in reply to Re (tilly) 3: Golf: Tree searching
in thread Golf: Tree searching
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:
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:
value entry in the hash.$|++; $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"; }
|
---|
Replies are listed 'Best First'. | |
---|---|
Re (tilly) 4: Golf: Tree searching
by tilly (Archbishop) on Apr 20, 2001 at 07:19 UTC | |
by tadman (Prior) on Apr 20, 2001 at 08:36 UTC | |
by tilly (Archbishop) on Apr 20, 2001 at 15:18 UTC |
In Section
Meditations