Select Case with AND ?

[Robert E. Carneal]

I am aware that I can do this:

I am going to be a plain about this for clarity.
(Assume Grade = 8)

I know I can do this:
Select Case Grade

Case 2, 4, 6, 8

Res$="EVEN"

Case 1, 3, 5, 7

Res$="ODD"

End select
Print Res$
End

That is, when a CASE clause contains multiple tests separated by commas, the logical OR is performed. Question: Can I make a SELECT CASE statement using the logical AND?? I.e., can I do this:?
Assume X = 75

Select Case X

Case >70, AND <100

Res$="PASS! Congratulations!!

Case >0, AND < 71

Res$="FLUNKO! Go to teacher for punishment!"

Case < 1

Res$="Report to office. You are being sent back a grade."

End Select
Print Res$

Is that possible?

Thank you.

Robert

[José Roca]

No with that syntax, but you can use:

   Select Case X

    Case 71 TO 99

        Res$="PASS! Congratulations!!

    Case 1 TO 70

        Res$="FLUNKO! Go to teacher for punishment!"

    Case < 1

        Res$="Report to office. You are being sent back a grade."

   End Select

[Bern Ertl]

If X / Grade are LONG variables, you should use SELECT CASE AS LONG ...

[Michael Mattias]

FUNCTION PBMAIN () AS LONG

  LOCAL I AS LONG
      FOR I = -51& TO 150 STEP 50
          CALL TestSelect (I)
      NEXT
      EXIT FUNCTION

END FUNCTION

FUNCTION TestSelect(Z AS LONG ) AS LONG
    
    
    SELECT CASE -1&    ' true
         CASE  (Z>0) AND (Z<70)
             MSGBOX USING$ ("# between 1 and 69", Z)
         CASE (Z>=70)
             MSGBOX USING$ ("# >=70", Z)
         CASE (Z=-1&)
             MSGBOX USING$("#  = minus one", Z)
         CASE ELSE
              MSGBOX FORMAT$(Z, "#;-#")  & " is else"
              
    END SELECT
END FUNCTION

(Oops, that's windows code. Well, it works with the Windows compilers. Give it a shot with the DOS compiler)

[Scott Slater]

Select Case Long is a "Windows" thing.

[Michael Mattias]

Guess that shows you how often Bern and I use the MS-DOS compiler, huh?

Best guess for me is not for about five years, but no new development for ten.

[Bern Ertl]

Oops... Sorry about that.

[Manuel Valdes]

Though Jose Roca's answer went straight to the point, I would like to recall that SELECT CASE is more aimed to improve the code clarity and efficiency. The good old IF, ELSEIF,......, ELSEIF, ELSE, END IF can do the same, and also handle a set of far more complex logical statements.

[Erich Schulman]

For your SELECT CASE X, you won't really lose anything much in performance or clarity if you just used a series of IF / END IF blocks. You should be able to see all of it on the screen at once. But if you anticipate adding more criteria and/or action per criterion, then the SELECT CASE would become desirable.

You can do your even/odd test with a single IF. Check the MOD 2 for true (odd) or false (even). But the SELECT CASE does convey more clearly what you intend to accomplish.

[Scott Slater]

You can do your even/odd test with a single IF. Check the MOD 2 for true (odd) or false (even). But the SELECT CASE does convey more clearly what you intend to accomplish.

You could AND it with 1 to make it really fast...

DIM X AS INTEGER

If (X AND 1) Then
   Print "ODD"
Else
   Print "EVEN"
End If

That simply checks BIT 1 which when ON always causes the value to be odd.

[Eric Pearson]

Scott --

Not if the number is negative. For example an INTEGER with "all bits on" is -32,768.

-- Eric

[Paul Dixon]

Eric,
did you post that before you had your first cup of coffee of the day?

Paul.

[Eric Pearson]

I don't drink coffee, but Yes, I posted that before my first Coca-Cola of the day. It seemed so logical at the time.

Scott, please disregard my response.

[Scott Slater]

no problem... you must have been up too late the night before writing code.

[Michael Mattias]

IsOdd = ISTRUE (X MOD 2)

[Scott Slater]

IsOdd = ISTRUE (X MOD 2)

IsOdd = (X AND 1)

Still faster because it translates better to assembly and does not have to do a divide.

[John Gleason]

!test x, 1

[Scott Slater]

!test x, 1

Is that faster than;

 
! and x, 1

?

[John Gleason]

Is that faster than;

 
! and x, 1

?

It is faster, but !test doesn't save the result in x like !and does, it just sets the flags and you would almost immediately need to !jz short label or similar based on the result.

[Michael Mattias]

I hate to interrupt this lovely (that's with an "oh" not an "eye") conversation, but did anyone using PB-DOS actually try the "SELECT CASE -1%" (changed to integer) construct to handle the 'compound conditions' which was, after all, the original question?