We don't bite newbies here... much PerlMonks

### Re: Re: Golf: Tree searching

by twerq (Deacon)
 on Apr 19, 2001 at 22:48 UTC ( #73921=note: print w/replies, xml ) Need Help??

in reply to Re: Golf: Tree searching

Here's 58 chars with a completely different approach. .

sub f{(\$s,\$t)=@_;\$\$t{d}eq\$s?\$t:f(\$s,\$\$t{l})||f(\$s,\$\$t{r})}

--twerq

Replies are listed 'Best First'.
Re (tilly) 3: Golf: Tree searching
by tilly (Archbishop) on Apr 20, 2001 at 01:55 UTC
As koolade pointed out, not all of these work. Yours in particular goes into deep recursion. Works in theory is not good enough, in golf you are going to learn a lot about the assumptions you cannot make. And yours has at least 2 separate (and serious) mistakes.

BTW the counts as I list them are just for the body of the sub, so if yours did work it would be 51 characters.

Playing independently I came in with several solutions that are small, and the following which is similar to yours really does work at 58:

```sub f {
my(\$s,\$t)=@_;\$t?\$\$t{d}eq\$s?\$t:f(\$s,\$\$t{l})||f(\$s,\$\$t{r}):0
}
And here, thanks to koolade, is the test that I use:
```\$t = {
d => 'd',
l => {
d => 'b',
l => { d => 'a', l => 0, r => 0, },
r => { d => 'c', l => 0, r => 0, },
},
r => {
d => 'f',
l => { d => 'e', l => 0, r => 0, },
r => { d => 'g', l => 0, r => 0, },
}
};

sub test {
my \$val = f(@_);
print \$val ? "\$val->{d}:\$val\n" : "\$val\n";
}

test('e',\$t);
test('O',\$t);
Both tilly and indigo have exactly the same result, which is a fix of twerq's including 1) my, and 2) the test for \$t which prevents infinite looping.

Here is a solution which is short, but not 'strict' compliant: sub f{for((\$s,\$t)=@_;\$_=\$\$t{(0,l,r)[\$\$t{d}cmp\$s]};\$t=\$_){}\$t}
Alas, although 4 characters lighter, it doesn't return 0 on misses properly. This does, but is admittedly 3 characters heavy: sub f{for((\$s,\$t)=@_;\$_=\$t&&(0,l,r)[\$\$t{d}cmp\$s];){\$t=\$\$t{\$_}||0}\$t}
Or, a tie, provided the tree has 0-value stubs as it does in other examples: sub f{for((\$s,\$t)=@_;\$_=\$t&&(0,l,r)[\$\$t{d}cmp\$s];){\$t=\$\$t{\$_}}\$t}
Fun test code below for a tree:
```    \$|++;
\$table = {
d => 'h',
l => {
d => 'd',
l => { d => 'b', l => { d => 'a' }, r => { d => 'c' } },
r => { d => 'f', l => { d => 'e' }, r => { d => 'g' } },
},
r => {
d => 'l',
l => { d => 'j', l => { d => 'i' }, r => { d => 'k' } },
r => { d => 'm', l => { d => 'l' }, r => { d => 'n' } },
}
};

foreach \$y (qw[ a b c d e f g h i j k l m n o p q ])
{
\$x = f(\$y,\$table);
print "Answer for \$y = ";
print \$x," ", \$\$x{d},"\n";
}
value entry in the hash.
Well it has to handle both failure and success. However by borrowing shamelessly from tye, I beat my previous, and if we want to be shamelessly non-strict about it, I can improve again.
```sub f {
\$t=pop;\$t=\$\$t{\$c>0?l:r}while\$c=\$\$t{d}cmp\$_[0]and\$t;\$t
}
By my count this is 53 characters.

Create A New User
Node Status?
node history
Node Type: note [id://73921]
help
Chatterbox?
 [jedikaiti]: Happy Monday [Tanktalus]: What's so happy about it? :P [Tanktalus]: ww: your parens don't line up. Sure, perl can look like line noise, but the parens still have to line up. ;) [Tux]: 1nickt did you build 5.24.3, 5.26.1, and 5.27.4 yet?

How do I use this? | Other CB clients
Other Users?
Others browsing the Monastery: (8)
As of 2017-09-25 16:29 GMT
Sections?
Information?
Find Nodes?
Leftovers?
Voting Booth?
During the recent solar eclipse, I:

Results (282 votes). Check out past polls.

Notices?