Is this a bug?

[JRS]

In most BASIC languages INT() is a function to return a integer number. SB has INT and FIX to give you more granularity on how it rounds. (up / down)

Could INT be a cast and achive the same result?

Code: OxygenBasic

1. float f=4.25
2. sys a
3. a=(int) f
4. print a
5.  

I'm spoiled as SB is a typleless scripting language and you don't have to DECLARE anything including arrays.

[JRS]

Here is a remake with INT as a cast.

Code: OxygenBasic

1.     include "$/inc/console.inc"
2.      
3.     dim a,b as double
4.     global r as int
5.     sub _print()
6.     ====================================
7.             if r=1 then
8.                     printl "YES is an exact root"
9.             else
10.                     printl "NO it is not an exact root"
11.             end if
12.             printl
13.             r=0
14.     end sub
15.      
16.      
17.     '====== MAIN ===============
18.    do
19.      
20.     r=0
21.     printl "input a number: "
22.     'a=input()
23.    a = ltrim rtrim Input ' per togliere caratteri da input tipo CR
24.    if a=0 then exit do
25. cls
26.     b = sqrt(a)
27.     print "a       = " a cr
28.     print "SQRT(a) = " b cr
29.      
30.     printl "Text if the result = (INT) result"
31.     if b = (int) b then
32.              r=1
33.     end if
34.     _print
35.      
36.     printl "Text if the FRAC(result) has a fractional component"
37.     if frac(b)=0 then
38.             r=1
39.     end if
40.     _print
41.      
42.     printl "Text if result = TRUNC(result)"
43.     if b=trunc(b) then
44.             r=1
45.     end if
46.     _print
47.      
48.     printl "Test if Result - (INT) result = 0"
49.     if b - (int) b = 0 then
50.             r=1
51.     end if
52.     _print
53.      
54.     printl "Test if (INT) Result ^ 2 = number in input"
55.     if (int) b * (int) b = a then
56.             r=1
57.     end if
58.     _print
59.      
60.     loop
61.  

a       = 25
SQRT(a) = 5

Text if the result = (INT) result
NO it is not an exact root

Text if the FRAC(result) has a fractional component
YES is an exact root

Text if result = TRUNC(result)
YES is an exact root

Test if Result - (INT) result = 0
NO it is not an exact root

Test if (INT) Result ^ 2 = number in input
NO it is not an exact root

---------------------------------------------------

a       = 3
SQRT(a) = 1.7320508075688772

Text if the result = (INT) result
NO it is not an exact root

Text if the FRAC(result) has a fractional component
NO it is not an exact root

Text if result = TRUNC(result)
NO it is not an exact root

Test if Result - (INT) result = 0
NO it is not an exact root

Test if (INT) Result ^ 2 = number in input
NO it is not an exact root

[Charles Pegge]

Hi John,

Yes you can cast (int) but the effect is to treat the floating point binary as integer binary. They have different encodings

But you can use convert

Code: [Select]

uses console
double g=5/3
print "g ="     tab g cr
print "cast"    tab str((int) g) cr
print "cast"    tab str(cast int g) cr
print "convert" tab str(convert int g) cr
wait


[JRS]

g =     1.6666666666666667
cast    -1431655765
cast    -1431655765
convert 2


Thanks for clearing up what a cast int is. The convert is a handy keyword.

[JRS]

Here is a remake using convert int.

Code: OxygenBasic

1.     include "$/inc/console.inc"
2.      
3.     dim a,b as double
4.     global r as int
5.     sub _print()
6.     ====================================
7.             if r=1 then
8.                     printl "YES is an exact root"
9.             else
10.                     printl "NO it is not an exact root"
11.             end if
12.             printl
13.             r=0
14.     end sub
15.      
16.      
17.     '====== MAIN ===============
18.    do
19.      
20.     r=0
21.     printl "input a number: "
22.     'a=input()
23.    a = ltrim rtrim Input ' per togliere caratteri da input tipo CR
24.    if a=0 then exit do
25. cls
26.     b = sqrt(a)
27.     print "a       = " a cr
28.     print "SQRT(a) = " b cr
29.      
30.     printl "Text if the result = convert int result"
31.     if b = convert int b then
32.              r=1
33.     end if
34.     _print
35.      
36.     printl "Text if the FRAC(result) has a fractional component"
37.     if frac(b)=0 then
38.             r=1
39.     end if
40.     _print
41.      
42.     printl "Text if result = TRUNC(result)"
43.     if b=trunc(b) then
44.             r=1
45.     end if
46.     _print
47.      
48.     printl "Test if Result - convert int result = 0"
49.     if b - convert int b = 0 then
50.             r=1
51.     end if
52.     _print
53.      
54.     printl "Test if convert int Result ^ 2 = number in input"
55.     if convert int b * convert int b = a then
56.             r=1
57.     end if
58.     _print
59.      
60.     loop
61.  

a       = 25
SQRT(a) = 5

Text if the result = convert int result
NO it is not an exact root

Text if the FRAC(result) has a fractional component
YES is an exact root

Text if result = TRUNC(result)
YES is an exact root

Test if Result - convert int result = 0
YES is an exact root

Test if convert int Result ^ 2 = number in input
YES is an exact root

[Nicola]

Hi Charles.
However apparently INT () works correctly. In fact, by definition, INT returns the nearest integer. For example INT (3.49) should equal 3, while INT (3.51) should equal 4.
The test I made proves it (see the attached txt file).

Code: OxygenBasic

1. use console
2.  
3. cls
4.  
5. double i
6. for i=1 to 15
7.   printl i tab sqrt(i) tab int(sqrt(i))
8. next
9. waitkey
10.  

I did research and found that CAST and CONVERT are used in SQL ...
Is there a place in O2 where they are explained?

[Charles Pegge]

Hi Nicola,

The problems arise when putting int() into a conditional expression. I will try to improve the situation but I strongly recommend using frac or round since these enable float-float comparisions directly on the Floating point processor (FPU).

One possible solution is for o2 to intercept int() and turn it into round()

frac translates to one assembler instruction frac, and round translates to frndint. This is much faster than integer-float conversions.

The cast and convert keywords are briefly described in the manual.

[Nicola]

Thanks Charlie, also for your patience.
At this point I like to see the results I get with the set of functions you have indicated to me.
For printing convenience I put the _str function. I'm sure you already have some function in o2 that does the same thing ...

Code: OxygenBasic

1. use console
2.  
3. function _str(string s,int d=20) as string
4.         function = s+ space(d-len(s))
5. end function
6.  
7. cls
8.  
9. double N,r
10. print "N" tab _str("sqrt(N)",22) _str("frac(sqrt(N))") _str("trunc(sqrt(N))") _str("ceil(sqrt(N))") _str("floor(sqrt(N))") cr
11. for N=1 to 16
12.   r=sqrt(N)
13.   print N tab _str(r,22) _str(frac(r)) _str(trunc(r)) _str(ceil(r)) _str(floor(r)) cr
14. next
15. waitkey
16.  

This is the result:

N       sqrt(N)               frac(sqrt(N))       trunc(sqrt(N))      ceil(sqrt(N))       floor(sqrt(N))
1       1                     0                   1                   1                   1
2       1.4142135623730951    0.4142135623730951  1                   2                   1
3       1.7320508075688772    0.7320508075688772  1                   2                   1
4       2                     0                   2                   2                   2
5       2.2360679774997898    0.2360679774997898  2                   3                   2
6       2.4494897427831779    0.4494897427831779  2                   3                   2
7       2.6457513110645907    0.6457513110645907  2                   3                   2
8       2.8284271247461903    0.8284271247461903  2                   3                   2
9       3                     0                   3                   3                   3
10      3.1622776601683795    0.1622776601683795  3                   4                   3
11      3.3166247903553998    0.3166247903553998  3                   4                   3
12      3.4641016151377544    0.4641016151377544  3                   4                   3
13      3.6055512754639891    0.6055512754639891  3                   4                   3
14      3.7416573867739413    0.7416573867739413  3                   4                   3
15      3.872983346207417     0.872983346207417   3                   4                   3
16      4                     0                   4                   4                   4


Hello

[Charles Pegge]

You can limit the decimal places by using a second param for str():

example:
str(1/3, 5)
0.33333

Using 0 will give you rounded integer strings.

[Nicola]

Great!
But what if you need some extra space?

[JRS]

Does O2 have a PRINT FORMAT like function?

[Charles Pegge]

We can use sprintf from msvcrt.dll.

But here is a function to support decimal-point alignment in a column, taking care of left and right spacing.

Code: [Select]

function dp_str(string num="", int wid=20, dpp=10) as string
===========================================================
  'num  number string
  'wid  column width
  'dpp  decimal point position in column
  '
  string buf 'column buffer
  int ln     'length of num
  int dp     'offset of num decimal point
  '
  buf=space(wid) ' column string
  ln=len(num)
  dp=instr(num,".")
  if dp=0 'integers
    dp=ln+1
  endif
  '
  int n=dpp-dp 'left spaces
  'check if too long
  if n<0 then n=0 'remove left space
  if ln>wid
    mid buf,1,"!!"+num 'error
  else
    mid(buf,1+n,num) 'copy number into position
  endif
  return buf
end function

'TESTS
uses console
print dp_str(pi(),30)      "|" cr
print dp_str(pi(),30,4)    "|" cr
print dp_str(str(pi(),5))  "|" cr
print dp_str(12345678)     "|" cr
print dp_str(123)          "|" cr
print dp_str(-123)         "|" cr
print dp_str(-123.45)      "|" cr
wait


[JRS]

Charles,

It might be worth translating ScriptBasic's PRINT FORMAT function from C to O2. A function I couldn't live without.

FORMAT

Not sure how O2 would deal with a varidic like argument passing.

[Charles Pegge]

Hi John,

I see that SB Format is based on sprintf.

This is a useful reference:

http://www.cplusplus.com/reference/cstdio/printf/

Variadics are supported for all msvcrt functions that use them.

Example:

Code: [Select]

uses corewin
uses console
char buf[100]
double d=pi()
char buf[100]
double d=pi()

'sprintf_s(buf,100,"floats: %4.2f %+.0e %E " cr, d, d, d)
'sprintf(buf, "floats: %4.2f %+.0e %E " cr, d, d, d)

   sprintf (buf, "Characters: %c %c ", asc("a"), 65)
   print buf cr
   sprintf (buf, "Decimals: %d %ld", 1977, 650000L)
   print buf cr
   sprintf (buf, "Preceding with blanks: %10d ", 1977)
   print buf cr
   sprintf (buf, "Preceding with zeros: %010d ", 1977)
   print buf cr
   sprintf (buf, "Some different radices: %d %x %o %#x %#o ", 100, 100, 100, 100, 100)
   print buf cr
   sprintf (buf, "floats: %4.2f %+.0e %E", double 3.1416, double 3.1416, double 3.1416)
   print buf cr
   sprintf (buf, "Width trick: %*d", 5, 10)
   print buf cr
   sprintf (buf, "%s", "A string")
   print buf cr
   print cr
wait


[JRS]

That's close enough for me. I do use the custom SB number formatting string a lot as well. It looks like it wouldn't be too hard to create a O2 FORMAT function using sprintf and a bit of extra code to handle the numeric formatting.

Code: Script BASIC

1. PRINT FORMAT("Line Number: %~0000~ Square root of 5 is %.3f\n", 25, SQR(5))

C:\ScriptBASIC\examples>sbc format.sb
Line Number: 0025 Square root of 5 is 2.236

C:\ScriptBASIC\examples>



The FORMAT function is in string.c SB source.