«

On Sep 21, 1:31 pm, Piotr Kobzda <pi...@gazeta.pl> wrote:
> Daniel Pitts wrote:
> >> enum C{C;{String s=" ";for(char
> >> c=64;c++<91System.out.println(c-64+(s+=(c-5)%10).replaceAll(c<91?"\\d":"!",""+c));System.exi t(0);}}
>
> Here is a slightly rewritten variant of it:
>
> enum C{C;{String s=" 0";for(char
> c=64;c++<91;s+=(c-4)%10)System.out.println(c-64+(c<91?s.replaceAll("\\d",""+c):s));System.exit( 0);}}
>
> Count is the same, 133, but the latter appears a bit nicer to me.
>
>
>
> > Nice...
>
> Thanks, and the same to you Daniel! Without your great Java hacks I
> would likely have never had come to my result.
>
> But the contest is still open!
>
> Hope we'll see shorter code soon.
>
> piotr

Perhaps a brute-force approach should be used. Write different java
files and compile them until one compiles, and when executed outputs
the expected output.

On Sep 21, 1:31 pm, Piotr Kobzda <pi...@gazeta.pl> wrote:
> Daniel Pitts wrote:
> >> enum C{C;{String s=" ";for(char
> >> c=64;c++<91System.out.println(c-64+(s+=(c-5)%10).replaceAll(c<91?"\\d":"!",""+c));System.exi t(0);}}
>
> Here is a slightly rewritten variant of it:
>
> enum C{C;{String s=" 0";for(char
> c=64;c++<91;s+=(c-4)%10)System.out.println(c-64+(c<91?s.replaceAll("\\d",""+c):s));System.exit( 0);}}
>
> Count is the same, 133, but the latter appears a bit nicer to me.
>
>
>
> > Nice...
>
> Thanks, and the same to you Daniel! Without your great Java hacks I
> would likely have never had come to my result.
>
> But the contest is still open!
>
> Hope we'll see shorter code soon.
>
> piotr

125 chars:

enum C{C;{String g="%d ",s=g+0;for(int i=0;i++<27;s+=i
%10)System.out.format(i<27?(g+="%2$c")+"\n":s,i,i
+64);System.exit(0);}}

Daniel Pitts wrote:

> 125 chars:
>
> enum C{C;{String g="%d ",s=g+0;for(int i=0;i++<27;s+=i
> %10)System.out.format(i<27?(g+="%2$c")+"\n":s,i,i
> +64);System.exit(0);}}

Super... I was thinking about %c before, but you did it!

Just one char less:

enum C{C;int i;{String g="%d ",s=g+0;for(;i++<27;s+=i%10)
System.out.format(i<27?(g+="%2$c")+"\n":s,i,i+64); System.exit(0);}}




piotr

Piotr Kobzda wrote:
> Daniel Pitts wrote:
>
>> 125 chars:
>>
>> enum C{C;{String g="%d ",s=g+0;for(int i=0;i++<27;s+=i
>> %10)System.out.format(i<27?(g+="%2$c")+"\n":s,i,i
>> +64);System.exit(0);}}
>
> Super... I was thinking about %c before, but you did it!
>
> Just one char less:
>
> enum C{C;int i;{String g="%d ",s=g+0;for(;i++<27;s+=i%10)
> System.out.format(i<27?(g+="%2$c")+"\n":s,i,i+64); System.exit(0);}}
>
>

See? Java does rival Ruby for compactness and obscurity of expression!

--
Lew

On Sep 22, 3:09 am, Piotr Kobzda <pi...@gazeta.pl> wrote:
> Daniel Pitts wrote:
> > 125 chars:
>
> > enum C{C;{String g="%d ",s=g+0;for(int i=0;i++<27;s+=i
> > %10)System.out.format(i<27?(g+="%2$c")+"\n":s,i,i
> > +64);System.exit(0);}}
>
> Super... I was thinking about %c before, but you did it!
>
> Just one char less:
>
> enum C{C;int i;{String g="%d ",s=g+0;for(;i++<27;s+=i%10)
> System.out.format(i<27?(g+="%2$c")+"\n":s,i,i+64); System.exit(0);}}
>
>
>
> piotr

Nice. I think I'm satisfied of the compactness of this program.

Unless you consider passing things in from the command line (could be
cheating, but whatever)
73 characters:

enum C{C;{System.out.format(System.getProperty("o"),
10);System.exit(0);}}

Run it as
java "-Do=1 A%1$c2 BB%1$c3 CCC%1$c4 DDDD%1$c5 EEEEE%1$c6 FFFFFF%1$c7
GGGGGGG%1$c8 HHHHHHHH%1$c9 IIIIIIIII%1$c10 JJJJJJJJJJ%1$c11 KKKKKKKKKKK
%1$c12 LLLLLLLLLLLL%1$c13 MMMMMMMMMMMMM%1$c14 NNNNNNNNNNNNNN%1$c15
OOOOOOOOOOOOOOO%1$c16 PPPPPPPPPPPPPPPP%1$c17 QQQQQQQQQQQQQQQQQ%1$c18
RRRRRRRRRRRRRRRRRR%1$c19 SSSSSSSSSSSSSSSSSSS%1$c20 TTTTTTTTTTTTTTTTTTTT
%1$c21 UUUUUUUUUUUUUUUUUUUUU%1$c22 VVVVVVVVVVVVVVVVVVVVVV%1$c23
WWWWWWWWWWWWWWWWWWWWWWW%1$c24 XXXXXXXXXXXXXXXXXXXXXXXX%1$c25
YYYYYYYYYYYYYYYYYYYYYYYYY%1$c26 ZZZZZZZZZZZZZZZZZZZZZZZZZZ%1$c27
012345678901234567890123456%1$c" C

You could also use the environment variables instead, but setting that
depends on the system. It shaves 4 off though (getProperty -
getenv). Also, depending on your system, you could replace all the
%1$c with an actual linefeed, and then you can change "format" to
"print", and remove the ",10" to save an additional 4.

Although, I really don't consider this type of solution as interesting.

Daniel Pitts wrote:
> Unless you consider passing things in from the command line (could be
> cheating, but whatever)
> 73 characters:
....
> Although, I really don't consider this type of solution as interesting.

Perhaps not to you, but it does represent a basic principle of good OO
programming - externalize your magic values. For one thing, with the text in
a resource or command-line argument, you have now a 73-character i18n version
of the algorithm; the original Ruby version was tied to American/Western
locale only.

Apropos of which, given the multinational character of the United States, and
the prevalence of at least Español as a widespread language there, it may be
that ASCII no longer suffices for American-locale strings anyway.

--
Lew

Daniel Pitts wrote:

> Nice. I think I'm satisfied of the compactness of this program.

Yes, it's compact. But we have a new one...

119 chars:

enum C{C;int i,j;{for(;j<27;j++)System.out.printf(j>0?"%c":
++i+" %c",j<i?i<27?i+64:j%10+48:11+(j=-1));System.exit(0);}}


> Unless you consider passing things in from the command line (could be
> cheating, but whatever)
> 73 characters:
>
> enum C{C;{System.out.format(System.getProperty("o"),
> 10);System.exit(0);}}
>
> Run it as
> java "-Do=1 A%1$c2 BB%1$c3 CCC%1$c4 DDDD%1$c5 EEEEE%1$c6 FFFFFF%1$c7
> GGGGGGG%1$c8 HHHHHHHH%1$c9 IIIIIIIII%1$c10 JJJJJJJJJJ%1$c11 KKKKKKKKKKK
> %1$c12 LLLLLLLLLLLL%1$c13 MMMMMMMMMMMMM%1$c14 NNNNNNNNNNNNNN%1$c15
> OOOOOOOOOOOOOOO%1$c16 PPPPPPPPPPPPPPPP%1$c17 QQQQQQQQQQQQQQQQQ%1$c18
> RRRRRRRRRRRRRRRRRR%1$c19 SSSSSSSSSSSSSSSSSSS%1$c20 TTTTTTTTTTTTTTTTTTTT
> %1$c21 UUUUUUUUUUUUUUUUUUUUU%1$c22 VVVVVVVVVVVVVVVVVVVVVV%1$c23
> WWWWWWWWWWWWWWWWWWWWWWW%1$c24 XXXXXXXXXXXXXXXXXXXXXXXX%1$c25
> YYYYYYYYYYYYYYYYYYYYYYYYY%1$c26 ZZZZZZZZZZZZZZZZZZZZZZZZZZ%1$c27
> 012345678901234567890123456%1$c" C

So what would you say about *0*-length solution?

Run it as:

jrunscript -e "for(i=0;i++<27{print(i+'
');for(j=0;j<i;++j)print(i<27?String.fromCharCode( i+64):j%10);println('');}"

(Requires JDK 1.6+)


> You could also use the environment variables instead, but setting that
> depends on the system. It shaves 4 off though (getProperty -
> getenv). Also, depending on your system, you could replace all the
> %1$c with an actual linefeed [...]

We don't need it since we have %n.

> Although, I really don't consider this type of solution as interesting.

Neither do I.


piotr

Piotr Kobzda wrote:

> 119 chars:
>
> enum C{C;int i,j;{for(;j<27;j++)System.out.printf(j>0?"%c":
> ++i+" %c",j<i?i<27?i+64:j%10+48:11+(j=-1));System.exit(0);}}

118:

enum C{C;int i,j;{for(;j<27;j++)System.out.printf(j>0?"%c":
++i+" %c",j<i?i<27?i+64:j%10+48:9-(j=-1));System.exit(0);}}




piotr

Piotr Kobzda wrote:
> Piotr Kobzda wrote:
>
>> 119 chars:
>>
>> enum C{C;int i,j;{for(;j<27;j++)System.out.printf(j>0?"%c":
>> ++i+" %c",j<i?i<27?i+64:j%10+48:11+(j=-1));System.exit(0);}}
>
>
> 118:
>
> enum C{C;int i,j;{for(;j<27;j++)System.out.printf(j>0?"%c":
> ++i+" %c",j<i?i<27?i+64:j%10+48:9-(j=-1));System.exit(0);}}

This gives one an interesting impression of what it may look like under
the hood when the compiler is doing optimizations ... and why the
resulting assembly/bytecode/whatever is frequently gibberish spaghetti
code nobody on earth could possibly hope to maintain. Not exactly, mind
you, since it optimizes for speed rather than space! But still ...