.1514751286948;
 Sun, 31 Dec 2017 12:14:46 -0800 (PST)
Path: 
 eternal-september.org!reader01.eternal-september.org!reader02.eternal-september.org!feeder.eternal-september.org!border1.nntp.ams1.giganews.com!nntp.giganews.com!peer03.ams1!peer.ams1.xlned.com!news.xlned.com!peer01.am4!peer.am4.highwinds-media.com!peer01.iad!feed-me.highwinds-media.com!news.highwinds-media.com!g80no4503606itg.0!news-out.google.com!b73ni17387ita.0!nntp.google.com!g80no4503602itg.0!postnews.google.com!glegroupsg2000goo.googlegroups.com!not-for-mail
Newsgroups: comp.lang.ada
Date: Sun, 31 Dec 2017 12:14:46 -0800 (PST)
In-Reply-To: <1c10eec9-040c-4d50-a557-11325883529a@googlegroups.com>
Complaints-To: groups-abuse@google.com
Injection-Info: glegroupsg2000goo.googlegroups.com;
 posting-host=2601:191:8303:2100:71ca:1cb4:da97:fead;
 posting-account=fdRd8woAAADTIlxCu9FgvDrUK4wPzvy3
NNTP-Posting-Host: 2601:191:8303:2100:71ca:1cb4:da97:fead
References: <1c10eec9-040c-4d50-a557-11325883529a@googlegroups.com>
User-Agent: G2/1.0
MIME-Version: 1.0
Message-ID: <79612a29-9183-41dc-b599-ba20e34cd280@googlegroups.com>
Subject: 
 =?UTF-8?B?UmU6IEFycmF5cywgc2xpY2VzLCBjYXNlLCBhbmQg4oCYaW7igJkgc3RyYXRlZ2llcw==?=
From: Robert Eachus <rieachus@comcast.net>
Injection-Date: Sun, 31 Dec 2017 20:14:47 +0000
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
X-Received-Body-CRC: 3342331512
X-Received-Bytes: 4446
Xref: reader02.eternal-september.org comp.lang.ada:49713
Date: 2017-12-31T12:14:46-08:00
List-Id: <comp.lang.ada>

On Friday, December 29, 2017 at 7:51:49 PM UTC-5, Mace Ayres wrote:
> I have a two dimensional array, gird, 1..9, 1..9 index of integer and arr=
ay of a record type named cell and array is named grid...=20
> I need to check, in addition to whether the proposed ingteger value is al=
ready in any cells in that row, or column,  whether the proposed value alre=
ady
> exists in any cell.value field for any cell in triad. The in parameters o=
f r,c can reveal what triad is in question, but it looks like some fat code=
 to to
> traverse the abstract 3x3 triads.
>=20
> I know I could create some triad types of arrays, but I am wondering if I=
 can slice the array grid into 9 collections and then check
> loop.. if numb (proposed number) already exists in the triad [this would =
be determined by row, column in parameters) OR also, the cell/record has a =
field with its triad,
> a constant property.

I would do this by putting the Triad number in each cell, and having a set =
type:

   type Sudoku_Set is array(Integer 1..9) of Boolean;
   Reset: constant Sudoku_Set :=3D (others =3D> False);
   Set_Array: array(Integer 1..9) of Sudoku_Set;
   Row_Array, Col_Array, Triad_Array: Set_Array :=3D (others =3D> Reset);

Now your tests for whether a number is already used is some ors:

   if not (Row_Array(Row) or Col_Array(Column) or
      Triad_Array(Cell(Row, Column).Triad)(Candidate) then...

Oring the three arrays together then chosing a candidate should be faster t=
han indexing them separately.  I didn't attempt to compile the above so it =
probably includes a syntax error or two.  The reason I didn't is that once =
you have the Ah Ha! moment of oring the sets together, you are likely to ch=
ange your tree search to do the oring into a local variable, then do someth=
ing like:

    function Recur(Row, Column: in Integer) is
      Local_Set: Sudoku_Set :=3D Row_Array(Row) or Col_Array(Column) or
      Triad_Array(Cell(Row, Column).Triad;
    begin
      if Cell(Row, Column) =3D 0
      then -- set as part of conditions.
        if Row < 9
        then return Recur(Row+1, Column);
        elsif Column < 9
        then return Recur(1, Column+1);
        else return True;
      end if;

      for I in Integer range 1..9 loop
        if not Local_Set(I)
        then
          Grid(Row_Column).Value :=3D I;
          Row_Array(Row)(I) :=3D True;
          Col_Array(Column)(I) :=3D True;
        end if;    =20
        if Row < 9
        then return Recur(Row+1, Column);
        elsif Column < 9
        then return Recur(1, Column+1);
        else return True;
        end if;=20
      end loop;
    end Recur;

This does a brute force search, but it should be fast enough.  You can add =
bells and whistles like filling in a row column or triad cell when eight nu=
mbers have been used.  I don't think that would make things any faster.  Us=
ing an access value instead of an index for triads might make it faster, bu=
t that's the type of optimization best left until last.