Greetings, all!
This is my second golf, and I'm having a wonderful time
reading all the unique solutions at work (thanks, Juho!)
It is dissapointing, however, that I am having several
issues (which I believe to be related to running ActivePerl
on Windows 2000) in getting several solutions to execute
(not to mention that my own solution which runs fine on
my machine is rejected upon submitting it to terje2) ANY
help or suggestions that anyone might have in relation to
this would be greatly appreciated.
For example, when running Juho's solution below, I don't
know how to turn the raw bytes into characters that
Windows Notepad can recognize. Trying to use the
ALT+xxxx shortcut, I run into situations where instead of
inserting a character, Windows asks me if I am sure I want
to close Notepad, or, for example, Windows will delete the
last character before the cursor.
Any thoughts?
In addition, can anyone tell me why the code below was
rejected?
map{$i=0;y/ #
/01/;map{$x[$i++].=$_}/(...)./g}<>;print
map{index('\xf6\xde$\x92\xe7\xce\xe7\x9e\xb7\x92\xf3\x9e\xf3\xde\xe4\x92\xf7
\xde\xf7\x9e',pack B16,$_)/[EMAIL PROTECTED]
-Riley (o0lit3)
"Juho Snellman" <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]
> I'm suprised at how different all the solutions for the LED reader
> minigolf were. Most solutions of course had the same basic approach of
> somehow hashing the led characters into a single byte, and then
> transforming the bytes into the correct ascii characters. (The main
> exception being tybalt who somehow managed to hash directly to the
> correct bytes.)
>
> Anyway, here's an explanation of my solution. There's a cute trick
> with vec() that I haven't seen used before (found with the usual
> approach of "pure luck").
>
> The solution:
>
> #!perl -lp
> s\#\++vec$a,$./2^2*pos,1\ge}{$_=$a;y/\x14\x10^\x1a\x198<\x15\x1c\x18/0-9/
>
> [The contents of the y/// are of course actually raw bytes, but they
> probably wouldn't fare too well in email transit... ]
>
> The hashing is done by finding all "#":s on a line, and modifying a
> corresponding bit in the output string with a vec in lvalue context.
> The bit to modify is calculated from pos and $. using the formula
> ($./2)^(2*pos). The results from this operation can be visualized with
> the following table:
>
> 024
> 0 024
> 1 135
> 1 135
> 2 206
> 2 206
>
> As is apparent from the table, multiple "#":s can map to the same
> bit. The cute trick I mentioned earlier is using ++ on vec, which
> effectively flips the value of the bit between 0 and 1 on each
> invocation (thanks to vec truncating the value). For example '0':
>
> ###
> # #
> # # => 0, 1, 1, 2, 2, 2, 0, 4, 5, 5, 6, 6 => 0010100
> # #
> ###
>
> This turns out to be unique for each led character. I originally used
> addition instead of xor for combining $. and pos, but that had the
> unfortunate side-effect of hashing one of the characters to "-", which
> required escaping when embedded in the y///.
>
> --
> Juho Snellman
