in reply to (Golf) Cryptographer's Tool #1

Here's mine. Not particularly interesting, really, and probably missing out on a few standard golf tricks.
```sub c {my(@a,%h,\$i)=split//,pop;my\$s=pop;\$s=~s!(.)!\$h{\$1}||=\$a[\$i++]!g
+e;\$s}
[download]```
67 chars.

Update: got rid of one character by using substr:

```sub c {my(\$a,%h,\$i)=pop;my\$s=pop;\$s=~s!(.)!\$h{\$1}||=substr\$a,\$i++,1!ge
+;\$s}
[download]```
66 chars.

Update 2: if you don't need strict:

```sub c {(\$s,\$a)=@_;my(%h,\$i);\$s=~s!(.)!\$h{\$1}||=substr\$a,\$i++,1!ge;\$s}
[download]```
61.

Update 3: Oh, and the reason I'm still initializing %h and \$i is so that the sub will work on repeated invocations. If we don't *have* to do that, this will do:

```sub c {(\$s,\$a)=@_;\$s=~s!(.)!\$h{\$1}||=substr\$a,\$i++,1!ge;\$s}
[download]```
51.

Update 4: Use \$_:

```sub c {(\$_,\$a)=@_;s!(.)!\$h{\$1}||=substr\$a,\$i++,1!ge;\$_}
[download]```
47.

Update 5: Okay, so none of these actually *work* per the spec. Thanks, tilly. :)

Replies are listed 'Best First'.
Re (tilly) 2: (Golf) Cryptographer's Tool #1
by tilly (Archbishop) on Jun 20, 2001 at 01:21 UTC
*ahem*

All 4 solutions fail if the alphabet includes a 0. (That is zero.)

I move that it be resolved that for the purposes of this golf the alphabet may not include a zero.

Is there a second?

:)

I move that it be resolved that the golf meet the original spec. Particularly since it was rather carefully and precisely defined.

Which means that you lazy bums also need to handle the ASCII character "\n" correctly. :-P

Absolutely. :)