Print custom alphabet

GolfScript 48 75 73 70 67 66 63 57 53

(91,65>.+.{32+}%+.@?>(;25<''+.,1>*\['''.']?[' 'n 0]=*

Online demos:

  • Test case 1 ('c')

  • Test case 2 ('H.')

  • Test case 3 (invalid input)

Update:

Now the last rule is also implemented. Thanks to Ventero for pointing out the issue.

Update:

I rewrote the code from scratch and found new ways to shorten it even further.

History of modifications:

.,3<\(:x;:§['''.']?)and{91,65>.+.{32+}%+.x?).{>25<''+{§n' 'if}%}{;;}if}''if
.,3<\(:x;:§['''.']?)*{91,65>.+.{32+}%+.x?).{>25<''+{§n' 'if}%}{;;}if}''if
.,3<\(:x;:§['''.']?)*{91,65>.+.{32+}%+.x?).!!@@>25<''+{§n' 'if}%*}''if
.,3<\(:x;:§['''.']?)*!!91,65>.+.{32+}%+.x?).!!@@>25<''+{§n' 'if}%**
.,3<\(:x;:§['''.']?)*91,65>.+.{32+}%+.x?).@@>25<''+{§n' 'if}%@@*!!* 
.(@,3<@:§['''.']?)*91,65>.+.{32+}%+@1$?).@@>25<''+{§n' 'if}%@@*!!*
.(@,3<@:§['''.']?)*91,65>.+.{32+}%+@1$?):x>25<''+{§n' 'if}%\x*!!*
.(@,3<@:§['''.']?)*91,65>.+.{32+}%+@1$?):x>25<''+§n' 'if*\x*!!*
(\:§['''.']?)91,65>.+.{32+}%+@1$?):x>25<''+§n' 'if*\x*!!*
(91,65>.+.{32+}%+.@?>(;25<''+.,1>*\['''.']?[' 'n 0]=*

C, 135 129 128 chars

Damn, so many different magic numbers, but no way to get rid of them.

Has to be run with the input as program parameter. Now follows the "wrong input" requirement.

c;main(a,b)char**b;{if(a==2&&isalpha(**++b)&&!(c=1[*b])||c==46&&!2[*b])for(;++a<28;)printf("%c%c",**b-=**b+6&31?-1:25,c?10:32);}

Explanation:

c;                   // Variable that will be used later
main(a,b)char**b;{   // There's one parameter => a = 2, b[1] = the parameter
                     // Wrong input checks: We want...
  if(
     a==2 &&         // 1 parameter and ...
     isalpha(**++b)  // lower- or uppercase letter as parameter,
                     // increase b so we can access it better
     &&              // and ...
     !(c=1[*b]) ||   //   either no further character,
                     //     save the character in c, or...
     (c==46&&!2[*b]) //   a dot as next character and no further characters
    )                // if check succeeded, execute the for loop, else do nothing
  for(;++a<28;)      // This will loop 26 times (2..27)
    printf("%c%c",   // Print two characters
                     // First character to print:
      **b            // We'll print the first character of the parameter,
       -=            // but decrement it before printing
       **b+6&31?     // if the last five bits (31 = 11111b) are not 26 (6 == -26 mod 32)
        -1           //   decrement it by -1 (aka increment it)
        :25,         //   else (char=z/Z) decrement by 25, so we start over at a/A
                     // Second character to print:
      c?             // c is either ASCII 0 or a dot (ASCII 46)
       10            //   dot     -> print a newline
       :32);         //   ASCII 0 -> print a space (ASCII 32)
}

The **b+6&31 part uses the fact that the ASCII codes for lowercase/uppercase character are the same if only looking at the last 5 bits and the remaining 5 bits are in range 1..26.

Version without "wrong input" requirement (82 chars):

main(a,b)char**b;{for(b++;++a<28;)printf("%c%c",**b-=**b+6&31?-1:25,1[*b]?10:32);}

Ruby, 72 71 61 characters

gets;25.times{$><<$_=$_.succ[0]+=$1?$/:' '}if~/^[a-z](\.)?$/i

This ruby version uses a regular expression to verify the input. Fortunately, the Ruby string method succ does most of the work for us (including the wrap-around).

Edit: 61 characters with the help of chron and Ventero.

Tags:

String

Code Golf

Recent Posts