Infix to RPN Post Mortem Back to Perl Golf
Accepted | Artistic/Unorthodox | Rejected

Referee comments are in italics. Blue rows denote golfer's best solution. Red text denotes special characters.

Accepted

ScoreGolferSubmit TimeCode
98.22 Eugene van der Pijll Thu Sep 5 22:19:40 2002 
-p sub
w{s/
([@_^+-])(.*)/ $2 $1/g}sub
f{s/\(([^()]+)\)/f($_=$+)/e?&f:w~w;$_}f
s/\s//g/s/\d+/$&
/g
99.24 BoB (Best of Breed - referees) Thu Sep 5 22:16:57 2002 
-p sub
w{s/
([@_^+-])(.*)/ $2 $1/g}sub
f{s/\(([^()]+)\)/f($_=$1)/e&&&f;w~w;$_}s/\s//g+s/\d+/$&
/g+f
optimize tiebreak
99.26 Eugene van der Pijll Thu Sep 5 22:04:21 2002 
-p s/\s//g;s/\d+/$&
/g;sub
w{s/
([@_^+-])(.*)/ $2 $1/g}sub
f{s/\(([^()]+)\)/f($_=$1)/e&&&f;w~w;$_}f
Into the double digits...
100.34 BoB (Best of Breed - referees) Thu Sep 5 21:26:40 2002 
-p s//m/;y/ ^I//dwhile
s^[+-]M(?![*/])|[*/]M|m-*\d+|.M\)^M^?push
_,$&=~/[*-9]+/g:s/m(.)/$+m/i;s/M/@_/
push a /g match of $& instead of $+
101.25 Eugene van der Pijll Thu Sep 5 22:03:22 2002 
-p s/\s//g;s/\d+/$&
/g;sub
w{s/
([@_^+-])(.*)/ $2 $1/g}sub
f{s/\(([^()]+)\)/f($_=$1)/e&&&f;w;w 7;$_}f
101.31 BoB (Best of Breed - referees) Thu Sep 5 20:56:07 2002 
-p s//m/;y/ ^I//dwhile
s^([+-])M(?![*/])|([*/])M|m(-*\d+)|.M\)^M^?~$+?push
_,$+:1:s/m(.)/$+m/i;s/M/@_/
Rick's 102, with the push moved outside the s///.
102.23 BoB (Best of Breed - referees) Thu Sep 5 18:27:42 2002 
-p sub
w{s/
([@_^+-])(.*)/ $2 $1/}sub
f{s-\(([^()]+)\)-f($_=$1)-e||w||w(7)?&f:$_}s/\s//g+s/\d+/$&
/g+f
remove local from eugene's 107, reorganize a bit to improve tie.
102.26 Eugene van der Pijll Thu Sep 5 21:41:11 2002 
-p s/\s//g;s/\d+/$&
/g;sub
w{s/
([@_^+-])(.*)/ $2 $1/}sub
f{s-\(([^()]+)\)-$_=$1;&f-e||w||w(7)?&f:$_}f
Of course. The true&tested method of "Can't possibly work, let's try it anyway."
102.30 Rick Klement Sat Sep 7 19:22:19 2002 
-p s//m/;y/ ^I//dwhile
s^(([+-])M(?![*/])|([*/])M|m(-*\d+))(?{push
_,$+})|.M\)^M^||s/m(.)/$+m/i;s/M/@_/
.03 tiebreak improvement
102.33 Rick Klement Thu Sep 5 14:18:47 2002 
-p s//m/;y/ ^I//dwhile
s^([+-])M(?![*/])|([*/])M|m(-*\d+)|.M\)^~$+&&push
_,$+;M^e||s/m(.)/$+m/i;s/M/@_/
that's y///... (actually, I think the copy/paste problem is on my end :( )

nice...

103.23 BoB (Best of Breed - referees) Tue Sep 3 21:18:20 2002 
-p y/ ^I//d+s//^I/;~$+&&push
a,$+while
s^([+-]) (?![*/])|([*/]) |^I(-*\d+)|. \)^ ^||s/\s(.)()/$1^I/;s/ /@a/
Heh, heh
103.30 BoB (Best of Breed - referees) Tue Sep 3 19:38:22 2002 
-p y/ ^I//d+s//m/;~$+&&push
a,$+while
s^([+-])M(?![*/])|([*/])M|m(-*\d+)|.M\)^M^||s/m(.)()/$1m/i+s/M/@a/
Rick's 104.31, slightly reorganized
103.33 Ton Hospel Sun Sep 8 03:48:17 2002 
-Xlp sub
c{($|--&&/\s/*s///g%/^./+/^-*\d+/?$&:c()).pop.($|--?/^[+-]/:s///
x/^[*\/]/and$".c(" $&"))}$_=c
103.34 Ton Hospel Sun Sep 8 03:05:03 2002 
-Xlp sub
c{($|--&&s///g%/^./+/^-*\d+/?$&:c()).pop.($|--?/^[+-]/:s///
x/^[*\/]/and$".c(" $&"))}/\s/;$_=c
103.61 Ton Hospel Sun Sep 8 02:56:36 2002 
-Xlp sub c{($|--&&s///g%/^./+/^-?\d+/?$&:c()).pop.($|--?/^[+-]/:s/// x/^[*\/]/and$".c(" $&"))}/\s/;$_=c
104.25 Ton Hospel Thu Sep 5 01:18:05 2002 
-p sub
c{$|--&&s///-/./+/^-*\d+/?$&+0:c(),@_,($|--?/^[+-]/:s///
x/^[*\/]/and$",c(" $&"))}y/ ^I//d;print
c
since i can't seem to find anything big, let's think small
104.26 Ton Hospel Wed Sep 4 22:45:06 2002 
-p sub
c{$|--&&s///%/./+/^-*\d+/?$&+0:c(),@_,($|--?/^[+-]/:s///
x/^[*\/]/and$",c(" $&"))}y/ ^I//d;print
c
almost forgot one more line
104.30 Ton Hospel Wed Sep 4 22:40:50 2002 
-p sub
c{$|--&&s///%/./+/^-*\d+/?$&+0:c(),@_,($|--?/^[+-]/:s///
x/^[*\/]/and$",c(" $&"))}y/ ^I//d;print c
Actually reading the tiebreaker code can be useful
104.31 Rick Klement Tue Sep 3 15:20:03 2002 
-p y/ ^I//d+s//m/;1while
s^([+-])M(?![*/])|([*/])M|m(-*\d+)|.M\)^~$+&&push
a,$+;M^e||s/m(.)/$+m/i+s/M/@a/
the y/// contains both a and a . Some tiebreak improvement :)

hmm .. closing in on double digits

104.57 Ton Hospel Wed Sep 4 17:17:57 2002 
-p sub c{$|--&&s///-/./+/^-?\d+/?$&+0:c(),@_,($|--?/^[+-]/:s/// x/^[*\/]/and$",c(" $&"))}y/ ^I//d;print c
another way to handle the \n
104.61 Ton Hospel Wed Sep 4 00:24:09 2002 
-Xlp sub c{($|--&&s///%/./+/^-?\d+/?$&:c()).pop.($|--?/^[+-]/:s/// x/^[*\/]/and$".c(" $&"))}s/\s//g;$_=c
Back to a string representation
105.62 Ton Hospel Wed Sep 4 00:10:41 2002 
-lp sub c{$|--&&s///%/./+/^-?\d+/?"$&":c(),@_,(--$|?s///*/^[*\/]/:/^[+-]/)?c("$&"):()}s/\s//g;$_="@{[c]}"
goodbye overload !
106.34 BoB (Best of Breed - referees) Tue Sep 3 11:58:24 2002 
-p y/^I //d;s//m/;1while
s!([+-])M(?=[
)+-])|([*/])M|m(-?\d+)|.M\)!~$+&&push@a,$+;M!e||s/m(.)/$+m/i;s/M/@a/
Rick's solution with a hard-linebreak inside the character group, instead of a $| outside it.
107.29 Eugene van der Pijll Thu Sep 5 17:55:59 2002 
-p s/\s//g;s/\d+/$&
/g;sub
w{s/
([@_^+-])(.*)/ $2 $1/}sub
f{s-\(([^()]+)\)-local$_=$1;&f-e||w||w(7)?&f:$_}f
107.44 BoB (Best of Breed - referees) Tue Sep 3 05:01:50 2002 
-pibless[s)$) y/ ^I//d;s/(?<!\)|\d)-*\d+/($^I $&)])/g;s/.?/"use
overload
'$&',sub{$^I $&)]}"/gee;eval$`;s///
Ton's solution with the tiebreaker improved beyond Rick's solution by applying linebreaks and changing '!' delimiters to ')'.
107.46 Rick Klement Mon Sep 2 13:45:07 2002 
-p y/  //d;s//m/;1while
s!([+-])M(?=$|[)+-])|([*/])M|m(-?\d+)|.M\)!~$+&&push@a,$+;M!e||s/m(.)/$+m/i;s/M/@a/
That's y///d at the start. 7:57 for v5 on my 200Mhz, should be fine on a 600 :)
107.74 Ton Hospel Mon Sep 2 23:26:19 2002 
-pibless[s!$! y/ ^I//d;s/(?<!\)|\d)-*\d+/($^I $&!])/g;s/.?/"use overload '$&',sub{$^I $&!]}"/gee;eval$`;s///
flog! flog!
108.30 Eugene van der Pijll Thu Sep 5 17:52:27 2002 
-p s/\s//g;s/\d+/$&
/g;sub
w{s-
([@_*/])(.*)- $2 $1-}sub
f{(s-\(([^()]+)\)-local$_=$1;&f-e||w||w"^")?&f:$_}f
108.61 Ton Hospel Tue Sep 3 23:33:32 2002 
-lp sub c{--$|?c():s///%/./+/^-?\d+/?"$&":c(),@_,(--$|?s///*/^[*\/]/:/^[+-]/)?c("$&"):()}s/\s//g;$_="@{[c]}"
grand unification
109.51 pom Sat Sep 7 10:01:20 2002 
-p y/ ^I//d;s/\b\D/x$&/g;s--$_=$1;s// $2 $1/while/x(.)(\PL+)/+/x([*\/])(\PL+)/;$_-ewhile/(.+)x/+/\(([^()]+)\)/
109.51 Alexey Rudenko Fri Sep 6 17:26:41 2002 
-lp $p=qr'(([^() ]|\(( |(??{$p}))*\))+)';s/\s//g;s//$1 $3 $2/while/(.*)([*\/])$p/|/(.*[\d)])([-+])$p/;y/()//d
109.78 Ton Hospel Mon Sep 2 23:05:52 2002 
-p y/ ^I//d;s#(?<!\)|\d)-?\d+#(bless[s!\$! $&!])#g;s/.?/"use overload '$&',sub{s!! $&!;shift}"/gee;eval$`;s///
Still haven't flogged this horse to death
110.36 Eugene van der Pijll Thu Sep 5 17:06:53 2002 
-p s/\s+//g;s/\d+/$&
/g;sub w{s#
([@_*/])(.*)# $2 $1#}sub
f{(s#\(([^()]+)\)#local$_=$1;&f#e||w||w"-+")?&f:$_}f
110.83 Ton Hospel Mon Sep 2 10:19:58 2002 
-lp s/\s//g;s#(?<!\)|\d)-?\d+#(bless[s!\$!$& !])#g;s/.?/"use overload '$&',sub{s!!$& !;shift}"/gee;eval$`;chop
the code is now completely free of the list of operators, so it will work for the full perl operator set
112.37 Eugene van der Pijll Thu Sep 5 16:34:33 2002 
-p s/\s+//g;s/\d+/$&
/g;sub w{s#
([@_*/])(.*)# $2 $1#}sub
f{1while s#\(([^()]+)\)#local$_=$1;&f#e||w||w"-+";$_}f
112.63 Michael Thelen Sat Sep 7 20:01:19 2002 
-lp y/()^I /^I_/d+s@^I|(?<=\w)\D@($"=" $&$\"")=~s#_(.*?)^I|(\S)([ */${":"x/\G[+-]/}]*)#$2 #;$+@eg;$_.=$";y/ //s+chop
Yet another stroke... today is a good day. Went back to using $" to avoid having to use /s or \C in the inner regex.
113.28 Michael Thelen Sat Sep 7 19:55:24 2002 
-lp y/()^I /^I_/d+s@^I|(?<=\w)\D@($/="
$&$/")=~s#_(\C*?)^I|(\S)([
*/${":"x/\G[+-]/}]*)#$2
#;$+@eg;$_.=$/;y/
/ /s+chop
Continuing with the theme of absolute obfuscation... ')' is now replaced by '_' to make the use of \w possible.
114.68 Ton Hospel Tue Sep 3 22:06:02 2002 
-lp sub a{join$",f(),@_,/^[+-]/?a("$&"):()}sub f{s///;/./;/^-?\d+/?"$&":a,@_,s///*m!^[*/]!?f("$&"):()}s/\s//g;$_=a
mm, recursive descent is running out of breath
114.77 Ton Hospel Mon Sep 2 03:28:11 2002 
-lp s/\s//g;s#(^|[*-/(])(-?\d+)#$1(bless[s!\$!$2 !])#g;s/.?/"use overload '$&',sub{s!!$& !;shift}"/gee;eval$`;chop
115.26 Michael Thelen Sat Sep 7 19:41:05 2002 
-lp y/)( ^I/ ^I/d+s@^I|(?<=\d| )\D@($/="
$&$/")=~s# (.*?)^I|(\S)([
*/${":"x/\G[+-]/}]*)#$2
#s;$+@eg;$_.=$/;y/
/ /s+chop
How absolutely confusing! Replacing parens with whitespace makes this probably the most obfuscated program I've ever written.
115.55 pom Fri Sep 6 22:35:18 2002 
-p y/ ^I//d;s/\b/x/g;s~~$_=$1;s/x([$1])(-?)x(\PL+)/ $2$3 $1/while/x(\D)/|/x([*\/])/;$_~ewhile/x(.+)x/|/\(([^()]+)\)/
115.73 Ton Hospel Mon Sep 2 03:16:10 2002 
-p s#(^|[*-/(])\s*(-?)\s*(\d+)#$1(bless[s!\$! $2$3!])#g;s/.?/"use overload '$&',sub{s!! $&!;shift}"/gee;eval$`;s///
Sigh. Sacrifise more characters
116.27 Qingning Huo Sat Sep 7 22:59:34 2002 
-p s/\s//g/s/(?<!\)|\d)-?\d+/
$&/g;1while
s/\(\)//||s#
(-?\d+)|(?![-+][*/])([^(])(?!
|\()|(.)$##&&push@0,$+;$_="@0
"
116.28 Qingning Huo Sat Sep 7 22:46:05 2002 
-p s/\s//g;s/(?<!\)|\d)-?\d+/
$&/g;1while
s/\(\)//||s#
(-?\d+)|(?![-+][*/])([^(])(?!
|\()|(.)$##&&push@0,$+;$_="@0
"
116.48 Rick Klement Mon Sep 2 00:27:49 2002 
-pi])M(?=$|[) y/  //d;s//m/;1while
s/\(M\)/M/|s!([^($^I+-])|([*/$^I*/])|m(-?\d+)!push@a,$+;M!e||s/M(.)/$+m/i;s/M/@a/
that y/// has both a space and a hard tab in it y/ \t//d
116.48 Jeff Moore Sat Sep 7 02:51:20 2002 
-p use
overload
nomethod,sub{bless[(map"@$_",@_[0,1]),pop]};y/ ^I//d;s/(?<!\d|\))-*\d+/bless(['$&'])/g;$_="@{+eval}
"
116.54 Jeff Moore Sat Sep 7 02:19:55 2002 
-pl use
overload
nomethod,sub{bless[(map"@$_",@_[0,1]),pop]};y/ ^I//d;s/(?<!\d|\))-*\d+/bless(['$&'])/g;$_="@{+eval}"
116.65 Stephen Turner Mon Sep 2 20:16:09 2002 
-pl use
overload'0+',sub{"@{$_[0]}"},nomethod,sub{bless\@_};y/ ^I//d+s/(?<![\d)])-?\d+/(bless["$&"])/g;$_=eval;y/ //s
Just a better tie breaker.
116.66 Stephen Turner Mon Sep 2 14:09:12 2002 
-pl use
overload'""',sub{"@{$_[0]}"},nomethod,sub{bless\@_};y/ ^I//d+s/(?<![\d)])-?\d+/(bless["$&"])/g;$_=eval;y/ //s
One shot worse than my 115.47 solution, but shouldn't use so much memory or time!

Sorry about the erronous rejection.

117.26 Qingning Huo Sat Sep 7 22:24:40 2002 
-p y/ ^I//d;s/(?<!\)|\d)-?\d+/
$&/g;1while
s/\(\)//||s#
(-?\d+)|(?![-+][*/])([^(])(?=[^(
])|(.)$##&&push@0,$+;$_="@0
"
117.52 Jeff Moore Fri Sep 6 23:48:09 2002 
-pl use
overload
nomethod,sub{bless["$_[0][0] $_[1][0] ".pop]};y/ ^I//d;s/(?<!\d|\))-*\d+/bless(['$&'])/g;$_=eval->[0]
117.64 Michael Thelen Sat Sep 7 05:42:38 2002 
-lp y/ ^I//d+s@\(|(?<=[\d)])\D@($"=" $&$\"")=~s!\)(.*?)\(|([^ (])([ */${':'x/\G[+-]/}]*)!$2 !;$+@eg;$_.=$";y/ //s+chop
Simple removal of an unnecessary space character. Now I just need to find the other 20 unnecessary characters.
117.66 Jeff Moore Thu Sep 5 22:21:07 2002 
-pl use
overload'""',sub{$_[0][0]},nomethod,sub{bless["@_"]};y/ ^I//d;s/(?<!\d|\))-*\d+/bless(['$&'])/g;$_=eval;y/ //s
118.45 Stephen Turner Sun Sep 1 20:16:22 2002 
-p sub
n{bless\@_}use
overload
nomethod,sub{n@{$_[0]},@{$_[1]},pop};y/ ^I//d+s/(?<![\d)])-?\d+/(n"$&")/g;$_="@{eval
}
"
Save a lot of strokes in the constructor by using arrays instead of scalars as the objects. WARNING: Uses a lot of memory for the final test. It does complete within the memory and speed constraints though.
118.54 Alexey Rudenko Wed Sep 4 11:14:24 2002 
-lp $p=qr'-?\d+|\(([^()]|(??{$p}))*\)';s/\s//g;s// $2 $1/while/([*\/])($p)/|/(?<=[-+\d)])([-+])($p([*\/]$p)*)/;y/()//d
118.57 Alain "TopMach" Dupuis Sun Sep 8 02:32:16 2002 
-p y/ ^I//d;sub t{$_=pop;for$+('*/','+-'){1while s/(\S)([$+])(-?(\d+| .)*)/$1 $3 $2/}$_}1while s/\(([^()]+)\)/t$1/e;t$_
118.64 Michael Thelen Sat Sep 7 00:31:12 2002 
-lp y/ ^I//d+s@\(|(?<=[\d)])\D@($"=" $& $\"")=~s!\)(.*?)\(|([^ (])([ */${':'x/\G[+-]/}]*)!$2 !;$+@eg;$_.=$";y/ //s+chop
Who knew you could interpolate $" as "$\""? Strange...
119.58 Michael Thelen Fri Sep 6 20:53:44 2002 
-lp y/ ^I//d+s@\(|(?<=[\d)])\D@($*=" $& $*")=~s!\)(.*?)\(|([^ (])([ */${':'x/\G[+-]/}]*)!$2 !;$+@eg;$_.=$*;y/ //s+s/ $//
Just a tiebreak improvement, but hey, every little bit helps.
119.59 Alain "TopMach" Dupuis Sun Sep 8 01:40:32 2002 
-lp s/\s//g;sub t{$_=pop;for$+('*/','+-'){1while s/(\S)([$+])(-?(\d+| .)*)/$1 $3 $2/}$_}1while s/\(([^()]+)\)/t$1/e;t$_
119.61 Michael Thelen Thu Sep 5 23:01:23 2002 
-lp y/ ^I//d+s@\(|(?<=[\d)])\D@($s=" $& $s")=~s!\)(.*?)\(|([^ (])([ */${':'x/\G[+-]/}]*)!$2 !;$+@eg;$_.=$s;y/ //s+s/ $//
Inline, inline, inline.
119.62 Alma Media Golf Team Tue Sep 3 09:46:30 2002 
-p s/\s//g;map{push@o,pop@1while@1*"$_$1[-1]"!~/\(|[\/*][-+]$|\d\D/;/\)/?pop@1:push@1,$_}/(?<![\d)])-?\d+|.?/g;$_="@o
"
numbers visit the stack now
120.27 Qingning Huo Sat Sep 7 08:06:32 2002 
-p y/ ^I//d;s/(?<!\)|\d)-?\d+/
$&/g;1while
s/\(\)//||s#
(-?\d+)|([^(])(?=[-+)])|([*/])(?=[*/])|(.)$##&&push@0,$+;$_="@0
"
120.80 Adam Antonik Fri Sep 6 16:07:18 2002 
-lp use overload'""',sub{${$_[0]}},nomethod,\*a;sub a{$_=pop;bless/g/?\"-$_[0]":\"@_$_"}s/-/- /g+s/\d+/a("$&")/g;*_=eval
121.42 pom Thu Sep 5 23:47:10 2002 
-p y/ ^I//d;s/\b/x/g;s/x(.+)x/($1)/;1while
s!\(([^()]+)\)!$_=$1;s/x([$1])(-?)x(\PL+)/ $2$3 $1/while/x(\D)/|/x([*\/])/;$_!e
121.59 Michael Thelen Thu Sep 5 20:55:00 2002 
-lp y/ ^I//d+s@\(|(?<=[\d)])\D@$z=$:x/\G[+-]/;($s=" $& $s")=~s+\)(.*?)\(|([^( ])([ */$z]*)+$2 +;$+@eg;$_.=$s;y/ //s+s/ $//
Accidentally stumbled across the fact that '+' does not need to be popped off the stack when '+' or '-' is seen. I think this is the first time I've ever used $:, ever.
121.60 Mykola Zubach Wed Sep 4 11:54:31 2002 
-lp y/() \t/<>/d;for(s/(?<!>|\d)-?\d+/<$&>/g;s/<(<[^>]+>)>/$1/||s@>([*/]|(.))<([^<>]+)>(?(2)(?![*/]))@ $3 $1>@;){}y/<>//d
121.60 Mark Hillebrand Thu Sep 5 08:47:44 2002 
-lp $_="($_)";y+(-9 ^I+b-s+d;1while s*b([^b]+?)c*$_=$1;for$/qw(di eg){s-(?<=\S)([$/])(g?[^eg$/]+)- $2 $1-g}y+b-s+(-9+;$_*e
one char & tie breaking
121.62 Chia-chi Yeh Tue Sep 3 17:56:06 2002 
-pl s/\s//g;$_="($_)";0while s!\(([^()]*)\)!$a=$1;for(qw(*/ +-)){0while$a=~s#(\S)([$_])(.*?\S)([*/+-]|$)#$1 $3 $2$4#}$a!e
121.69 Ton Hospel Tue Sep 3 18:27:29 2002 
-Xlp sub f{(s/^-?\d+//?$&:a(!s/.//).!s///).pop.(s!^[*/]!!&&$".f(" $&"))}sub a{f.pop.(s/^[+-]//&&$".a($".$&))}s/\s//g;$_=a
Back to playing with a real parser
122.27 Qingning Huo Sat Sep 7 07:04:26 2002 
-p y/ ^I//d;s/(?<!\)|\d)-?\d+/
$&/g;1while
s/\(\)//||s#([-+*/])(?=[-+)])|([*/])(?=[*/])|
(-?\d+)|(.)$##&&push@o,$+;$_="@o
"
122.43 pom Tue Sep 3 21:57:18 2002 
-p y/ ^I//d;s/\b/x/g;s/x(.+)x/($1)/;1while
s!\(([^()]+)\)!$_=$1;s/x([$1])(-?)x(\PL+)/ $2$3 $1/while/x([*\/])/||/x(\D)/;$_!e
122.64 Mark Hillebrand Wed Sep 4 14:56:24 2002 
-lp $_="($_)";y+)(-9 ^I+a-s+d;1while s|b([^ab]+)a|$_=$1;for$/qw(di eg){s:(?<=\S)([$/])(g?[^eg$/]+): $2 $1:g}y+d-s+*-9+;$_|e
Abuse of $/ saves one char. 2nd try
122.76 Adam Antonik Thu Sep 5 19:17:41 2002 
-lp sub a{$_=pop;bless/g/?\"-$_[0]":\"@_$_"}use overload'""',sub{${$_[0]}},nomethod,\*a;s/-/- /g+s/0*(\d+)/a($1)/g;*_=eval
123.34 Eugene van der Pijll Thu Sep 5 16:27:30 2002 
-p s/\s+//g;s/\d+/$&
/g;sub w{s#
([@_*/])(.*)# $2 $1#}sub
f{local
$_=pop;1while s#\(([^()]+)\)#f($1)#e||w||w"-+";$_}
$_=f$_
climbing fast :)
123.59 Michael Thelen Wed Sep 4 20:46:34 2002 
-lp y/ ^I//d+s@\(|(?<=[\d)])\D@$z="+-"x!/\G[+-]/;($s=" $& $s")=~s/\)(.*?)\(|([^( ])([^($z]*)/$2 /;$+@eg;$_.=$s;y/ //s+s/ $//
Tiebreaker improvement. Hopefully the tab char comes through okay.
123.60 Alma Media Golf Team Mon Sep 2 16:32:21 2002 
-p s/\s//g;map{push@0,pop@1while@1*"$_$1[-1]"!~/\d|\(|[*\/][+-]/;/\)/?pop@1:push@{1-/\d/},$_}/(?<![\d)])-?\d+|.?/g;$_="@0
"
123.64 Michael Thelen Sat Sep 7 17:21:25 2002 
-lp y/ ^I//d+s@\(|(?<=[\d)])\D@$z=!/\G[+-]/&&'"';($"=" $&$\"")=~s+\)(.*?)\(|([^ (])([*/$:$z]*)+$2 +;$+@eg;$_.=$";y/ //s+chop
Not my best solution, but using the usage of $: is interesting.
123.65 Michael Thelen Wed Sep 4 20:18:13 2002 
-lp s/\s//g;s@\(|(?<=[\d)])\D@$z="+-"x!/\G[+-]/;($s=" $& $s")=~s!\)(.*?)\(|([^( ])([^($z]*)!$2 !;$+@eg;$_.=$s;y/ //s;s/ $//
Two more strokes with y///s. Now, how to avoid the need for s/ $// ?
123.65 Chia-chi Yeh Tue Sep 3 17:46:59 2002 
-pl s/\s//g;$_="($_)";0while s!\(([^()]*)\)!$a=$1;for$b('*/','+-'){0while$a=~s#(\S)([$b])(.*?\S)([*/+-]|$)#$1 $3 $2$4#}$a!e
123.76 Adam Antonik Wed Sep 4 17:31:34 2002 
-pl sub a{$_=pop;bless[/g/?"-$_[0]":"@_$_"]}use overload nomethod,\*a,'""',sub{$_[0][0]};s/-/- /g+s/0*(\d+)/a($1)/g;$_=eval
124.43 Eugene van der Pijll Wed Sep 4 17:40:19 2002 
-lp sub
_{bless\"@_"}use
overload
neg,sub{_"-$_[0]"},qw/nomethod
_
""/,sub{${$_[0]}};s/-/
-/g;s/\d+/_("$&")/g;$_=eval;y/ //s
Slightly better tiebreaker. There must be a better way.
124.49 Eugene van der Pijll Tue Sep 3 20:27:48 2002 
-lp sub
_{bless\"@_"}use
overload
neg,sub{_"-$_[0]"},'""',sub{${$_[0]}},nomethod,'_';s/-/
-/g;s/\d+/_("$&")/g;$_=eval;y/ //s
125.38 Mtv Europe Tue Sep 3 17:21:00 2002 
-p y/) ^I/
/d;$a.=/\d/?!push
o,$_:/[^(0]/+/[*\/]/&&do{push o,$/while($/=chop$a)=~//;$/.$_}x/./for/(?<!\d|
)-?\d+|./sg;$_="@o
"
catch pom again!
125.50 pom Tue Sep 3 09:39:06 2002 
-pl $_="($_)";s/\s//g;s/\b/x/g;1while
s!\(([^()]+)\)!$_=$1;s/x([$1])(-?)x(\PL+)/ $2$3 $1/while/x([*\/])/||/x(\D)/;$_!e||s/x//
125.59 Michael Thelen Wed Sep 4 19:52:08 2002 
-p s/\s//g;s@\(|(?<=[\d)])\D@$z="+-"x!/\G[+-]/;($s=" $& $s")=~s/\)(.*?)\(|([^( ])([^($z]*)/$2 /;$+@eg;$_.=$s;$_="@{[split]}
"
Whoa! One of the thousand combinations I tried actually yielded a better score!
126.50 Stephen Turner Sun Sep 1 17:43:03 2002 
-pl sub
n{my$p="@_";bless\$p}use
overload
nomethod,sub{n${$_[0]},${$_[1]},pop};y/ ^I//d+s/(?<![\d)])-?\d+/(n"$&")/g;$_=${eval
}
Save two strokes by building a string inside the constructor instead of inside the overloaded operator.
126.59 Michael Thelen Wed Sep 4 18:46:44 2002 
-p s/\s//g;s@\(|(?<=[\d)])\D@$z=/\G[+-]/||"+-";($s=" $& $s")=~s/\).*?\(|(?<=[^( ])[^($z]*/ /;$&@eg;$_.=$s;$_="@{[/[*-9]+/g]}
"
Okay, that \G is freaking me out. I don't know why I even tried it.
127.37 Mtv Europe Tue Sep 3 16:55:57 2002 
-p y/) ^I/
/d;$a.=/\d/?!push
o,$_:/\(/?0:do{push
o,$/while$/=chop$a,/[*\/]/?$/=~//:$/;$/.$_}x/./for/(?<!\d|
)-?\d+|./sg;$_="@o
"
ahh, today rate is done
127.60 Michael Thelen Wed Sep 4 18:27:02 2002 
-p s/\s//g;s@\(|(?<=[\d)])\D@$z="+-"x$&!~/[+-]/;($s=" $&$s ")=~s/\).*?\(|(?<=[^( ])[^($z]*/ /;$&@eg;$_.=$s;$_="@{[/[*-9]+/g]}
"
Trying to start to use whitespace wisely. Failing quite miserably so far.
127.65 Markus Laire Sat Sep 7 01:21:17 2002 
-pl s/\s//g;$x=xx;$a=s#[*/]#x$_#?'(([+-] )?x)':x,s/[-+(]$/x$_ x/,$x=~s/$a/$_ $1/for/(?<!\d|\))-?\d+|./g;$_=$x;y/()x / /s;s/ $//
128.38 Mtv Europe Mon Sep 2 12:55:50 2002 
-p y/) ^I/
/d;$a.=/\d/?!push
o,$_:/\(/?0:do{push
o,$/while$/=chop$a,/\*|\//?$/=~//:$/;"$/$_"x/./}for/(?<!\d|
)-?\d+|./sg;$_="@o
"
128.47 Michael Thelen Wed Sep 4 17:53:08 2002 
-p s/\s//g;s@\(|(?<=[\d)])\D@$z="+-"x$&!~/[+-]/;($s="$& $s")=~s/\)(.*?)\(|(?<=^[^(])([^($z]*)/ /;$+@eg;s/$/
$s/;$_="@{[split]}
"
Negation of =~ instead of the character class saves another stroke.

hmm .. you sure do keep a ref buzy :)

128.49 Stephen Turner Sun Sep 1 13:35:36 2002 
-pl sub
n{my$p=pop;bless\$p}use
overload
nomethod,sub{n"${$_[0]} ${$_[1]} ".pop};y/ ^I//d+s/(?<![\d)])-?\d+/(n"$&")/g;$_=${eval
}
Better tie-breaker.
128.49 pom Mon Sep 2 13:03:03 2002 
-p $_="($_)";y/ ^I//d;s/\b/x/g;1while
s!\(([^()]+)\)!$_=$1;s/x(\Q$1\E)(-?)x(\PL+)/ $2$3 $1/while/x([*\/])/||/x([+-])/;$_!e;y/x//d
128.52 Stephen Turner Sun Sep 1 13:32:16 2002 
-pl sub
n{my$p=pop;bless\$p}use
overload
nomethod,sub{n"${$_[0]} ${$_[1]} ".pop};y/ ^I//d+s/(?<![\d)])-?\d+/(n"$&")/g;$_=${+eval}
Another stroke bites the dust.
128.52 Alexey Rudenko Tue Sep 3 11:04:39 2002 
-lp $p=qr'-?\d+|\(([^()]|(??{$p}))*\)';y/ \t//d;1while s/([-+\d)])([-+])($p([*\/]$p)*)/$1 $3 $2/||s/([*\/])($p)/ $2 $1/;y/()//d;
128.59 Alain "TopMach" Dupuis Sun Sep 8 00:27:13 2002 
-lp s/\s//g;$-=/\(/,s!\({$-}([^()]+)\){$-}!$_=$1;for$+('*/','+-'){1while s/(\S)([$+])(-?(\d+| .)*)/$1 $3 $2/}$_!ewhile/\S[^\w ]/
Tie breaker optimization
128.59 Mykola Zubach Tue Sep 3 14:07:03 2002 
-lp y/() \t/<>/d;for(s/(?<![>\d])-?\d+/<$&>/g;s/<(<[^>]+>)>/$1/||s@>([/*]|([+-]))<([^<>]+)>(?(2)([^*/]|$))@ $3 $1>$4@;){}y/<>//d
There was invisible "\n" in "links" browser in previously rejected version. Got new one, also improved.
128.66 Alain "TopMach" Dupuis Sat Sep 7 22:50:47 2002 
-lp s/\s//g;$%=/\(/,s!\({$%}([^()]+)\){$%}!$_=$1;for$o('*/','+-'){1while s#(\S)([$o])(-?(\d+| .)*)#$1 $3 $2#}$_!ewhile/\S[^\w ]/
129.47 Michael Thelen Wed Sep 4 17:50:18 2002 
-p s/\s//g;s@\(|(?<=[\d)])\D@$z="+-"x$&=~/[^+-]/;($s="$& $s")=~s/\)(.*?)\(|(?<=^[^(])([^($z]*)/ /;$+@eg;s/$/
$s/;$_="@{[split]}
"
Oops, I had an unnecessary ( in a character class. Now if I could just get rid of $z...
129.52 Stephen Turner Sun Sep 1 13:24:27 2002 
-pl sub
n{my$p=pop;bless\$p}use
overload
nomethod,sub{n"${$_[0]} ${$_[1]} ".pop};y/ ^I//d+s/(?<![\d)])-?\d+/(n"$&")/g;$_=${eval()}
Cheaper to handle unary minus ourselves.

And BoB is beaten 7,5 hours after the rules were available .. impressive

129.64 Markus Laire Fri Sep 6 22:01:11 2002 
-pl s/\s//g;$x=xx;s/$/ /,/[*\/]/&&$x=~s/[+-] x|x/x$_$&/||s/\(|([+-] )/x$1x/+$x=~s/x/$_/for/(?<!\d|\))-?\d+|./g;$_=$x;s/ \) | x//g
129.76 Ton Hospel Mon Sep 2 22:22:29 2002 
-Xlp sub n{/\G-?\d+/cg?$&:!/./g.&a.!//g}sub f{n.pop.(m!\G[*/]!cg&&$".f(" $&"))}sub a{f.pop.(/\G[+-]/cg&&$".a($".$&))}s/\s//g;$_=a
Ok, I broke down and wrote an actual parser Let's start with recursive descent
129.79 Eugene van der Pijll Tue Sep 3 18:17:55 2002 
-lpX sub _{bless\"@_"}use overload neg,sub{_"-@_"},'""',sub{${$_[0]}},nomethod,'_';s/-/ -/g;s/\d+/_("$&")/g;$_=eval;y/ //s;s/ $//
This should work... Third time lucky. No "Unknown filetest" errors now.
130.47 Michael Thelen Wed Sep 4 17:34:56 2002 
-p s/\s//g;s@\(|(?<=[\d)])\D@$z="+-"x$&=~/[(*\/]/;($s="$& $s")=~s/\)(.*?)\(|(?<=^[^(])([^($z]*)/ /;$+@eg;s/$/
$s/;$_="@{[split]}
"
More removal of unnecessary complexity. Thank goodness for $+.
131.38 Mtv Europe Mon Sep 2 12:54:16 2002 
-p y/) ^I/
/d;/\d/?push
o,$_:/\(/?$a.=0:do{push
o,$/while$/=chop$a,/\*|\//?$/=~//:$/;$a.="$/$_"x/./}for/(?<!\d|
)-?\d+|./sg;$_="@o
"
hmm
131.66 Alain "TopMach" Dupuis Sat Sep 7 22:11:34 2002 
-lp s/\s//g;$%=/\(/,s!\({$%}([^()]+)\){$%}!$_=$1;for$o('*/','+-'){1while s#(\S)([$o])(-?\d+( .|\d+)*)#$1 $3 $2#}$_!ewhile/\S[^\w ]/
131.82 TheodoreYoung Tue Sep 3 19:45:10 2002 
-lp sub _{bless\"@_"}use overload'""',sub{${$_[0]}},neg,sub{_"-$_[0]"},nomethod,sub{_"@_[0,1,3]"};s/-/ -/g;s/\d+/ _('$&')/g;$_=eval
Didn't quite retire the idea, eh? :-)
132.35 Mtv Europe Mon Sep 2 12:33:44 2002 
-p y/) ^I/
/d;/\d/?push
o,$_:/\(/?$a.=0:do{push
o,$/while$/=chop$a,/\*|\//?$/=~//:$/;/
/
or$a.=$/.$_}for/(?<!\d|
)-?\d+|./sg;$_="@o
"
outrun pom at tie again!
132.37 pom Mon Sep 2 11:11:43 2002 
-p $_="($_)";y/ ^I//d;s/\b/x/g;1while
s!\(([^()]+)\)!$_=$1;$*=$1,s/y(-?)x(\PL+)/ $1$2 $*/while
s-x([*/])-y-||s/x([+-])/y/;$_!e;y/x//d
Sorry, I sent a wrong version, this one works!
132.59 Alma Media Golf Team Mon Sep 2 15:00:20 2002 
-p s/\s//g;map{push@0,pop@1while@1*')-++-*//*'=~/\Q$_\E.*\Q$1[-1]/;/\)/?pop@1:push@{1-/\d/},$_}"($_)"=~/(?<![\d)])-?\d+|./g;$_="@0
"
132.65 Chia-chi Yeh Tue Sep 3 05:09:22 2002 
-pl s/\s//g;$_="($_)";@p=qw(*/ +-);0while s&\(([^(]*?)\)&$a=$1;0while$a=~s#(\S)([$p[$i]])(.*?\S)([*/+-]|$)#$1 $3 $2$4#||($i^=1);$a&e
132.72 Mark Hillebrand Wed Sep 4 11:10:41 2002 
-lp $_="($_)";y:)(-9 ^I:a-s:d;s.([b-g])g.$1-.g;1while s!b([^ab]+)a!$_=$1;for$k('di','eg'){s:([$k])([^eg$k]+): $2 $1:g}y:d-s:*-9:;$_!e
1st try
132.80 Jasvir Nagra Sun Sep 8 04:56:58 2002 
-pl use overload nomethod,sub{($a,$b,$c,$_)=@_;/"/?"@{$a}":bless[ref$b?@_[$c,!$c,3]:-"@_"]};s/-/- /g;s/0*(\d+)/bless([$1])/g;$_=eval
133.23 Markus Laire Fri Sep 6 19:43:56 2002 
-p s/\s//g;$x="

";s/$/ /,/[*\/]/&&$x=~s/[+-] 
|
/
$_$&/||s/\(|([+-] )/
$1
/+$x=~s/
/$_/for/(?<!\d|\))-?\d+|./g;$_=$x;s/ \) //g;s/ $//
It seems it fails this test: Input: 1 + 2 * 3 Output: 1 2 + 3 *
133.36 pom Mon Sep 2 11:00:54 2002 
-p $_="($_)";y/ ^I//d;s/\b/x/g;1while
s!\(([^()]+)\)!$_=$1;s/y(.)(-?)x(\PL+)/ $2$3 $1/while
s-x([*/])-y$1-||s/x([+-])/y$1/;$_!e;y/x//d
133.41 Mtv Europe Mon Sep 2 10:35:29 2002 
-p y/ ^I//d;/\(/?$#a++:/\d/?push
o,$_:do{push
o,$,while$,=pop
a,/[*\/]/?$,=~//:$,;/\)/||push
a,$,,$_}for/(?<!\d|\))-?\d+|.?/g;$_="@o
"
damn, this looks like the limit for this algorithm, and i wonder what pom will answer?
133.47 Michael Thelen Wed Sep 4 17:30:30 2002 
-p s/\s//g;s@\(|(?<=[\d)])\D@$z="+-"x$&=~/[(*\/]/;($s="$& $s")=~s/\)(.*?)\(|^([^(])([^($z]*)/$2 /;$1||$3@eg;s/$/
$s/;$_="@{[split]}
"
10-stroke gain, yay! It's nice to get rid of an entire statement.
133.53 pom Mon Sep 2 08:41:27 2002 
-p $_="($_)";y/ ^I//d;s/\b/x/g;1while
s!\(([^()]+)\)!$_=$1;$_="$` $2$3 $1$'"while/x([*\/])(-?)x(\PL+)/||/x(\D)(-?)x(\PL+)/;$_!e;y/x//d
133.62 Keith Calvert Ivey Sun Sep 8 03:03:59 2002 
-p s/\s//g;map{$0=m+[*/]+?'*/':'^(';$0[@0]=pop@1until/\(|\d/|"@1"!~/[$0]$/;/\)/?pop@1:push@{1-/\w/},$_}/(?<![\d)])-?\d+|.?/g;$_="@0
"
I hope I retyped this correctly from my laptop, which currently has no Internet connection, into my parents' computer, which has no Perl. Who needs to visit family when I can beat Andrew?

You, sir, are a true golfer.

133.63 Chia-chi Yeh Tue Sep 3 04:34:34 2002 
-pl s/\s//g;$_="($_)";@p=qw(*/ +-);0while s/\(([^(]*?)\)/$a=$1;0while$a=~s#(\S)([$p[$i]])(.*?\S)([*\/+-]|$)#$1 $3 $2$4#||($i^=1);$a/e
133.69 Andrew Savige Sat Sep 7 17:33:47 2002 
-lp y+ ^I++d;$*.=/\d/?"$_ ":($s=~s~^.*?(?=@{['|'x/\(/.'-|\+|'x/[*\/]/]}\(|$)~$_ ~,$&),$s=~s+\) . ++for/(?<![\d)])-?\d+|.?/g;$_=$*;chop
133.79 Adam Antonik Tue Sep 3 17:28:17 2002 
-l use overload nomethod,sub{$_=pop;bless[/g/?"-$_[0]":"@_$_"]},'""',sub{$_[0][0]};$_=<>;s/-/- /g;s/0*(\d+)/\(bless[$1])/g;print eval
134.69 Andrew Savige Sat Sep 7 08:06:47 2002 
-lp y+ ^I++d;$*.=/\d/?"$_ ":($s=~s~^.*?(?=@{['^|'x/\(/.'-|\+|'x/[*\/]/]}\(|$)~$_ ~,$&),$s=~s+\) . ++for/(?<![\d)])-?\d+|.?/g;$_=$*;chop
I feel a bit embarassed flogging the same horse, but I am too tired to think of anything new. Gee, I thought Keith had given up, but he is right behind me now and I fear he will finish stronger.
134.78 Jasvir Nagra Sun Sep 8 04:52:11 2002 
-pl use overload nomethod,sub{$_[3]=~/"/?"@{$_[0]}":bless[ref$_[1]?@_[$_[2],!$_[2],3]:-"@_"]};s/-/- /g;s/0*(\d+)/bless([$1])/g;$_=eval
136.33 Eugene van der Pijll Thu Sep 5 16:12:54 2002 
-p s/\s+//g;s/\d+/a$&
/g;sub w{s#
([@_*/])(-?)a(.*)# $2$3 $1#}sub
f{local
$_=pop;1while
s#\(([^()]+)\)#f($1)#e||w||w"-+";$_}$_=f$_;s/a//
136.48 Mtv Europe Mon Sep 2 07:26:50 2002 
-p s/\s//g;/\(/?$#:++:/\d/?push
o,$_:do{push
o,pop@:while/[*\/]$/?"@:"=~//:$:[-1];/\)/?$#:--:push@:,$_}for/(?<!\d|\))-?\d+|.?/g;$_="@o
"
looks like my previous attempt failed because i misunderstood 'negative' numbers.
136.70 Andrew Savige Sat Sep 7 03:05:48 2002 
-lp y+ ^I++d;$*.=/\d/?"$_ ":($s=~s~^.*?(?=@{["[@{['-+'x/[*\/]/]}(]|\$"x!/\(/]})~$_ ~,$&),$s=~s+\) . ++for/(?<![\d)])-?\d+|.?/g;$_=$*;chop
I hope `/ refs this one for I have smashed the refs 138 wall with my trusty one iron (much to my surprise I must admit).
136.81 Jasvir Nagra Sun Sep 8 04:25:36 2002 
-pl use overload'""',sub{"@{$_[0]}"},nomethod,sub{bless[defined$_[1]?@_[$_[2],!$_[2],3]:-"@_"]};s/-/- /g;s/0*(\d+)/bless([$1])/g;$_=eval
137.31 Qingning Huo Sat Sep 7 06:57:12 2002 
-p y/ ^I//d;s/(?<!\)|\d)-?\d+/
$&/g;s/\(\)//||s#([-+*/])(?=\))|([*/])(?=[-+*/])|([-+])(?=[-+])|
(-?\d+)|(.)$##&&push@o,$+while/./;$_="@o
"
137.36 pom Sun Sep 1 17:59:51 2002 
-p $_="($_)";y/ ^I//d;s/\b/x/g;1while
s!\(([^()]+)\)!$_=$1;1while
s-x([*/])(.?)x(\PL+)- $2$3 $1-||s/x(\D)(-?)x(\PL+)/ $2$3 $1/;$_!e;y/x//d
The tiebreaker for this hole is quite fun...

makes the pm easier to read :)

138.48 BoB (Best of Breed - referees) Sun Sep 1 05:21:32 2002 
-pl sub
e{map$_&&=/(.*\w)(\/|\*)/|/(.*\w)(\+|-)/&&e($1,$',$2)||e($o{$_})||$_,@_=@_;"@_"}s/\s//g;$o{$~++}=$1while
s/\(([^()]+)\)/$~/;$_=e$_
Okay, let's see whether submitting works.
138.55 Mykola Zubach Tue Sep 3 10:54:38 2002 
-lp y/() \t/<>/d;s/(?<![>\d])(-?\d+)/<$1>/g;for($a="<([^<>]+)>";s/<$a>/<$1>/||s@$a([
/*]|([+-]))$a(?(3)([^*/]|$))@<$1 $4 $2>$5@;){}y/<>//d
It's my first try to play perlgolf

Looking good!

138.63 Keith Calvert Ivey Sat Sep 7 11:31:37 2002 
-p s/\s//g;map{if(!/\(|\d/){$0=m+[*/]+?'*/':'^(';$0[@0]=pop@1while"@1"=~/[$0]$/}/\)/?pop@1:push@{1-/\w/},$_}/(?<![\d)])-?\d+|.?/g;$_="@0
"
That push() didn't last long. It's only 0.01 improvement, but I'm running out of time.

Hurry up, Andrew is not far ahead! :-)

138.64 Keith Calvert Ivey Sat Sep 7 11:20:06 2002 
-p s/\s//g;map{if(!/\(|\d/){$0=m+[*/]+?'*/':'^(';push@0,pop@1while"@1"=~/[$0]$/}/\)/?pop@1:push@{1-/\w/},$_}/(?<![\d)])-?\d+|.?/g;$_="@0
"
Changing back to push() made it a lot faster
138.75 Andrew Savige Fri Sep 6 21:48:38 2002 
-lp y~ ^I~~d;$x.=/\d/?"$_ ":($s=~s~@{[/\(/?v94:".*?(?=[@{['-+'x/[*\/]/]}(]|\$)"]}~$_ ~,$&),$s=~s~\) . ~~for/(?<![\d)])-?\d+|.?/g;$_=$x;chop
Caught the refs!!! Amazingly just by flogging the same ol' dead horse. Gee, Amir is moving up fast from behind. Recite litany against fear.
139.81 Jasvir Nagra Sun Sep 8 03:59:35 2002 
-pl use overload'""',sub{$_[0][0]},neg,sub{bless[-"@_"]},nomethod,sub{bless["@_[$_[2],!$_[2],3]"]};s/-/- /g;s/0*(\d+)/bless([$1])/g;$_=eval
140.23 Markus Laire Fri Sep 6 14:26:55 2002 
-p s/\s//g;s/(?<!\d|\))-/f/g;$x="

";s/$/ /,/[*\/]/&&$x=~s/[-+]? 
/ 
$_$&/||s/\(|([+-] )/
$1
/+$x=~s/
/$_/for/\w+|./g;$_=$x;y/f) /- /s;s/ $//
140.46 Amir Karger Fri Sep 6 20:27:18 2002 
-pl y/ ^I//d;s/(?<![\d)])-/^I/g;$*=qr/\(([^()]+|(??{$*}))*\)|^I?\d+/;@*=qw(+- */);s+([$*[$-%2]])($*( $* [*/])*)+ $2 $1+gwhile$-++-1e4;y/^I()/-/d
No mas!
140.47 Stephen Turner Sun Sep 1 12:10:38 2002 
-pl sub
n{my$p=pop;bless\$p}use
overload
nomethod,sub{n$_[3]lt
9?"${$_[0]} ${$_[1]} ".pop:"-${$_[0]}"};s/-/
-/g+s/\d+/(n"$&")/g;$_=${eval()}
Minor improvements
140.53 Amir Karger Fri Sep 6 20:09:49 2002 
-pl s/\s//g;s/(?<![\d)])-/x/g;$*=qr/\(([^()]+|(??{$*}))*\)|x?\d+/;@*=qw(+- */);s#([$*[$-%2]])($*( $* [*/])*)# $2 $1#gwhile$-++<1e4;y/x()/-/d
Gotcha, Andrew! (Maybe.)
140.63 Keith Calvert Ivey Sat Sep 7 04:24:56 2002 
-p s/\s//g;map{if(!/[(\d]$/){$0=m+[*/]+?'*/':'^(';@0=(@0,pop@1)while"@1"=~/[$0]$/}/\)/?pop@1:push@{1-/\w/},$_}/(?<![\d)])-?\d+|.?/g;$_="@0
"
140.66 Alain "TopMach" Dupuis Sat Sep 7 21:59:52 2002 
-lp s/\s//g;$%=/\(/,s!\({$%}([^()]+)\){$%}!$_=$1;for$o('*/','+-'){1while s#(\S)([$o])(-?\d+( .|\d+)*)#$1 $3 $2#}$_!ewhile/(\d|[^\w ])[^\w ]/
140.75 Andrew Savige Fri Sep 6 13:02:27 2002 
-lp y~ ^I~~d;$x.=/\d/?"$_ ":($s=~s~@{[/\(/?v94:".*?(?=[@{['-+'x/[*\/]/]}(]|\$)"]}~$_ ~,$&),$s=~s~\) . ~~for/(?<![\d)])-?\d+|./g;$_=$x.$s;chop
That horse just keeps on going.
141.50 Stephen Turner Sun Sep 1 11:32:40 2002 
-pl sub
n{my$p=pop;bless\$p}use
overload
nomethod,sub{n$_[3]lt 9?"${$_[0]} ${$_[1]} $_[3]":"-${$_[0]}"};s/-/
-/g;s/\d+/(n"$&")/g;$_=${eval()}
Let Perl sort it out.

That's very cool!

141.65 Chia-chi Yeh Tue Sep 3 04:05:20 2002 
-p s/\s//g;$_="($_)";@p=qw(*/ +-);$i=0while s/\(([^(]*?)\)/$a=$1;0while$a=~s#(\S)([$p[$i]])(.*?\S)([*\/+-]|$)#$1 $3 $2$4#||++$i%2;$a/e;$_.=$/
142.56 Rick Klement Sun Sep 1 21:28:46 2002 
-p y/  //d;s//M/;s!([)\d])-!$1=!g;1while
s/\(M\)/M/|s!([^(])M(?=$|[=+)])|([*/])M(?=$|[*/)])|M(-?\d+)!push@a,$+;M!e||s/M(.)/$1M/;s/M/@a/;y/=/-/
resubmitted. That y/ //d at the beginning contains both a space and a tab. Please make sure your copy/paste is not changing it. This script passes version 3 of the test program.
142.56 Amir Karger Fri Sep 6 16:18:29 2002 
-pl s/\s//g;$z=qr/\(([^()]+|(??{$z}))*\)|-?\d+/;s#(?<=\S)([*/])($z)# $2 $1#g,s#((?<=[\d)*/])[+-])($z( $z [*/])*)# $2 $1#gwhile$q++<1e4;y/()//d
Aha! One more rung! This may be my last gasp, though.
142.58 John Williams Fri Sep 6 06:49:03 2002 
-pl $_="(@{[/\d+|\S/g]} )";s/([^\d)] -) /$1/g;1while s#\([^()]+\)#$_=$&,s= ([*/]) (\S+)=,$2,$1=||s/ ([+-]) (\S+)/,$2,$1/||y/() //d,$_#e;y/,/ /
142.80 TheodoreYoung Tue Sep 3 19:05:16 2002 
-l sub n{bless\"@_"}use overload'""',sub{${$_[0]}},neg,sub{n"-$_[0]"},nomethod,sub{n"@_[0,1,3]"};$_=<>;s/(-?)[0 ]*(\d+)/ $1 n($2)/g;print eval
Sorry for so many posts in one day. I think it is time to retire this idea.
142.82 Jasvir Nagra Sun Sep 8 03:48:07 2002 
-pl use overload'""',sub{$_[0][0]},neg,sub{bless[-"@_"]},nomethod,sub{bless["@_[$_[2],!$_[2]] ".pop]};s/-/- /g;s/0*(\d+)/bless([$1])/g;$_=eval
Oh my. I really really had given up - just happened to read perldoc overload one last time!
143.39 pom Sun Sep 1 14:35:06 2002 
-p $_="($_)";y/ ^I//d;s/(?<!\d|\))-?\d+/x$&x/g;1while
s!\(([^()]+)\)!$_=$1;1while
s-x([*/])x(\PL+)- $2 $1-||s/x([+-])x(\PL+)/ $2 $1/;$_!e;y/x//d
Thanks to mtve for the \PL trick!
143.52 Michael Thelen Wed Sep 4 17:16:56 2002 
-p s/\s//g;s@\(|(?<=[\d)])\D@$y=$&eq')'&&'\(';$z="+-"x$&=~/[(*\/]/;($s="$& $s")=~s/^(\)|([^(]))([^($z]*)$y/$2 /;$3@eg;s/$/
$s/;$_="@{[split]}
"
First stab at consolidating the various cases into a single substitution.
143.55 Andrew Savige Fri Sep 6 12:13:13 2002 
-lp y~ ^I~~d;$x.=/\d/?"$_ ":
($s=~s~@{[/\(/?v94:".*?(?=[@{['-+'x/[*\/]/]}(]|\$)"]}~$_ ~,$&),
$s=~s~\) . ~~,for/(?<![\d)])-?\d+|./g;$_=$x.$s;chop
Much cleaner now.
143.72 Kuang-che Wu Mon Sep 2 08:56:10 2002 
-lp sub e{my($i)=@_;$i--?do{e($i);eval"e $i;push\@1,'$&'"while$i?/\G[+-]/cg:/\G[*\/]/cg}:/\G-?\d+/cg?push@1,$&:/./g+e(2)+//}s/\s//g+e 2;$_="@1"
144.51 Michael Thelen Wed Sep 4 06:08:07 2002 
-p s/\s//g;s@([()])|(?<=[\d)])\D@$s="$& $s";$1?$s=~s/\)(.*?)\(//&&$1:($z="+-"x$&!~/[+-]/,$s=~s/(?<=.)[^($z]*/ /,$&)@eg;s/$/
$s/;$_="@{[split]}
"
Replaced [-+*/] with \D, that's pretty much it for now.
144.57 Amir Karger Fri Sep 6 16:08:05 2002 
-pl s/\s//g;$z=qr/\(([^()]+|(??{$z}))*\)|-?\d+/;s#(?<=\S)([*/])($z)# $2 $1#g,s#((?<=[\d)*/])[+-])($z( $z [*/])*)# $2 $1#gwhile$q++<1e4;s/[()]//g
Kill some apparently unneeded groups.
144.65 Michiel van Leeuwen Sat Sep 7 22:44:13 2002 
-p s/\s//g;s%(?<!\d|\))-?\d+|.%$_="$& ";s|\D |$b='*/'x/[^( ]/."+-"x/[)+-]/."]*)".'\( 'x/\)/;$s=~s~([ $b~"$_"x/[^) ]/~e;$1|e;$_%eg;$_.=$s;s/ $/
/
144.67 Alain "TopMach" Dupuis Sat Sep 7 21:11:54 2002 
-lp s/\s//g;$%=/\(/,s!\({$%}([^()]+)\){$%}!$_=$1;for$o('*/','+-'){1while s#([^SO])([$o])(-?\d+(O.|S\d+)*)#$1S$3O$2#}$_!ewhile/(\d|\W)\W/;y/SO/ /
145.57 John Williams Thu Sep 5 01:12:44 2002 
-pl $_="(@{[/\d+|\S/g]} )";s/(?<![\d)] )- /-/g;1while s#(\([^()]+\))#$_=$1,s= ([*/]) (\S+)=,$2,$1=||s/ ([+-]) (\S+)/,$2,$1/||y/() //d,$_#e;y/,/ /
s/(.).../$1.../ was rather silly
147.21 Markus Laire Fri Sep 6 13:49:33 2002 
-p s/\s//g;s/(?<!\d|\))-/f/g;$x=$/x2;s/$/ /,/[*\/]/&&$x=~s/[-+]? 
/ 
$_$&/
or
s/[+-] / 
$_
/,s/\(/

/,$x=~s/
/$_/for/\w+|./g;$_=$x;y/f) /- /s;s/ $//
Finally getting forward
147.57 Andrew Savige Fri Sep 6 08:02:17 2002 
-lp y~ ^I~~d;$x.=/\d/?"$_ ":!/\(/&&
($s=~s~.*?(?=[@{['-+'x/[*\/]/]}(]|$)~~,$&),
$s="$_ $s",$s=~s~-?\d+ |\)...~~for/(?<![\d)])-?\d+|./g;$_=$x.$s;chop
Getting slower and grottier. Time for a new approach.
147.61 Mtv Europe Sun Sep 1 14:02:13 2002 
-pi<([^>]+)> y/ ^I//d;s/(?<!\d|\))-?\d+/<$&>/g;s/.*/($&)/;1while
s#\(([^()]+)\)#$_=$1;for$i('*/','+-'){s/$^I([$i])$^I/<$1 $3 $2>/&&redo}$_#e;y/<>//d
147.65 Keith Calvert Ivey Sat Sep 7 04:15:21 2002 
-p s/\s//g;map{if(/[^(\d]$/){$0=/[)+-]/?'^(':'*/';@0=(@0,pop@1)while"@1"=~/[$0]$/}/\)/?pop@1:push@{1-/\w/},$_}"($_)"=~/(?<![\d)])-?\d+|./g;$_="@0
"
147.75 Kuang-che Wu Sun Sep 1 21:26:11 2002 
-lp @x=qw(*/ +-);sub e{my($i)=@_;$i--?do{e($i);eval"e $i;push\@a,'$&'"while/\G[$x[$i]]/cg}:/\G-?\d+/cg?push@a,$&:/\G./g+e(2)+//}s/\s//g;e 2;$_="@a"
148.48 Michael Thelen Wed Sep 4 04:24:48 2002 
-p s/\s//g;s@([()])|(?<=[\d)])([-+])|[*/]@$s="$& $s";$1?$s=~s/\)(.*?)\(//&&$1:($z=!$2&&"+-",$s=~s/(.)([^($z]*)/$1 /,$2)@eg;s/$/
$s/;$_="@{[split]}
"
From where I started, I never thought I would get this far, but it's amazing what a whole lot of staring can accomplish.

Looking good

148.57 Amir Karger Fri Sep 6 16:00:18 2002 
-pl s/\s//g;$z=qr/\(((?>[^()]+)|(??{$z}))*\)|-?\d+/;s#(?<=\S)([*/])($z)# $2 $1#g,s#((?<=[\d)*/])[+-])($z( $z [*/])*)# $2 $1#gwhile$q++<1e4;s/[()]//g
That's more like it! Now as long as noone else moves up at all, I can call it a respectable finish. Hm. Maybe I could hack sf.net...
149.25 Jukka Suomela Mon Sep 2 15:05:48 2002 
-p s//ac/;s/\s//g;1while s,(a.*)(c(-?\d+)|([*/])
(?=[*-/])|([-+])
(?=[-+])),$3$4$5 $1
,||s,
([^)]),$1c,||s,c.,dc,||s,d
.,
,||s,(a.*)(.),$2 $1,;s/ a//
149.50 Andrew Savige Fri Sep 6 07:16:22 2002 
-lp y~ ^I~~d;$x.=/\d/?"$_ ":
!/\(/&&($s=~s~.*?(?=[@{['-+'x/[*\/]/]}(]|$)~~,$&),
/\d/||($s="$_ $s",$s=~s~\)...~~)
for/(?<![\d)])-?\d+|./g;$_=$x.$s;chop
grotty 'next' ... gone! grotty 'substr' ... gone! ... solution still grotty :-( Oh well, at least I broke 150, which I never expected to. Mad Dutch described his stupendous 98 as "quick and dirty". I wonder, is it just plain filthy or is it *really* pornographic? I suppose Rick could take his revenge and label the person who created it as "clearly insane". :)
149.64 Chia-chi Yeh Tue Sep 3 03:33:15 2002 
-p s/\s//g;$_="($_)";@p=qw(*/ +-);$i=0while s/\(([^\(\)]*)\)/$a=$1;0while$a=~s#(\S)([$p[$i]])(.*?\S)([*\/+-]|$)#$1 $3 $2$4#||(++$i&&$i<2);$a/e;$_.=$/
Learning Perl is a nice book. ^^"
150.35 Eugene van der Pijll Thu Sep 5 16:06:27 2002 
-p s/\s+//g;s/((?<!\)|\d)-)?\d+/a$&
/g;sub w{s#
([@_])a(.*)# $2 $1#}sub
f{local
$_=pop;1while 
s#\(([^()]+)\)#a.f($1)#e||w"*/"or w"-+";s/a//;$_}$_=f$_
151.67 Alain "TopMach" Dupuis Fri Sep 6 18:26:03 2002 
-lp s/\s//g;$%=/\(/,s!\(?([^()]+)\){$%}!$_=$1;for$o('*/','+-'){1while s#(-?\d+)([$o])(-?\w+)#"$1S$3So".ord$2#e}$_!ewhile/\d\W/;y/S/ /;s/o(\d+)/chr$1/eg
151.69 Michiel van Leeuwen Fri Sep 6 22:22:41 2002 
-p s/\s//g;s%(?<!\d|\))-?\d+|.%$_="$& ";s/\) /$s=~s~(.*?)\( ~~;$1/e||s|\D |$b=" */"x/[^( ]/;$a="+-"x/[+-]/;$s=~s~[ $b$a]*~$_~;$&|e;$_%eg;$_.=$s;s/ $/
/
153.35 Rick Klement Sun Sep 1 17:31:10 2002 
-p0 s//M/;s!([-M+*/(])-(?=\d)!$1=!g;s/\s//while
s/\(M\)/M/|s!([-+*/])M(?=[-+)]|$)|([*/])M(?=[*/)]|$)|M(=?\d+)!push@a,$+;M!e||s/M(.)/$1M/
;$_="@a
";y/=/-/
to handle (-1+2)

thanks for the new test

153.48 Chris Dolan Fri Sep 6 18:13:16 2002 
-p sub
f{s/^-?\d+//?push@t,$&:s/.//+e("+-")+s///}sub
e{eval$x;while(s/^[@_]//){my$p=$&;eval$x;push@t,$p}}$x='$_[0]=~/-/?e("*/"):f';s/\s//g;e"+-";s/^/@t
/
Oops, last submission had a copy-n-paste error
153.58 Andrew Savige Fri Sep 6 00:30:42 2002 
-lp y~ ^I~~d;/\d/&&($x.="$_ ",next),
/\(/||($s=~s~.*?(?=[@{['-+'x/[*\/]/]}(]|$)~~,$x.=$&),
$s=/\)/?substr$s,2:"$_ $s"for/(?<![\d)])-?\d+|./g;$_=$x.$s;chop
Hurrah, exactly 100 strokes whittled from my first entry. Never whittled that much before. And puts me a nostril ahead of Chris Dolan. ^.^
153.69 Michael Thelen Wed Sep 4 03:32:05 2002 
-lp s/\s//g;s@([()])|(?<=[\d)])([-+])|[*/]@$".($1?($s=$1.$s)=~s/\)(.*?)\(//&&$1:($x=$&,$z=!$2&&"+-",$s=~s/^[^($z]*/$x /,$&))@eg;$_=join$","$_ $s"=~/\S+/g
Starting to cut down on the parentheses, finally!
154.43 Chris Dolan Fri Sep 6 18:11:48 2002 
-p sub
f{s/^-?\d+//?push@t,$&:s/.//+e("+-")+s///}sub
e{eval$x;while(s/^[@_]//){my$p=$&;eval$x;push@t,$p}}$x='$_[0]=~/-/?e("*/"):f';s/
\s//g;e"+-";s/^/@t
/
What a horrible hack for just one stroke... Sigh...
154.47 Chris Dolan Fri Sep 6 17:52:33 2002 
-p sub
f{s/^-?\d+//?push@t,$&:s/.//+&e+s///}sub
t{f;while(s+^[*/]++){my$m=$&;f;push@t,$m}}sub
e{t;while(s/^[+-]//){my$p=$&;t;push@t,$p}}s/\s//g;e;s/^/@t
/
just a tiebreaker improvement over my previous solution
154.67 Keith Calvert Ivey Sat Sep 7 04:08:54 2002 
-p s/\s//g;for("($_)"=~/(?<![\d)])-?\d+|./g){if(/[^(\d]$/){$0=/[)+-]/?'^(':'*/';@0=(@0,pop@1)while"@1"=~/[$0]$/}/\)/?pop@1:push@{/\w/?\@0:\@1},$_}$_="@0
"
154.71 Michiel van Leeuwen Wed Sep 4 21:32:54 2002 
-p s/\s//g;s%(?<!\d|\))-?\d+|.%$_="$& ";s/\( /$s=$_.$s;""/e;s/\) /$s=~s~(.*?)\( ~~;$1/e||s|\D |$a="+-"x/[+-]/;$s=~s~[ */$a]*~$_~;$&|e;$_%eg;$_.=$s;s/ $/
/
Negative lookback is a positive thing here
154.77 Chris Dolan Mon Sep 2 03:59:07 2002 
-lp sub f{s/^-?\d+//?push@t,$&:s/.//+&e+s///}sub t{f;while(s+^[*/]++){my$m=$&;f;push@t,$m}}sub e{t;while(s/^[+-]//){my$p=$&;t;push@t,$p}}s/\s//g;e;$_="@t"
three part recursion for expression, term and factor components
155.28 Amir Karger Fri Sep 6 15:52:35 2002 
-pl s/\s//g;
$z=qr/\((?:(?>[^()]+)|(??{$z}))*\)|-?\d+/;
s#(?<=\S)([*/])($z)# $2 $1#g,
s#((?<=[\d)*/])[+-])($z( $z [*/])*)# $2 $1#gwhile$q++<1e4;
s/[()]//g;
155.51 Andrew Savige Thu Sep 5 08:41:52 2002 
-lp y~ ^I~~d;/\d/&&($x.="$_ ",next),
/\(/||($s=~s~.*?(?=[@{['-+'x/[*\/]/]}(]|$)~~,$x.=$&),
$s=/\)/?substr$s,2:"$_ $s"for/(?<![\d)])-?\d+|\S/g;
$_=$x.$s;chop
By pure luck, I happened to be reading one of my favourite books, the shiny ball book by Joseph N Hall. He suggests on page 62, "Let's ignore white-space ... it could have been tr-ed out beforehand". Of course, "tr-ed" in golfspeak is "y-ed". ;) This simple idea just saved me 20 strokes. Amazing!

Took you long enough ;)

156.68 Keith Calvert Ivey Sat Sep 7 03:52:35 2002 
-p s/\s//g;$_="($_)";for(/(?<![\d)])-?\d+|./g){if(/[^(\d]$/){$0=/[)+-]/?'^(':'*/';@0=(@0,pop@1)while"@1"=~/[$0]$/}/\)/?pop@1:push@{/\w/?\@0:\@1},$_}$_="@0
"
156.79 Jasvir Nagra Sun Sep 8 02:01:44 2002 
-lp use overload'""',sub{$_[0][0]},neg,sub{bless[-"@_"]},map{my$z=$_;$_,sub{bless["@_[$_[2],!pop] $z"]}}qw!+ - * /!;s/-/- /g;s/0*(\d+)/bless([$1])/g;$_=eval
157.26 Jukka Suomela Sun Sep 1 23:19:50 2002 
-p s//ac/;s/\s//g;1while s,(a.*)c(-?\d+),$2 $1
,||s,(a.*)([*/](?=
[*-/])|[-+](?=
[-+])),$2 $1,||s,
([*-/]),$1c,||s,c.,dc,||s,d
.,
,||s,(a.*)(.),$2 $1,;s/ a//
157.68 Alain "TopMach" Dupuis Fri Sep 6 16:58:04 2002 
-lp s/\s//g;$q=/\(/?'':'?',s!\($q([^()]+)\)$q!$_=$1;for$o('*/','+-'){1while s#(-?\d+)([$o])(-?\w+)#"$1S$3So".ord$2#e}$_!ewhile/\d\W/;y/S/ /;s/o(\d+)/chr$1/eg
158.35 Eugene van der Pijll Thu Sep 5 16:02:07 2002 
-p s/\s+//g;s/((?<!\)|\d)-)?\d+/a$&
/g;sub w{1while s#
([@_])a(.*)# $2 $1#}sub
f{local
$_=pop;1while
s#\(([^()]+)\)#a.f($1)#e;w('*/');w('-+');y/a//d;$_}$_=f$_
159.55 John Williams Wed Sep 4 23:52:36 2002 
-pl $_="(@{[/\d+|\S/g]} )";s/(?<![\d)] )- /-/g;1while s#(\([^()]+\))#$_=$1,s=(\S+) ([*/]) (\S+)=$1,$3,$2=||s/(\S+) ([+-]) (\S+)/$1,$3,$2/||y/() //d,$_#e;y/,/ /
remove sub,local,extra\s
160.68 Michael Thelen Wed Sep 4 02:56:00 2002 
-lp s/\s//g;s@(\()|(\))|(?<=[\d)])([-+])|[*/]@$".($1?$s!~s/^/(/:$2?$s=~s/(.*?)\(//&&$1:($x=$&,$z=!$3&&"+-",$s=~s/^[^($z]*/$x /,$&))@eg;$_=join$","$_ $s"=~/\S+/g
Hooray! 12th place at long last! It's mine, all mine!
161.26 Markus Laire Fri Sep 6 13:17:17 2002 
-p s/\s//g;s/(?<!\d|\))-/f/g;$x=$/x2;s/$/ /,/[*\/]/&&$x=~s/[-+]? 
/ 
$_$&/m&&next,s/[+-] / 
$_
/,s/\( /

/,s/\)/_/,$x=~s/
/$_/mfor/\w+|./g;$_=$x;y/f_ /- /s;s/ $//
161.43 Markus Laire Sun Sep 1 10:57:42 2002 
-pl %w='*b/c+d-e'=~/./g;s/\s//g;s/(?<!\d|\))-/f/g;$_="($_)";1
while
s#\(([^()]+)\)#$a=$1;for$b('*/','+-'){1
while$a=~s/([$b])(\w+)/a$2a$w{$1}/}$a#e;y#a-f# */+-#s
Final optimizations, better logic needed.
161.70 Michiel van Leeuwen Wed Sep 4 20:13:07 2002 
-pl s/\s//g;s%((?<=[(*+/-])-|^-)?\d+|.%$_="$& ";s/\( /$s=$_.$s;""/e;s/\) /$s=~s~(.*?)\( ~~;$1/e||s|\D |$a="+-"x/[+-]/;$s=~s~[ */$a]*~$_~;$&|e;$_%eg;$_.=$s;s/ $//
162.70 Prakash Kailasa Sat Sep 7 19:42:08 2002 
-lp $t='(=[^=]+=|(?<=[*-/])-\d+|\d+)';sub r{for$o(qw|*/ -+|){1while s~$t([$o])$t~$_="$1 $3 $2";y/=//d;"=$_="~e}$_}s/\s//g;1while s~\(([^()]+)\)~$_=$1;r~e;r;y/=//d
162.78 Jasvir Nagra Fri Sep 6 02:18:28 2002 
-lp use overload'""',sub{$_[0][0]},neg,sub{bless[-$_[0][0]]},map{my$z=$_;$_,sub{bless["@_[$_[2],!$_[2]] $z"]}}qw!+ - * /!;s/-/- /g;s/0*(\d+)/bless([$1])/g;$_=eval
163.69 Prakash Kailasa Thu Sep 5 19:02:34 2002 
-lp $t='(=[^=]+=|(?<=[*-/])-\d+|\d+)';sub r{for$o(qw|*/ -+|){1while s~$t([$o])$t~$_="$1 $3 $2";y/=//d;"=$_="~e}$_}s/\s+//g;1while s~\(([^()]+)\)~$_=$1;r~e;r;y/=//d
163.69 Michael Thelen Wed Sep 4 02:44:40 2002 
-lp s/\s//g;s@(\()|(\))|(?<=[\d)])([-+])|[*/]@$".($1?$s!~s/^/(/:$2?$s=~s/(.*?)\(//&&$1:($x=$&,$z=!$3&&"+-",$s=~s/^[^($z]*/$x /,$&)).$"@eg;$_=join$","$_ $s"=~/\S+/g
Slowly inching my way toward 12th place...
163.82 TheodoreYoung Tue Sep 3 18:17:52 2002 
-l sub n{my$c=shift;bless\$c}use overload'""',sub{${$_[0]}},'neg',sub{n "-$_[0]"},'nomethod',sub{n "@_[0,1,3]"};$_=<STDIN>;s/(-?)[0\s]*(\d+)/ $1 n($2)/g;print eval
There is a perl bug here. bless shift workds on Windows 5.6.1 but not on redhat 5.6.0
164.71 Alain "TopMach" Dupuis Thu Sep 5 22:04:04 2002 
-lp $_=join' ',/\d+|\S/g;s#(^|[^)\d] )- #$1-#g;sub z{$_=pop;for$o('*/','+-'){1while s!(\S+) ([$o]) (\S+)!"$1s$3s$2"!e}$_}1while s#\( ([^(]+?) \)#z$1#e;$_=z$_;y!s! !
165.25 Jukka Suomela Sun Sep 1 23:03:40 2002 
-p s//ac/;s/\s//g;1while s,(a.*)c(-?\d+),$2 $1
,||s,(a.*)(([*/])|[-+])(
(?(3)[-+*/]|[-+])),$2 $1$4,||s,
([-+*/]),$1c,||s,c\(,dc,||s,d
\),
,||s,(a.*)(.),$2 $1,;s/ a//
165.45 tinzo (berlin.pm) Fri Sep 6 11:09:59 2002 
-pl $*='(\w+)';$/='([()+-]|^|$)';1while
s/\($*\)|\s/$+/|s#([^\d)])(-\d+)|$/$*(([*/])$*|([+-])$*$/)#"$1$3_".(push@/,$5?"$4 $7$9 $6$8":$+).$10#e;1while
s#_$*#$/[$+-1]#
Still not short enough...
165.65 Alain "TopMach" Dupuis Fri Sep 6 02:59:48 2002 
-lp $_=join' ',/\d+|\S/g;$q=/\(/?1:0,s#(\( ){$q}([^()]+)( \)){$q}#$_=$2;s!(^| \D )- !$1-!g;for$o('*/','+-'){1while s!(\S+) ([$o]) (\S+)!$1S$3S$2!}$_#ewhile/ /;y!S! !
165.68 Eugene van der Pijll Thu Sep 5 15:56:55 2002 
-lp s/\s+//g;s/((?<!\)|\d)-)?\d+/a$&z/g;sub w{1while s#z([@_])a(.*?)z# $2 $1z#}sub f{local $_=pop;1while s#\(([^()]+)\)#a.f($1).z#e;w('*/');w('-+');y/az//d;$_}$_=f$_
165.69 Alain "TopMach" Dupuis Thu Sep 5 01:38:47 2002 
-lp @a=/\d+|\S/g;$_="@a";s#(^|[(/*+-] )- #$1-#g;sub z{$_=pop;for$o('*/','+-'){1while s!(\S+) ([$o]) (\S+)!"$1s$3s$2"!e}$_}1while s#\( ([^(]+?) \)#z$1#e;$_=z$_;y!s! !
165.75 Keith Calvert Ivey Sat Sep 7 03:37:21 2002 
-p s/\s//g;$_="($_)";for(/(?<![\d)])-?\d+|./g){if(/[^(\d]$/){$r=/\)/?'^(':'*/'.'+-'x/[+-]/;push@o,pop@_ while"@_"=~/[$r]$/}/\)/?pop@_:push@{/\w/?\@o:\@_},$_}$_="@o
"
166.56 Michael Thelen Wed Sep 4 02:12:00 2002 
-p s/\s//g;s@(\()|(\))|(?<=[\d)])([-+])|[*/]@$/.($1?!($s="$& $s"):$2?$s=~s/(.*?)\(//&&$1:($x=$&,$z=!$3&&"+-",$s=~s/^[^($z]*/$x /,$&)).$/@eg;$_.="
$s";$_="@{[split]}
"
Yet a few more strokes shaved...
166.70 Prakash Kailasa Thu Sep 5 18:53:13 2002 
-lp $t='(=[^=]+=|(?<=[*-/])-\d+|\d+)';sub r{for$o(qw|*/ -+|){1while s~$t([$o])$t~$_="$1 $3 $2";y/=//d;"=$_="~e}$_}s/\s+//g;1while s~\(([^()]+)\)~$_=$1;r~e;$_=r;y/=//d
167.22 Jukka Suomela Sun Sep 1 23:00:15 2002 
-p s//ac/;s/\s//g;1while s,(a.*)c(-?\d+),$2 $1
,||s,(a.*)(([*/])|[-+])(
(?(3)[-+*/]|[-+])),$2 $1$4,||s,
([-+*/]),$1c,||s,c\(,dc,||s,d
\),
,||s,(a.*)(.)
,$2 $1
,;s/ a//
168.27 Amir Karger Fri Sep 6 15:45:44 2002 
-pl s/\s//g;
$p=qr/\((?:(?>[^()]+)|(??{$p}))*\)/;
$z=qr/$p|-?\d+/;
s#(?<=\S)([*/])($z)# $2 $1#g,
s#((?<=[\d)*/])[+-])($z( $z [*/])*)# $2 $1#gwhile($q++<1e4);
s/[()]//g;
That was a whole lot of work to go through for a measly two spots in the rankings...

For some reason this scores 174.29 when i test it. Hopefully PGAS isn't missing something.

169.66 Alain "TopMach" Dupuis Fri Sep 6 01:42:40 2002 
-lp $_=join' ',/\d+|\S/g;s#(^|[^)\d] )- #$1-#g;$q=/\(/?1:0,s#(\( ){$q}([^()]+)( \)){$q}#$t=$2;for('*/','+-'){1while$t=~s!(\S+) ([$_]) (\S+)!$1S$3S$2!}$t#ewhile/ /;y!S! !
169.71 Michiel van Leeuwen Wed Sep 4 17:34:48 2002 
-pl s/\s//g;s%((?<=[(*+/-])-|^-)?\d+|.%$_="$& ";s/\( /$s=$_.$s;""/e;s/\) /$s=~s~(.*?)\( ~~;$1/e||s|^\D |$a="+-"x/[+-]/;$s=~s~[ */$a]*~~;$s=$_.$s;$&|e;$_%eg;$_.=$s;s/ $//
2 simple optimizations gains me 9 strokes!
171.54 Mtv Europe Wed Sep 4 17:41:16 2002 
-pl BEGIN{@^H{integer,binary}=(w)x2;sub
w{bless\$_[0]}use
overload'""',sub{${$_[0]}},neg,sub{w"-$_[0]"};for
my$i(qw(+ - * /)){*{"($i"}=sub{w"@_$i"}}$^H=16e4}s/-/
-/g;$_=eval
who said dead horse? unfortunately changing ${$_[0]} to $${&w} doesn't work
171.55 Michael Thelen Wed Sep 4 01:41:08 2002 
-p s/\s//g;s@(\()|(\))|(?<=[\d)])([-+])|[*/]@$/.($1?!($s="$& $s"):$2?$s=~s/(.*?)\(//&&$1:($x=$&,$z=" */".($3&&"+-"),$s=~s/^[$z]*/$x /,$&)).$/@eg;$_.="
$s";$_="@{[split]}
"
A few strokes improvement.
172.31 Qingning Huo Fri Sep 6 18:14:49 2002 
-p sub
p{push@o,$+}y/ ^I//d;{s/(\(+)?(-?\d+)//;p$_.=$1;while(s/^\)//||/^
/){p
while
s/([^(
])$//;chop}if(s/.//){$_.=$&;p
while!/[+-][*\/]$/&&s/([^(
])(?=.$)//;redo}}$_="@o
"
172.39 Jasper McCrea Fri Sep 6 18:01:44 2002 
-lp sub
r{$_=pop;for$r(qw{?)(\+
)(&
)(\*|/}){$_="@_[0,2,1]"if@_=map
r($_),/(.*$r)(.+)/}y#&-/#a-j#;$_}s/\s//g;s#(\)|\d)-#$1&#g;1while
s#\(([^()]+)\)#r$1#e;$_=r$_;y#a-j#-'-/#
tiebreaker fiddling (what's the point) and fix the tr bit. If I had more time, I might work on a different algorithm.
172.59 Jukka Suomela Sun Sep 1 22:33:27 2002 
-pl s/^/ac/;s/\s//g;1while s,(a.*)c(-?\d+),$2 $1b,||s,(a.*)(([*/])|([-+]))(b(?(3)[-+*/]|[-+])),$2 $1$5,||s,b([-+*/]),$1c,||s,c\(,dc,||s,db\),b,||s,(a.*)(.)b,$2 $1b,;s/ ab//
172.70 Prakash Kailasa Thu Sep 5 18:40:09 2002 
-lp $t=qr~=[^=]+=|(?<=[*-/])-?\d+|^\d+~;sub r{for$o(qw|*/ -+|){1while s~($t)([$o])($t)~$_="$1 $3 $2";y/=//d;"=$_="~e}$_}s/\s+//g;1while s~\(([^()]+)\)~$_=$1;r~e;$_=r;y/=//d
173.62 Andrew Savige Thu Sep 5 04:28:48 2002 
-lp $_="  @{[/\d+|\S/g]}";s~(?<=[^\d)] -) ~~g;
/\d/&&($x.="$_ ",next),/\(/||($s=~s~.*?(?=[@{['-+'x/[*\/]/]}(]|$)~~,$x.=$&),
$s=/\)/?substr$s,2:"$_ $s"for/\S+/g;$_=$x.$s;chop
*cough* This game is improving my regex knowledge. Variable width look-behind would be nice; I wonder if Larry is going to do that for Perl 6.
173.68 Jasper McCrea Fri Sep 6 15:03:56 2002 
-lp sub r{$_=pop;for$r(qw{?)(\+ )(& )(\*|/}){$_="@_[0,2,1]"if@_=map{r($_)}/(.*$r)(.+)/}y#*-/&#a-g#;$_}s/\s//g;s#(\)|\d)-#$1&#g;1while s#\(([^()]+)\)#r$1#e;$_=r$_;y#a-g#*-/-#
this is probably my last entry. teriffic.
174.72 Alain "TopMach" Dupuis Wed Sep 4 23:15:39 2002 
-lp s/\s//g;sub z{$_=pop;s#(^|\W)-#$1_#g;for$o('*/','+-'){1while s!(\w+)([$o])(\w+)!$r=$2;$r=~y|*/+-|tdpm|;"$1s$3s$r"!e}$_}1while s#\(([^(]+?)\)#z$1#e;$_=z$_;y!mpdts_!-+/* -!
At last...
175.40 Michael Thelen Mon Sep 2 17:44:09 2002 
-p s/\s//g;s@(\()|(\))|(?<=[\d)])([-+])|[*/]@$/.($1?($s="$&
$s")&&'':$2?$s=~s/(.*?)\(//s&&$1:($x=$&,$z="
*/".($3&&"+-"),$s=~s/^[$z]*/$x
/,$&)).$/@eg;$_.="
$s";$_="@{[split]}
"
Looking better, but it still feels way too verbose.

Might be verbose, but it's working and you still have a whole week to trim it. Courage!

175.48 Mtv Europe Tue Sep 3 14:03:13 2002 
-pl BEGIN{sub
w{bless\$_[0]}use
overload
neg,sub{w"-$_[0]"},'""',sub{${$_[0]}};
for
my$i(qw(+ - * /)){*{"($i"}=sub{w"@_$i"}}@^H{integer,binary}=(\&w)x2;$^H=16e4}s/-/
-/g;$_=eval
sorry, i missed last stroke. mostly complete now.
175.67 Eugene van der Pijll Thu Sep 5 15:51:13 2002 
-lp s/\s+//g;s/((?<!\)|\d)-)?\d+/a$&z/g;sub f{local$_=pop;1while s#\(([^()]+)\)#a.f($1).z#e;1while s#z([*/])a(.*?)z# $2 $1z#;1while s#z([-+])a(.*?)z# $2 $1z#;y/az//d;$_}$_=f$_
Here are some solutions I thought up during work today. Let's see how far I can go with this.
176.29 Rick Klement Sun Sep 1 16:52:25 2002 
-p0 s//M/;s/\s+//g;s!([-M+*/])-(?=\d)!$1=!g;1while
s/M(=?\d+)/push@a,$1;M/e
|s/\(M\)/M/
||s!([-+*/])M(?=[-+)]|$)|([*/])M(?=[*/)]|$)!push@a,$+;M!e
||s/M(.)/$1M/
;$_="@a
";y/=/-/
just to get on the board, so I don't have to go through that infinite selection list :)
176.48 Mtv Europe Tue Sep 3 13:53:18 2002 
-pl BEGIN{sub
w{bless\$_[0]}use
overload
neg,sub{w"-$_[0]"},'""',sub{${$_[0]}};
for
my$i(qw(+ - * /)){*{"($i"}=sub{w"@_$i"}}@^H{integer,binary}=(\&w)x2;$^H=152e3}s/-/
-/g;$_=eval
*twitch*
176.64 Jonathan Wang Wed Sep 4 16:24:22 2002 
$_=<>;$_="($_)";y/ \t//d;s/\d+/ $& /g;s#([^ )])([+-]) #$1 $2#g;while(s#\(([^()]+)\)#$_=$1;foreach$a('/*','+-'){while(s! (\S+) ([$a]) (\S+) ! $1s$3s$2 !){}}$_#e){}y/s / /d;print
177.37 Qingning Huo Fri Sep 6 17:01:09 2002 
-p sub
p{push@o,$+}y!-+/* ^I!bBaA!d;{s/(\(+)?(b?\d+)//;p$_.=$1;while(s/^\)//||/^
/){p
while
s/(\w)$//;chop}if(s/.//){$x=$&;p
while
s/(a|$x)$//i;$_.=$x;redo}}$_="@o
";y!bBaA!-+/*!
177.58 Jed Davis Tue Sep 3 18:40:11 2002 
-l sub n{my($x)=@_;bless\$x,'main'}use
overload(map{my$o=$_;$o,sub{n($_[2]?"$o$_[0]":"$_[0] $_[1] $o")}}qw{+
- * /}),'""',sub{${$_[0]}};$_=<>;s/-/- /g;s/\d+/n("$&")/g;print
eval
Never underestimate the power of code reuse.
177.63 Jonathan Wang Wed Sep 4 15:45:07 2002 
$_=<>;$_="($_)";y/ \t//d;s/\d+/ $& /g;s#([*/+-])([+-]) #$1 $2#g;while(s#\(([^()]+)\)#$_=$1;foreach$a('/*','+-'){while(s! (\S+) ([$a]) (\S+) ! $1s$3s$2 !){}}$_#e){}y/s / /d;print
178.47 Mtv Europe Tue Sep 3 12:27:45 2002 
-pl BEGIN{sub
w{bless\$_[0]}use
overload
neg,sub{w"-$_[0]"},'""',sub{${$_[0]}};
for
my$i(qw(+ - * /)){*{"($i"}=sub{w"@_$i"}}@^H{integer,binary}=(\&w)x2;$^H|=151552}s/-/
-/g;$_=eval
is it shorterst overload?
178.63 Andrew Savige Thu Sep 5 03:35:35 2002 
-lp $_="  @{[/\d+|\S/g]}";s~(?<=^  -|[^\d)] -) ~~g;
/\d/&&($x.="$_ ",next),/\(/||($s=~s~.*?(?=[@{['-+'x/[*\/]/]}(]|$)~~,$x.=$&),
$s=/\)/?substr$s,2:"$_ $s"for split;$_=$x.$s;chop
Just a pathetic regex tweak. I suck.
178.69 Prakash Kailasa Thu Sep 5 18:30:03 2002 
-lp $t=qr~=[^=]+=|(?<=[*-/])-?\d+|^\d+~;sub r{$_=$1;for$o(qw|*/ -+|){1while s~($t)([$o])($t)~$_="$1 $3 $2";y/=//d;"=$_="~e}$_}s/\s+//g;1while s~\(([^()]+)\)~r~e;s~(.*)~r~e;y/=//d
178.70 Michiel van Leeuwen Wed Sep 4 17:24:54 2002 
-pl s/\s//g;s%((?<=[(*+/-])-|^-)?\d+|.%$_="$& ";s/\( /$s=$_.$s;""/e;s/\) /$s=~s~(.*?)\( ~~;$1/e||s|[*+/-] |$a="+-"x/[+-]/;$s=~s~[ */$a]*~~;$s=$_.$s;$&|e;$b.=$_%eg;$_=$b.$s;s/ $//
More regex'es
179.69 Jonathan Wang Wed Sep 4 15:39:45 2002 
$_=<>;$_="($_)";y/ \t//d;s/\d+/m$&m/g;s#([*/+-])([+-])m#$1m$2#g;while(s#\(([^()]+)\)#$_=$1;foreach$a('/*','+-'){while(s!m([^m]+)m([$a])m([^m]+)m!m$1 $3 $2m!){}}$_#e){}s/m//g;print
179.72 Amir Karger Fri Sep 6 05:09:46 2002 
-pl @a=qw(+- */);s/\s//g;do{push@c,$o[$l]||();($l%3>1&&(s/^\)//||s/^-?\d+/$f=push@c,$&;""/e))||($l%3>1&&s/^\(//||!$f||($o[$l]=s#^[$a[$l%3]]##&&$&))&&($l+=2,$f=0)}while$l--;s/$/@c/
A last (pathetic) gasp?
181.40 tinzo (berlin.pm) Thu Sep 5 23:18:43 2002 
-pl $w='(\w+)';$/='([()+-]|^|$)';$e='push@_,$3?"$2 $5$7 $4$6":$2;"$1_$#_$8"';1
while
s/(\($w\)|\s)/$2/|s/([^\d)])(-\d+)/$e/ee|s#$/$w(([*/])$w|([+-])$w$/)#$e#ee;1
while
s/_$w/$_[$+]/
181.69 Jonathan Wang Wed Sep 4 14:36:12 2002 
$_=<>;$_="($_)";s/[ \t]//g;s/\d+/=$&=/g;s#([*/+-])([+-])=#$1=$2#g;while(s#\(([^()]+)\)#$_=$1;foreach$a('/*','+-'){while(s!=([^=]+)=([$a])=([^=]+)=!=$1 $3 $2=!){}}$_#e){}s/=//g;print
182.57 Andrew Savige Wed Sep 4 12:13:48 2002 
-lp $_="@{[/\d+|\S/g]}";s~(?<=[^\d)] -) ~~g;s~^- ~-~;
/\d/&&($x.=$_.$",next),$:='-+'x/[*\/]/,
/\(/||($s=~s~.*?(?=[$:(]|$)~~,$x.=$&),
$s=/\)/?substr$s,2:"$_ $s"for split;$_=$x.$s;chop
Come on Keith!
182.69 Jonathan Wang Wed Sep 4 08:57:05 2002 
$_=<>;$_="($_)";s/[ \t]//g;s/\d+/=$&=/g;s#([*/+-])([+-])=#$1=$2#g;while(s#\(([^()]+)\)#$_=$1;foreach$a('/*','+-'){while(s!=([^=]+)=([$a])=([^=]+)=!=$1 $3 $2=!){}}$_#e){}s/=//g;print;
182.73 Amir Karger Fri Sep 6 04:07:13 2002 
-pl @a=qw(+- */);s/\s//g;do{push@c,$o[$l]||();if($l%3>1&&s/^((\))|-?\d+)//){$f=push@c,$1if!$2}elsif($l%3>1&& s/^\(//||!$f||($o[$l]=s#^[$a[$l%3]]##&&$&)){$l+=2;$f=0}}while$l--;s/$/@c/
Still don't know why my 159 solution failed...
183.58 John Williams Tue Sep 3 20:13:56 2002 
-pl $_="( @{[m=(\d+|\S)\s*=g]} )";s/(?<![\d)] )- /-/g;sub i{local$_=pop;s=(\S+) ([*/]) (\S+)=$1,$3,$2=||s/(\S+) ([+-]) (\S+)/$1,$3,$2/||y/() //d;$_}while(s/(\([^()]+\))/i$1/e){}y/,/ /
183.60 Jukka Suomela Sun Sep 1 22:24:02 2002 
-pl s/^/ac/;s/\s//g;1while s,(a.*)c(-?\d+),$2 $1b,||s,(a.*)([*/])(b[*/]),$2 $1$3,||s,(a.*)([-+*/])(b[-+]),$2 $1$3,||s,b([-+*/]),$1c,||s,c\(,dc,||s,db\),b,||s,(a.*)(.)b,$2 $1b,;s/ ab//
183.67 Jonathan Wang Wed Sep 4 08:50:26 2002 
$_=<>;s/.+/($&)/;s/[ \t]//g;s/\d+/=$&=/g;s#([*/+-])([+-])=#$1=$2#g;while(s#\(([^()]+)\)#$_=$1;foreach$a('/*','+-'){while(s!=([^=]+)=([$a])=([^=]+)=!=$1 $3 $2=!){}}$_#e){}s/=//g;print;
First time

good luck!

184.66 Jasper McCrea Thu Sep 5 17:29:39 2002 
-lp sub r{$_=pop;for$r(qw{?)(\+ )(- )(\*|/}){$_=r($_[0]).$".r(pop).$".pop if@_=/(.*$r)(.*)/}y#-/*+#a-d#;$_}s/\s//g;s#(^|[^0-9)])-#$1~#g;1while s#\(([^()]+)\)#r$1#e;$_=r$_;y#a-d~#-/*+-#
did somebody mention a dead horse?
185.69 Prakash Kailasa Thu Sep 5 16:06:07 2002 
-lp $t=qr~=[^=]+=|(?<=[*-/])-?\d+|^\d+~;sub r{$_=$1;for$o(qw|*/ -+|){1while s~($t)([$o])($t)~$_="$1 $3 $2";y/=//d;"=$_="~e}$_}s/\s+//g;1while s~\(([^()]+)\)~r~e||s~($t(.$t)+)~r~e;y/=//d
187.71 Michiel van Leeuwen Tue Sep 3 21:39:07 2002 
-pl s/\s//g;$_=~s%((?<=[(*+/-])-|^-)?\d+|.%$_="$& ";/\d/?$b.=$_:(/\(/?$s=$_.$s:(/\)/?do{$s=~s/(.*?)\( //;$b.=$1}:do{$a="+-"x/[+-]/;$s=~s/[ *\/$a]*//;$b.=$&;$s=$_.$s}))%eg;$_=$b.$s;s/\s$//
That $a did the trick :-)
189.65 Jasper McCrea Thu Sep 5 14:37:59 2002 
-lp sub r{$_=pop;for$r(qw{+?)(\+ *)(- *)(\*|/}){if(@_=/(.$r)(.*)/){$_=r($_[0]).$".r(pop).$".pop}}y#-/*+#a-d#;$_}s/\s//g;s#(^|[(*/+-])-#$1~#g;1while s#\(([^()]+)\)#r$1#e;$_=r$_;y#a-d~#-/*+-#
I'd like to say this was my first solution. ( can't get that damn ))+(( soltion to work at all! )
190.58 John Williams Tue Sep 3 20:11:50 2002 
-pl $_="( @{[m=(\d+|\S)\s*=g]} )";s/(?<![\d)] )- /-/g;sub i{local$_=pop;s=(\S+) ([*/]) (\S+)=$1,$3,$2=||s/(\S+) ([+-]) (\S+)/$1,$3,$2/||s/. (\S+) ./$1/;$_}while(s/(\([^()]+\))/i$1/e){}y/,/ /
190.76 Keith Calvert Ivey Fri Sep 6 02:36:04 2002 
-p s/\s//g;$_="($_)";for(/(?<![\d)])-?\d+|./g){if(/\)/){@o=(@o,$_=pop@_)until/\(/;pop@o;next}if(/[^(\d]$/){$r='*/'.'+-'x/[+-]/;@o=(@o,pop@_)while"@_"=~/[$r]$/}push@{/\w/?\@o:\@_},$_}$_="@o
"
Still on the wrong path, but submitting this to at least pass Qingning Huo.
191.45 Qingning Huo Thu Sep 5 18:33:25 2002 
-p y/ ^I//d;my@o;{s/^(\(+)?(-?\d+)//;$_.=$1;push@o,$2;while(s/^\)//||/^
/){push@o,$&while
s#[^(
]$##;chop}if(s/^.//){$_.=$&;push@o,$&while
s#[-+*/](?=[-+]$)##||s#[*/](?=[*/]$)##;redo}}$_="@o
"
191.59 Andrew Savige Wed Sep 4 07:56:24 2002 
-lp $_="@{[/\d+|\S/g]}";s~(?<=[^\d)] -) ~~g;s~^- ~-~;
$x.="$_ "x/\d/,$s=/\(/.$s,/[\d(]/&&next,
$:='-+'x/[*\/]/,$s=~s~.*?(?=[$:1]|$)~~,$x.=$&,
$s=/\)/?substr$s,1:"$_ $s"for split;$_=$x.$s;chop
I am shocked to go sailing gaily path Keith.
192.39 Michael Thelen Mon Sep 2 16:55:48 2002 
-p s/\s//g;s/(?<![\d)])-/~/g;s@(\()|(\))|([-+])|([*/])@$/.($1?($s="$&
$s")&&'':$2?$s=~s/(.*?)\(//s&&$1:($x=$&,$z="
*/".($3&&"+-"),$s=~s/^[$z]*/$x
/,$&)).$/@eg;$_.="
$s";$_="@{[split]}
";y/~/-/
193.48 Markus Laire Sun Sep 1 10:21:01 2002 
-pl %w=split//,'*b/c+d-e';sub r{my($a)=@_;
for$b('*/','+-'){1 while$a=~s#(\w+)([$b])(\w+)#a$1a$3a$w{$2}#}$a}
s/\s//g;s/(?<!\d|\))-/f/g;1 while s/\(([^()]+)\)/r$1/e;
$_=r$_;y#a-f# */+-#s;s/^\s//
Easy optimizations, score <200
194.68 Prakash Kailasa Wed Sep 4 21:16:10 2002 
-lp sub r{$_=$1;for$o(qw|*/ -+|){1while s~($t)([$o])($t)~$_="$1 $3 $2";y/{}//d;"{$_}"~e}$_}$t=qr~(?:^|(?<=[-+*/]))-?\d+|{[^}]+}~;s/\s+//g;1while s~\(([^()]+)\)~r~e||s~($t([-+*/]$t)+)~r~e;y/{}//d
195.42 Qingning Huo Thu Sep 5 18:24:09 2002 
-p y/ ^I//d;my@o;{s/^(\(+)?(-?\d+)//;$_.=$1;push@o,$2;while(s/^\)//||/^
/){push@o,$&while
s#[^(
]$##;chop}if(s/^.//){$_.=$&;push@o,$1while
s#([-+*/])([-+])$#$2#||s#([*/])([*/])$#$2#;redo}}$_="@o
"
195.75 Keith Calvert Ivey Sun Sep 1 21:57:06 2002 
-p s/\s//g;$_="($_)";for(/(?<![\d)])-?\d+|./g){if(/\)/){@o=(@o,$_=pop@_)until/\(/;pop@o;next}if(/[*\/+-]$/){$r='*/'.'+-'x$&=~/[+-]/;@o=(@o,pop@_)while"@_"=~/[$r]$/}push@{/\w/?\@o:\@_},$_}$_="@o
"
197.43 Qingning Huo Thu Sep 5 18:11:32 2002 
-p y/ ^I//d;my@o;while(s/^(\(+)?(-?\d+)//){$_.=$1;push@o,$2;while(s/^\)//||/^
/){push@o,$&while
s#[^(
]$##;chop}s/^.//||last;$_.=$&;push@o,$1while
s#([-+*/])([-+])$#$2#||s#([*/])([*/])$#$2#}$_="@o
"
197.78 Andrew Savige Wed Sep 4 04:52:42 2002 
-lp $_="@{[/\d+|\S/g]}";s~(?<=[^\d)] -) ~~g;s~^- ~-~;map{$x.="$_ "x/\d/;$s=/\(/.$s;$s=~s~(.*?)1~~,$x.=$1if/\)/;$s=~s~.*?(?=[@{['-+'x/[*\/]/]}1]|$)~~,$x.=$&,$s="$_ $s"if!/[\d()]/}split;$_=$x.$s;chop
To my great surprise and delight, I have almost caught Keith! I am sure he will spurt away from me, however.
198.55 John Williams Tue Sep 3 16:35:50 2002 
-pl @x=m=(\d+|[()+/*-])\s*=g;$_="( @x )";sub i{local$_=pop;s=(\S+) ([*/]) (\S+)=$1,$3,$2=||s/(\S+) ([+-]) (\S+)/$1,$3,$2/||s/. (\S+) ./$1/;$_}s/(?<![0-9)] )- /-/g;while(s/(\([^()]+\))/i$1/e){}y/,/ /
Under 200!
198.68 Prakash Kailasa Wed Sep 4 20:20:57 2002 
-lp sub r{$_=$1;for$o(qw|*/ -+|){1while s~($t)([$o])($t)~y/{}//dfor@x=($1,$3,$2);"{@x}"~e};$_}$t=qr~(?:^|(?<=[-+*/]))-?\d+|{[^}]+}~;s/\s+//g;1 while s~\(([^()]+)\)~r~e||s~($t([-+*/]$t)+)~r~e;y/{}//d
200.77 Keith Calvert Ivey Sun Sep 1 20:52:40 2002 
-p s/\s//g;for(/(?<![\d)])-?\d+|./g){if(/\)/){@o=(@o,$_=pop@_)until/\(/;pop@o;next}if(/[*\/+-]$/){$r='*/'.'+-'x$&=~/[+-]/;@o=(@o,pop@_)while"@_"=~/[$r]$/}push@{/\w/?\@o:\@_},$_}$_="@{[@o,reverse@_]}
"
Bleah! Still disgusting.
203.63 Eike Grote Thu Sep 5 20:18:03 2002 
-lp s/\s//g;s/(?<![)\d])-/y/g;$_="($_)";while(/\(/){eval($w='$z[$=++]="$2 $4 $3"while s|(\([^()]*?)(\w+)([*/])(\w+)([^()]*\))|$1x$=$5|');$w=~s|\*/|-+|;eval$w;s/\((\w+)\)/$1/}s/x$=/pop@z/e&y/y/-/while--$=
205.69 Prakash Kailasa Wed Sep 4 13:26:11 2002 
-lp sub n{y/{}//dfor@x=($1,$3,$2);"{@x}"}sub r{$_=$1;for$o(qw|*/ -+|){1while s~($t)([$o])($t)~n~e};$_}$t=qr~(?:^|(?<=[-+*/]))-?\d+|{[^}]+}~;s/\s+//g;1while s~\(([^()]+)\)~r~e||s~($t([-+*/]$t)+)~r~e;y/{}//d
Just double the leaders' score. Should be a reasonable starting point ;-)
205.77 Michael Thelen Sun Sep 1 21:17:13 2002 
-ln s/\s//g;/\d/?push@g,$_:/\)/?($f=pop@s)ne'('&&push(@g,$f)&&redo:(/([+-])|[*\/]/&&($z="*/".($1&&"+-"),"@s"=~/[$z]$/)&&push(@g,pop@s)&&redo,push@s,$_)for/(?:(?<![\d)])-)?\d+|./g;push@g,reverse@s;print"@g"
206.56 Greg Allen Fri Sep 6 02:44:17 2002 
-pl sub a{$_=$1;0while s/\(([^()]+)\)/$0[++$-]=&a;"%$-"/e;s&([^+^I]*\d)([*/])([^+^I]+)&a(" $3 $2")&eg;$_=a(" $' $2")if/^(.+)([+^I])/;"$_@_";}s/\s//g+s/(\d|\))-/$1^I/g+s/(.*)/a/e;s/%(\d+)/$0[$1]/ewhile/%/+y/^I/-/
sub 200 would be tolerable
206.70 me ELOISE Sat Sep 7 03:38:31 2002 
-l chomp($_=<>);$_="($_)";1while s!\(\s*([^(]+?)\)!$a=$1;map{1while$a=~s#((^|[*/+-]\s*)(-))?\s*(\w+)\s*([$_])\s*(-?)\s*(\w+)\s*#$h{A.++$i}="$3$4 $6$7 $5";"$2A$i"#e}'*/','-+';$a!e;1while s|A\d+|$h{$&}|;print
207.57 Greg Allen Fri Sep 6 02:22:49 2002 
-pl sub a{$_=$1;0while s/\(([^()]+)\)/$0[++$-]=&a;"%$-"/e;s&([^+^I]*\d)([*/])([-%\d]+)&a(" $3 $2")&eg;$_=a(" $' $2")if/^(.+)([+^I])/;"$_@_";}s/\s//g+s/(\d|\))-/$1^I/g+s/(.*)/a/e;s/%(\d+)/$0[$1]/ewhile/%/+y/^I/-/
sub 200 would be tolerable
207.58 Andrew Savige Wed Sep 4 00:25:17 2002 
-lp $_="@{[/\d+|\S/g]}";s;(?<=[^\d)] -) ;;g;s;^- ;-;;map{
$x.="$_ "x/\d/;$s=/\(/.$s;$s=~s/(.*?)1//,$x.=$1if/\)/;
if(/[-+*\/]$/){$s=~s/..//,$x.=$&while$s&&$s<1&$s=~/^[-+]/<=//;$s="$_ $s"}
}split;$_=$x.$s;chop
Still flogging the same horse. Just one dirty trick after another (the latest being the default regex). I still think Mtv's Lemma holds the key to this game. Like Rick's Troppo Sort game, I suspect (from the leaderboard turnaround times) that the leaders are using a much slower algorithm than mine.
209.69 nashdj Sun Sep 1 10:37:43 2002 
-l $e=qr{(??{$t})(([+-])(??{$t})(?{push@_,$2}))*};$t=qr{(??{$f})(([*/])(??{$f})(?{push@_,$2}))*};$f=qr{\((??{$e})\)|((-|)\d+)(?{push@_,$1})};$_=<>;s/ |\t//g;1while(s/\(\(([^)])\)\)/($1)/);$_=~$e;$,=' ';print@_
temporary HACK to avoid segfault on linux
210.77 Ala Qumsieh Tue Sep 3 21:13:23 2002 
-pl s/\s//g;%p=qw|* 2 / 2 + 1 - 1|;/\d/?push@_,$_:/\(/?push@@,$_:/\)/?do{push@_,$*while'('ne($*=pop@@)}:do{push@_,pop@@while$p{$@[-1]}>=$p{$_};push@@,$_}for/(?<=[\d)])[*\/+-]|-?\d+|./g;@_=(@_,reverse@@);$_="@_"
Next submission on a different horse. Promise!
213.53 John Williams Tue Sep 3 07:23:02 2002 
-pl @x=m=(\d+|[()+/*-])\s*=g;$_="( @x )";sub i{local$_=shift;s=(\S+) ([*/]) (\S+)=$1,$3,$2=||s/(\S+) ([+-]) (\S+)/$1,$3,$2/;"( $_ )"}s/(?<![0-9)] )- /-/g;while(/ /){s/\( (\S+) \)/$1/||s/\( ([^()]+) \)/i$1/e}y/,/ /
test 44 took over an hour before I eliminated $& from my code
213.70 Michiel van Leeuwen Tue Sep 3 21:29:43 2002 
-pl s/\s//g;$_=~s%((?<=[(*+/-])-|^-)?\d+|.%$_="$& ";/\d/?$b.=$_:(/\(/?$s=$_.$s:(/\)/?do{$s=~s/(.*?)\( //;$b.=$1}:do{/[+-]/?do{$s=~s/[ *\/+-]*//;$b.=$&}:do{$s=~s/[ *\/]*//;$b.=$&};$s=$_.$s}))%eg;$_="$b$s";s/\s$//
Aargh. those egative integers... And why took I so long before using google?
214.46 Qingning Huo Thu Sep 5 18:02:54 2002 
-p y/ ^I//d;my@o;while(s/^(\(+)?(-?\d+)//){$_.=$1;push@o,$2;while(s/^\)//){push@o,$&while
s#[^(
]$##;chop}s/^.?//||last;$_.=$&;push@o,$1while
s#([-+*/])([-+])$#$2#||s#([*/])([*/])$#$2#}push@o,$&while
s/.$//;$_="@o
"
214.82 Jasvir Nagra Thu Sep 5 22:23:59 2002 
-lp sub r{bless["@_[$_[2],!$_[2]] $_[3]"]};use overload'+'=>sub{r@_,'+'},'-'=>sub{r@_,'-'},'*'=>sub{r@_,'*'},'/'=>sub{r@_,'/'},'""'=>sub{$_[0][0]},neg=>sub{bless[-$_[0][0]]};s/0*(\d+)/bless([$1])/g;s/-/ -/g;$_=eval
216.69 me ELOISE Tue Sep 3 03:48:29 2002 
-l chomp($_=<>);$_="($_)";1while s!\(\s*([^(]+?)\)!$a=$1;map{1while$a=~s#((^|[*/+-]\s*)(-))?\s*([A0-9]+)\s*([$_])\s*(-?)\s*([A0-9]+)\s*#$h{A.++$i}="$3$4 $6$7 $5";$2."A$i"#e}'*/','-+';$a!e;1while s|A\d+|$h{$&}|e;print
216.79 Ala Qumsieh Tue Sep 3 20:08:25 2002 
-pl %p=qw|* 2 / 2 + 1 - 1|;s/\s//g;/\d/?push@_,$_:/\(/?push@@,$_:/\)/?do{push@_,$*while@@&&'('ne($*=pop@@)}:do{push@_,pop@@while$p{$@[-1]}>=$p{$_};push@@,$_}for/(?<=[\d)])[*\/+-]|-?\d+|./g;push@_,pop@@while@@;$_="@_"
dead horse, I know. At least I'm ahead of /-\ ;)
218.57 Greg Allen Fri Sep 6 00:56:10 2002 
-pl sub a{$_=$1;0while s/\(([^()]+)\)/$0[++$-]=&a;"%$-"/e;s&([^+^I]+)([*/])([-%\d]+)&$0[++$-]=a(" $3 $2");"%$-"&eg;$_=a(" $' $2")if/(.+)([+^I])/;"$_@_"}s/\s//g+s/(\d|\))-/$1^I/g+s/(.*)/a/e;s/%(\d+)/$0[$1]/ewhile/%/|y/^I/-/
I see the light
218.60 Andrew Savige Tue Sep 3 13:25:09 2002 
-lp $_="@{[/\d+|\S/g]}";s;(?<=[^\d)] -) ;;g;s;^- ;-;;map{
$x.="$_ "x/\d/;$s=Y x/\(/.$s;$s=~s/(.*?)Y//,$x.=$1if/\)/;
if(/[-+*\/]$/){$s=~s/..//,$x.=$&while$s&&$s=~/^[-+]/<=/[-+]/&$s!~/^Y/;$s="$_ $s"}
}split;$_=$x.$s;chop
*Cough* there were a lot of blunders in my earlier efforts. And plenty more remain. Kept my original algorithm, just corrected some blunders and changed from array-based to string-based.
220.68 me ELOISE Tue Sep 3 03:21:58 2002 
-l chomp($_=<>);$_="($_)";1while s!\(\s*([^(]+?)\)!$a=$1;for$b('*/','-+'){1while $a=~s#((^|[*/+-]\s*)(-))?\s*([A0-9]+)\s*([$b])\s*(-?)\s*([A0-9]+)\s*#$h{A.++$i}="$3$4 $6$7 $5";$2."A$i"#e}$a!e;1while s|A\d+|$h{$&}|e;print
222.58 Greg Allen Fri Sep 6 00:03:35 2002 
-pl sub a{$_=$1;0while s/\(([^()]+)\)/$0[++$-]=&a;"%$-"/e;s&([^+~]+)([*/])([-%\d]+)&$0[++$-]=a(" $3 $2");"%$-"&eg;s/(.+)([+~])(.+)/a(" $3 $2")/e;$_.="@_"}s/\s//g;s/(\d|\))-/$1~/g;s/(.*)/a/e;s/%(\d+)/$0[$1]/ewhile/%/|y/~/-/
nearly down to double figures :)
224.79 Ala Qumsieh Tue Sep 3 16:08:26 2002 
-pl %p=qw|* 2 / 2 + 1 - 1|;s/\s//g;map{/\d/?push@_,$_:/\(/?push@@,$_:/\)/?do{push@_,$*while@@&&'('ne($*=pop@@)}:do{push@_,pop@@while$p{$@[-1]}>=$p{$_};push@@,$_}}m:\b[*/+-]|(?<=\))[*/+-]|-?\d+|.:g;push@_,pop@@while@@;$_="@_"
As usual, start with the brute force approach.
225.72 Amir Karger Fri Sep 6 02:35:48 2002 
-pl @a=qw(+- */);
s/\s//g;
do{push@c,$o[$l]||();if($l%3==2&&s/^(\)|(-?\d+))//){$l--;if(defined$2){push@c,$2;$found=1}}elsif($l%3==2&&s/^\(//||!$found||($o[$l]=s#^[$a[$l%3]]##&&$&)){$l++;$found=0}else{$l--}}until$l==-1;s/$/@c/
New algorithm. Hopefully this one will work on your computers too!
226.53 Qingning Huo Thu Sep 5 16:06:42 2002 
-p y/ ^I//d;my@o;while($.&&s/^-?\d+//||s/^.//){if($.=!($&=~y/0-9)//)){$_.=$&;push@o,$1while
s#([-+*/])([-+])$#$2#||s#([*/])([*/])$#$2#}elsif($&eq')'){push@o,$&while
s#[^(
]$##;chop}else{push@o,$&}}push@o,$&while s/.$//;$_="@o
"
227.80 Mtv Europe Sun Sep 1 14:45:27 2002 
-pl BEGIN{$OVERLOAD{d}++;*{'()'}=sub{};eval"*{'($_'}=sub{bless\\\"\@_$_\",main}"for
qw(+ - * /);*{'(neg'}=sub{bless\"-$_[0]",main};*{'(""'}=sub{${$_[0]}};$^H{integer}=$^H{binary}=sub{bless\$_[0],main};$^H|=151552}s/--/-
-/;$_=eval
abuse of perl

hmm .. another (ab)use of overload

233.70 me ELOISE Tue Sep 3 02:26:15 2002 
-l chomp($_=<>);$_="($_)";1while s!\(([^(]+?)\)!$a=$1;$a=~s#\s*((^|[*/+-])\s*)-\s*#$1=#g;for$b('*/','-+'){1while$a=~s#\s*(=?[A0-9]+)\s*([$b])\s*(=?[A0-9]+)\s*#$h{A.++$i}="$1 $3 $2";"A$i"#e}$a!xe;1while s|(A\d+)|$h{$1}|e;s/=/-/g;print
at least I'm getting near the ballpark
244.60 Andrew Savige Tue Sep 3 07:02:19 2002 
-n $_="@{[/\d+|\S/g]}";s;(?<=[^\d)] -) ;;g;s;^- ;-;;map{
/\d/?push@o,$_:/\(/&&unshift@~,Y;if(/\)/){push@o,$:while($:=shift@~)!~Y}
if(/[-+*\/]$/){push@o,shift@~while@~&&$~[0]=~/[-+]/<=/[-+]/&$~[0]!~Y;@~=($_,@~)}
}/-?\d+|\S/g;push@o,@~;print"@o
"
Dirty trick: changed '(' token to 'Y'. Since the actual letter is arbitrary, I naturally chose 'Y' to honour our esteemed referee, Dr `/anick Shampoo.
253.50 Andrew Savige Tue Sep 3 00:43:36 2002 
-n $_="@{[/\d+|\S/g]}";s;([^\d)]) - ;$+ -;g;s;^- ;-;;
map{/\d/?push@o,$_:/\(/&&unshift@~,$_;
if(/\)/){push@o,$&while+shift(@~)=~/[^(]/}
if(/[-+*\/]$/){push@o,shift@~while@~&&$~[0]=~/[-+]/<=/[-+]/&$~[0]ne'(';@~=($_,@~)}
}/-?\d+|\S/g;
push@o,@~;print"@o
"
Interesting problem (kudos to Juho), clumsy opening bid. I think this is the first time I've ever used unshift in a golf game (Ton would never approve ;) -- a sure sign something is wrong. I wonder what The Aliens are doing? I expect they're sitting around the campfire sucking hard on the Regex crack-pipe chanting "Mtv's Lemma, Mtv's Lemma, Mtv's Lemma ...".

253, and it's not signed André? Oh my. (okay, I know it's not right to use of my ref status to gloat, but then again, power corrupts and all that :) )

266.69 Greg Allen Thu Sep 5 11:13:26 2002 
-pl sub t{$_=pop;s/\((.*)\)/a($1)/e;"$_@_"}$b=qr/\((?:(?>[^()]+)|(??{$b}))*\)/;$p=qr/$b|-?\d+/;sub p{$_=shift;/(\*|\/)($p)$/?p("$`",'',t(" $1@_",$2)):t@_,$_}sub a{$_=shift;/[\d)]([-+]($p(?:[*\/]$p)*))+$/&&/(.)(\Q$+\E)$/?a("$`",'',p($2," $1@_")):p$_,@_}y/ ^I//d;$_=a$_
guess this approach won't get too close to 104 then...
267.25 Honza Pazdziora Fri Sep 6 14:09:30 2002 
-n close STDERR;
s!(^|[(*/+-])\s*(-?)\s*(\d+)!$1 d("$2$3")!g;
$_=`$^X -MO=Deparse,-p -e'$_'`;
print map({
if (/\d/) {
        ' 'x($i++>0).$_;
} elsif (/\)/) {
        $i++;shift@a;
} else {
        /\(/?(unshift@a,''):($a[0]=" $_");();
} 
}/-?\d+|[-*+\/()]/g),@a,'
'
The question is whether your machine won't break on the long -e option. Mine doesn't.
276.71 Greg Allen Wed Sep 4 17:08:15 2002 
-pl sub t{$_=pop;/^\((.*)\)$/?a($1,@_):"$_@_"}$b=qr{\((?:(?>[^()]+)|(??{$b}))*\)};$p=qr{$b|-?\d+};sub p{$_=shift;/(\*|\/)($p)$/?p("$`",'',t(" $1@_",$2)):t@_,$_}sub a{$_=shift;/(?<=[\d)])(([-+])($p(?:[*\/]$p)*))+$/&&/(.)(\Q$+\E)$/?a("$`",'',p($2," $1@_")):p$_,@_}s/\s//g;$_=a$_
parsing by regex still sucks
278.34 Markus Laire Sun Sep 1 09:43:33 2002 
-p %w=('+'=>'a','-'=>'s','*'=>'M','/'=>'D');
sub rpn($) {
my($a)=@_;
1 while$a=~s#(\w+)([*/])(\w+)#_$1_$3_$w{$2}#;
1 while$a=~s/(\w+)([+-])(\w+)/_$1_$3_$w{$2}/;
$a
}
s/\s//g;s/(?<!\d|\))-/n/g;
1 while s/\(([^()]+)\)/rpn$1/e;
$_=rpn$_;

y#saMDn#-+*/-#;
y#_# #s;
s/^\s//;
$_.="\n";
First working solution, so I don't bother with optimization yet.
289.71 Greg Allen Wed Sep 4 16:45:19 2002 
-pl sub t{$_=shift;/^\((.*)\)$/?a($1,@_):"$_@_"}$b=qr{\((?:(?>[^()]+)|(??{$b}))*\)};sub p{$_=shift;/(\*|\/)($b|-?\d+)$/?p("$`",'',t($2," $1@_")):t $_,@_}$p=qr{(?:$b|-?\d+)};sub a{$_=shift;/(?<=[\d)])(([-+])($p(?:[*\/]$p)*))+$/&&/(.)(\Q$+\E)$/?a("$`",'',p($2," $1@_")):p$_,@_}s/\s//g;$_=a$_
parsing by regex sucks
290.48 Keith Calvert Ivey Sun Sep 1 15:37:22 2002 
$_=<>;s/\s//g;@s=/(?<![\d)])-?\d+|\S/g;L:for(@s){
/\w/and@o=(@o,$_),next;
/\(/and@_=(@_,$_),next;
if(/\)/){/\(/?next L:(@o=(@o,$_))while$_=pop@_;next}
if(/[+-]/){@o=(@o,pop@_)while@_&&$_[-1]!~/\(/}
if(/[*\/]/){@o=(@o,pop@_)while$_[-1]=~/[*\/]/}
/\S/&&push@_,$_
}@o=(@o,reverse@_);print"@o
"
Submitting my first working solution, for the record. Haven't shortened it at all yet, obviously.
292.43 Constantin M Sat Sep 7 19:18:08 2002 
$s='-+*\/';
$d='-?\d+';
$_=<>;
s/\s+//g;
@_=(m/(^$d|(?<![$s(])[$s]|$d|.)/g);
while(~($_=shift@_)){
/\(/?$.++:/\)/?$.--:/$d/?$r.=$_.' ':/./?push@o,[$_,$.]:0;
$_=$_[0]||'';
while((/[$s)]$/||!@_ )&&@o&&$o[$#o][1]==$.&&!($o[$#o][0]=~/[-+]/&&/[*\/]/)){$r.=${pop@o}[0].' '}}
$_=$r;
s/\s$/\n/;
print
295.21 Amir Karger Wed Sep 4 19:57:25 2002 
-pl @a=qw(+- */);
sub e2r {
    my $level = pop;
    if ($level == 2) {
^Is/^\(// && &e2r x s/^\)// || s/-?\d+// && $&
    } else {
^Imy $exp = &e2r($level+1);
^Iwhile (s#^[$a[$level]]##) {
^I    my $op = $&;
^I    $e2 = &e2r($level+1);
^I    $exp .= " $e2 $op";
^I    }
^I$exp;
    }
}
s/\s+//g;
$_=e2r
Minimal squeezing, but arithmetic & multiplicative expressions are almost the same, so shoehorn all three types into one subroutine that's called with a "level".
316.87 Peter Haworth Fri Sep 6 18:46:57 2002 
-p $t='-?';s#\s##g;{if(s#^$t\d+##){$\.="$& "}elsif(s#^\(##){$n=$O.=$&}elsif(s#^\)##){$\.="$& "while$O=~s#[^(]$##;$O=~s#\($##}elsif(s#^[+-]##){$o=$&;$\.="$& "while$O=~s#[-+/*]$##;$n=$O.=$o}elsif(s#^[/*]##){$o=$&;$\.="$& "while$O=~s#[/*]$##;$n=$O.=$o}else{last}($t,$n)=$n&&'-?';redo}$\.="$o "while$o=chop$O;$\=~s/ $/
/
This is feeble, but I've run out of time
317.75 me ELOISE Mon Sep 2 15:54:39 2002 
-l sub g{my$t=pop;$t=~s|\s*((^\|[*/+-])\s*)-\s*|$1=|g;$t=~s|^\s*(-?\d+)\s*$|$h{A.++$i}=$1;"A$i"|ge;for$a('*/','-+'){$t=~s|\s*(=?[A0-9]+)\s*([$a])\s*(=?[A0-9]+)\s*|$h{A.++$i}="$1 $3 $2";"A$i"|ewhile$t=~/[$a]/}$t}chomp($_=<>);s|\(([^()]+)\)|$h{A.++$i}=g$1;"A$i"|gewhile/\(/;$_=g$_;s|(A\d+)|$h{$1}|while/A/;s|=|-|g;print
not sure why I'm bothering to refine this algorithm since I know it can't compete...
322.75 GONZO Wed Sep 4 11:44:23 2002 
-l @s=@o=();%p=split//,'-1+1/2*2_3';($_=<>)=~s/\s//g;s/(^|[-+\/*(])-(\d+)/$1_$2/g;@l=split/([-+\/*()])/;for(@l){s/_/-/g;/\d/&&push@o,$_;/\(/&&push@s,$_;if(/\)/){while(defined($x=pop@s)and'('ne$x){push@o,$x}}if(/^[-+\/*]$/){while($p{$s[$#s]}>=$p{$_}){$x=pop@s;push@o,$x}push@s,$_}}while($x=pop@s){push@o,$x}print join' ',@o
Better than nothing...
327.75 Michael Thelen Sun Sep 1 07:09:38 2002 
-ln0 s/\s//g;s!(^|[+/*-])-!$1~!g;@t=/[\(\)+\/*-]|~?\d+/g;for(@t){s/~/-/;/\d/?push(@g,$_):/\(/?push(@s,$_):/\)/?${$f=pop@s,($f ne'('&&(push(@g,$f),redo))}:/[+-]/?${$s[-1]=~/[\*\/+-]/&&($f=pop@s,push(@g,$f),redo);push@s,$_}:/[*\/]/?${$s[-1]=~/[\*\/]/&&($f=pop@s,push(@g,$f),redo);push@s,$_}:1}$f=pop@s,push@g,$f while@s;print"@g"
Holy moly, what an ugly drive. Now to hack my way out of the rough...
395.72 me ELOISE Mon Sep 2 15:32:22 2002 
-l sub g{my$t=pop;$t=~s|\s*((^\|[*/+-])\s*)-\s*|$1=|g;$t=~s|^\s*(-?\d+)\s*$|$h{A.++$i}=$1;"A$i"|ge;$t=~s|\s*(=?[A0-9]+)\s*([*/])\s*(=?[A0-9]+)\s*|;$h{A.++$i}="$1 $3 $2";"A$i"|ewhile$t=~/[*\/]/;$t=~s|\s*(=?[A0-9]+)\s*([+\-])\s*(=?[A0-9]+)\s*|$h{A.++$i}="$1 $3 $2";"A$i"|ewhile$t=~/[+-]/;$t}chomp($_=<>);s|\(([^()]+)\)|$h{A.++$i}=g$1;"A$i"|gewhile/\(/;$_=g$_;s|(A\d+)|$h{$1}|while/A/;s|=|-|g;print
*sigh* ... at least I'm on the board...
438.24 Qingning Huo Thu Sep 5 13:16:50 2002 
-lp s/\s//g;
$o=0;
@s=@o=();
while (/\S/) {
^I$o ? s/.// : s#-?\d+|\(##;
^I$n = $&;
^Iif ($n eq '(') {
^I^Ipush @s, $n;
^I^I$o = 0;
^I} elsif ($n eq ')') {
^I^Iwhile (($a=pop(@s)) ne '(') {
^I^I^Ipush @o, $a;
^I^I}
^I^I$o = 1;
^I} elsif ($n !~ m#\d#) {
^I^Iwhile (@s && $s[-1] ne '(' && ($s[-1] =~ m#[*/]# || $n =~ m#[-+]#)) {
^I^I^Ipush @o, pop @s;
^I^I}
^I^Ipush @s, $n;
^I^I$o = 0;
^I} else {
^I^Ipush @o, $n;
^I^I$o = 1;
^I}
}
while (@s) { push @o, pop @s; }
$_ = "@o";
472.29 Peter Haworth Wed Sep 4 19:32:43 2002 
-p 
$t='-?';
s#\s##g;
{
  if(s#^$t\d+##){
    $\.="$& "
  }elsif(s#^\(##){
    $n=$ops.=$&
  }elsif(s#^\)##){
    while($ops=~s#[^(]$##){
      $\.="$& "
    }
    $ops=~s#\($##
  }elsif(s#^[+-]##){
    $op=$&;
    while($ops=~s#[-+/*]$##){
      $\.="$& ";
    }
    $n=$ops.=$op
  }elsif(s#^[/*]##){
    $op=$&;
    while($ops=~s#[/*]$##){
      $\.="$& ";
    }
    $n=$ops.=$op
  }else{
    last
  }

  ($t,$n)=$n&&'-?';
  redo  
}
$\.="$o "while$o=chop$ops;
$\=~s/ $/
/
Even after golfing, this is still going to be too big
589.25 Peter Haworth Tue Sep 3 17:31:26 2002 
-p 
$t=1;
while(1){
  s#^\s+##;
  if($t && s#^-\s*0*(\d+)##){
    $\.= "-$1 ";
  }elsif(s#^0*(\d+)##){
    $\.= "$1 ";
  }elsif(s#^\(##){
    push @ops,'(';
    $n=1;
  }elsif(s#^\)##){
    while(($op=pop @ops) ne '('){
      $\.= "$op ";
    }
  }elsif(s#^([+-])##){
    $op=$1;
    while(@ops && $ops[-1]=~m#[+-/*]#){
      $\.= pop(@ops)." ";
    }
    push @ops,$op;
    $n=1;
  }elsif(s#^([/*])##){
    $op=$1;
    while(@ops && $ops[-1]=~m#[/*]#){
      $\.= pop(@ops)." ";
    }
    push @ops,$op;
    $n=1;
  }else{
    last;
  }
  ($t,$n)=$n
}
($\.=join' ',reverse @ops)=~s/ ?$/
/;
693.32 Amir Karger Wed Sep 4 19:50:50 2002 
-nl 
use strict;

s/\s+//g; # first get rid of whitespace
my $a = a2r();
print $a;

# arithmetic expression is mult. exp. followed by N pairs of 
# arithmetic operator and mult. exp.
sub a2r {
    my $mult = m2r();
    while (s/^[+-]//) {
^Imy $op = $&;
^Imy $m2 = m2r();
^I$mult .= " $m2 $op";
    }
    $mult;
}

# multiplicative expression is unary exp. followed by N pairs of 
# mult. operator and unary exp.
sub m2r {
    my $unary = u2r();
    while (s#^[*/]##) {
^Imy $op = $&;
^Imy $u2 = u2r();
^I$unary .= " $u2 $op";
    }
    $unary;
}

# Unary expression is integer or parenthesized arith. exp.
sub u2r {
    if (s/^\(//) {
^Imy $a = a2r();
^Is/^\)//;
^I$a;
    } else {
^Is/-?\d*//;
^I$&;
    }
}
I'm entering this contest against my better judgement. My eyeballs still ache from mushing them back in on August 8...
1666.36 André Savigne Wed Sep 4 01:50:04 2002 
eval eval '"'.
 
 '\\$_=\\"\\@\\{[<>=~/\\\\'.('`'|'$').'+|\\\\'.('{'^'(').'/'.('`'
 |"'").']\\}\\";'.('['^'(').';(?<=[^\\\\'.('`'|'$').')]'.('{'^'['
 ).((                                                        '-')
 ).((                                                        ')')
 ).(#  Ovian Joy                           `/anick Champoux  '{'^
 '{'^                                                        '[')
 .';'                                                        .';'
 .''.                                                        ('`'
 |"\'").                                                  (';').(
 '['^"\(").                                            ';^-'.('{'
 ^'[').';-;;'.                                      ('`'|('-')).(
 '`'|'!').("\["^                                  '+').'\\{\\$'.(
 '['^'#').'.=/\\'.                              '\\'.('`'|('$')).
 ('/?\\$_.\\$\\":').                          "'"."'".';\\$,=/\\'
 .'\\(/.\\$,;/\\\\)/'                        .'&&(\\$,=~'.(('[')^
 '(').'/(.*?)'.("\^"^(                      '`'|"\/")).'//,\\$'.(
 '['^"\#").'.=\\$+);'.(                    '`'|')').('`'|'&').'('
 .'/[-+*\\\\/]\\$/)\\{'.                  '\\$,=~'.('['^'(').'/.'
 .'.//,\\$'.('['^'#').'.'                .'=\\$&'.('['^',').('`'|
 '(').('`'|')').('`'|',').              ('`'|'%').'\\$,&&\\$,<!'.
 '\\$%&\\$,=~/^[-+]/<=//;'              .'\\$,=\\"\\$_'.('{'^'[')
 .'\\$,\\"\\}\\}/\\\\'.('{'            ^'(').'+/'.('`'|"'").';\\'
 .'$_=\\$'.('['^'#').'.\\$'            .',;'.('['^'(').'/.\\$//;'
 .('['^'+').('['^')').('`'|')').('`'|'.').('['^'/').'\\$_.\\$/"';
 $:='.'^'~';$~='@'|'(';$^=')'^'[';$/='`'|'.';$_='('^'}';$,=('`');
"Undies Clad Figure Trudging". To mark the changing of the guard in Canadian golf. Ala only needs to finish in 41st place to overtake `/anick Champoux on the Career Money List. And to celebrate the launch of the new "Ovian Joy"(R) range of designer underwear by `/anick Champoux Fashion, Inc. (R)"Ovian Joy" is a registered trademark of `/anick Champoux Fashion, Inc. Used with permission.

And, of course, I had to be the one to ref that entry... Oh, the irony, the cruel irony of it...

Artistic/Unorthodox

ScoreGolferSubmit TimeCode
141.36 Rick Klement Sat Sep 7 14:53:42 2002 
-p sub o{print$"x$|++,shift}
sub f{s/^-*\d+//?o$&:e(s/.//)+s/.//}
sub t{f;o$&x1,f while s+^[*/]++}
sub e{t;o$&x1,t while s/^[+-]//}
e y/ ^I//d
My first ever entry to "Artistic". It's a classical recursive descent, and I think it looks quite elegant and even pretty. It has an interesting mechanism for local storage in the recursive routines :) Also, if the lines are combined into one line, it's a 137, one shorter than the original BoB!

pretty :)

142.35 Rick Klement Sat Sep 7 14:52:30 2002 
-p sub o{print$"x$|++,shift} 
sub f{s/^-*\d+//?o$&:e(s/.//)+s/.//}
sub t{f;o$&x1,f while s+^[*/]++}
sub e{t;o$&x1,t while s/^[+-]//}
e y/ ^I//d
My first ever entry to "Artistic". It's a classical recursive descent, and I think it looks quite elegant and even pretty. It has an interesting mechanism for local storage in the recursive routines :) Also, if the lines are combined into one line, it's a 137, one shorter than the original BoB!
158.44 Eugene van der Pijll Thu Sep 5 23:20:42 2002 
-lpX use
base
B::Deparse;s/-/$:/g;s/\d+/a$&/g;sub
maybe_parens{$_=$_[1];s/ (.*) (.*)/ $2 $1/;y/ /_/;$_}$_=main->new->coderef2text(eval"sub{$_}");y/_
-';-}/ /d
193.44 TheodoreYoung Tue Sep 3 16:03:52 2002 
-l package A;
use overload
'""'=>sub{${$_[0]}},
'neg'=>sub{n A "-$_[0]"},
'nomethod'=>sub{n A "@_[0,1,3]"};
sub n{my $c=$_[1];bless \$c}
$_=<STDIN>;
s/(-?)[0 ]*(\d+)/ $1 A->n($2)/g;
print eval;

Rejected

ScoreGolferSubmit TimeCode
17.37 Taipei.pm Sun Sep 1 13:34:21 2002 
-wle 'print %]';
111.75 Ton Hospel Mon Sep 2 02:57:31 2002 
-p s#(^|[*-/(])\s*(-?\s*\d+)#$1(bless[s!\$! $2!])#g;s/.?/"use overload '$&',sub{s!! $&!;shift}"/gee;eval$`;s///
113.75 Ton Hospel Mon Sep 2 02:50:14 2002 
-p s#(^|[*-/(])\s*(-?\s*\d+)#$1(bless[s!\$! $2!])#g;eval"$_;".s/./"use overload '$&',sub{s!! $&!;shift}"/gee;s///
more regex repeating
115.47 Stephen Turner Mon Sep 2 10:41:48 2002 
-p use
overload
nomethod,sub{bless[@{$_[0]},@{$_[1]},pop]};y/ ^I//d+s/(?<![\d)])-?\d+/(bless["$&"])/g;$_="@{eval
}
"
We don't need to make the constructor explicit. WARNING: Like my last program, this uses a lot of memory. Maybe even more than my last program.

The memory usage is problematic, but it seems that even with the full 4 GB (allowed by the rules) of real memory this would be way too slow.

115.76 Ton Hospel Mon Sep 2 02:38:49 2002 
-p s#(^|[*-/(])\s*(-?\s*\d+)#$1(bless[s!\$! $2!])#g;eval"$_;".s/./"use overload '$&',sub{s!\$! $&!;shift}"/gee;s///
The output returned for '1 - - 2' is '1 - 2', but it should be something like '1 -2 -'
116.66 Stephen Turner Mon Sep 2 13:08:08 2002 
-pl use
overload'""',sub{"@{$_[0]}"},nomethod,sub{bless\@_};y/ ^I//d+s/(?<![\d)])-?\d+/(bless["$&"])/g;$_=eval;y/ //s
While my rude 115.47 is still pending, try this one, which is one stroke longer, but much more polite.

This has now taken 15 minutes of CPU on a 750 MHz Athlon, which isn't swapping. Sorry, but I'll have to reject the solution.

116.78 Ton Hospel Mon Sep 2 02:32:18 2002 
-lp s#(^|[*-/(])\s*(-?\s*\d+)#$1(bless[s!\$!$2 !])#g;eval"$_;".s/./"use overload '$&',sub{s!\$!$& !;shift}"/gee;chop
mm, the horse still is not dead
120.62 Chia-chi Yeh Tue Sep 3 19:02:01 2002 
-pl s/\s//g;$_.=")";0while s!\(?([^()]*)\)!$a=$1;for(qw(*/ +-)){0while$a=~s#(\S)([$_])(.*?\S)([*/+-]|$)#$1 $3 $2$4#}$a!e
test 45 has not been completed after 20 minutes of execution in a Pentium Celeron at 300 MHz
128.78 Ton Hospel Mon Sep 2 00:12:12 2002 
-lp s#(^|[-+*/(])\s*(-?\d+)#$1(bless[\$q.="$2 "])#g;s/./"use overload'$&',sub{\$q.='$& ';shift}"/eegfor$q="+-*/";eval;$_=$q;chop
Mmm, i'm starting to doubt this horse

Unfortunaty (again), it fails the test with '1 - -\t2'.

131.35 pom Mon Sep 2 11:11:04 2002 
-p $_="($_)";y/ ^I//d;s/\b/x/g;1while
s!\(([^()]+)\)!$_=$1;$*=$1,s/ (-?)x(\PL+)/$1$2 $*/while
s-x([*/])- -||s/x([+-])/ /;$_!e;y/x//d
A slight improvement...
131.49 Mykola Zubach Tue Sep 3 13:18:26 2002 
-lp y/() \t/<>/d;for(s/(?<![>\d])(-?\d+)/<$1>/g;s/<(<[^>]+>)>/$1/||s@>([/*]|([+-]))<
([^<>]+)>(?(2)([^*/]|$))@ $3 $1>$4@;){}y/<>//d
old version, but improved

This only seems to remove whitespace. For example the first test: Input: 1 - - 2 Output: 1--2

131.74 Eugene van der Pijll Mon Sep 2 17:03:54 2002 
-lp s/\s//g;$n='((\B-)?\w+)';push@a,"$1 $4 $3"while s#\($n\)#$1#g||s#$n([*/])$n#a.@a#e||s#$n([-+])$n#a.@a#e;1while s#a(\d+)#$a[$1]#
Alas, alas, the program fails for (18*16*16*5-1+12+15+18*1-8+6/7-6-2-(19)*(17))+8+((9/14))*16+15*11*13/20*4+16- 7-19+9+(12)+(4)+20-1*(16)+(6)+19*13/1*6*5/17-17+(2)*6*4-19*17-4*(1-20*11*3*19 +3/(8+9*20/16+(16)/8/3))-((3))*(6+2-(17)*2*2-(10))*(14-15*4+6/1
134.71 Mtv Europe Sun Sep 1 22:00:48 2002 
-p /\(/?push@;,0:/\d/?push@o,$_:do{push@o,pop@;while/\*|\//?$;[-1]=~//:$;[-1];/\)/?pop@;:push@;,$_}for/(?<!\d|\))-?\d+|\S/g,a;$_="@o
"
and yet another way how to kill the cat. fast.

Unfortunaly, it fails the test with '1 - -\t2'

140.77 Ton Hospel Sun Sep 1 22:24:24 2002 
-pl s/./,"$&",sub{bless["\@_$&"]}/gfor$a="+-*/";s#(^\s*|[-+*/(]\s*)(-?\d+)#$1(bless["$2"])#g;s/^/use overload'""',sub{\$_[0][0]}$a;/;$_=eval
Unfortunaly, it fails the test with '1 - -\t2'.
140.84 Eugene van der Pijll Tue Sep 3 17:26:15 2002 
-lpX sub a{bless\"@_"}use overload neg,sub{a('-'.shift)},'""',sub{${$_[0]}},nomethod,sub{a shift,shift,pop};s/-/ -/g;s/\d+/a("$&")/g;$_=eval
Finally, a working solution. I hope.

fail's on input (-1 + 2), outputs nothing .. check if you submitted the right one, or made a paste error

142.56 Rick Klement Sun Sep 1 20:34:13 2002 
-p s//M/;y/  //d;s!([)\d])-!$1=!g;1while
s/\(M\)/M/|s!([^(])M(?=$|[=+)])|([*/])M(?=$|[*/)])|M(-?\d+)!push@a,$+;M!e||s/M(.)/$1M/;s/M/@a/;y/=/-/
Fails on 1 - -2.
144.67 Alain "TopMach" Dupuis Sat Sep 7 19:51:05 2002 
-lp s/\s//g;$%=/\(/,s!\(?([^()]+)\){$%}!$_=$1;for$o('*/','+-'){1while s#([^SO])([$o])(-?[S\d]+(O.|S\d+)*)#$1S$3O$2#}$_!ewhile/(\d|\W)\W/;y/SO/ /
Test 45 is taking ages in the machines of two referees. It seems this program does not satisfy the time constraint stated in the rules.
155.49 Mtv Europe Sun Sep 1 17:06:19 2002 
-p s/\d+/\$$&/g;s/-/ -/g;close STDERR;$_=`$^X -MO=Deparse,-p -e '$_'`;
$q=qr/\(((??{$q})|-?\$\d+)( .)?( (??{$q})| -?\$\d+)?\)/;
1while s/$q/$1$3$2/;y/$;//d
hmm, artistic or unortodox? at least it's platform dependant (needs s/'/#/g for windows). and speed is not my fault, it's Deparse's! :)

too slow and platform dependant .. even arthistic / unorthodox need to pass the judge :)

155.81 TheodoreYoung Tue Sep 3 17:59:34 2002 
-l sub n{bless\shift}use overload'""',sub{${$_[0]}},'neg',sub{n "-$_[0]"},'nomethod',sub{n "@_[0,1,3]"};$_=<STDIN>;s/(-?)[0\s]*(\d+)/ $1 n($2)/g;print eval
doesnt pass any tests ..
159.73 Amir Karger Wed Sep 4 20:01:29 2002 
-pl @a=qw(+- */);sub e{my($l,$o)=@_;if($l-2){do{e($l+1);$o&&push@c,$o}while$o=s#^[$a[$l]]##&&$&}else{s/^\(//?e()+s/^\)//:s/-?\d+//+push@c,$&}}s/\s//g;e;s/$/@c/
Squeeze out lots o' whitespace. Sadly, I'm still very very far from the top. But it's a good enough score to at least enter the contest.

never finish test1

179.70 nashdj Sun Sep 1 09:56:31 2002 
-l $e=qr/(??{$t})(([+-])(??{$t})(?{push@_,$2}))*/;$t=qr/(??{$f})(([*\/])(??{$f})(?{push@_,$2}))*/;$f=qr/\((??{$e})\)|((-|)\d+)(?{push@_,$1})/;$_=<>;s/ |\t//g;$_=~$e;$,=' ';print@_
Consistently causes Perl 5.6.1 on Linux to segfault in test 33.
182.29 Qingning Huo Tue Sep 3 07:57:22 2002 
-pl s#\s##g;
s#(.*?)(^|[-+*/(])(-?\d+)#$1$2\n$3\n#g;
1while
s#\((\n.+\n)\)#$1#||
s#^([^)]*?)(.+)\n([*/])\n(.+)#$1$2 $4 $3#||
s#(.+)\n([-+])\n(.+)(?=\n|$)(?!\n[*/])#$1 $3 $2#;
s#\n##g
Way too slow. Test 45 (v7) has now taken 12 minutes of CPU time on the specified machine, and the really long tests are still to come.
184.43 TheodoreYoung Tue Sep 3 13:14:47 2002 
-l package A;
$_=<STDIN>;
s/(-?)[0 ]*(\d+)/ $1 A->n($2)/g;
print eval;
use overload
'""'=>sub{${$_[0]}},
'neg'=>sub{n A "-$_[0]"},
'nomethod'=>sub{n A "@_[0,1,3]"};
sub n{bless \$_[1]}
This doesn't seem to produce any output besides a newline for any input.
189.56 Greg Allen Sat Sep 7 17:09:47 2002 
-pl $*='(?:-?\d+(?: (?:[-*\/+]|-?\d+))*)';sub a{$_=~$1?$1:$`;0while s/\(([^()]+)\)/&a/e;s&($*(?:[*/]$*)*)([*/])($*)&a(" $3 $2")&eg;$_=a(" $3 $2")if/()([-+])($*)$/;"$_@_";}s/\s//g+s/(.*)/a/e
hurrah, some semblance of respectability

Hangs on test1.

190.43 Qingning Huo Fri Sep 6 08:20:58 2002 
-p y/  //d;my@o;{s/^(\(+)?(-?\d+)//;$_.=$1;push@o,$2;while(s/^\)//||/^
/){push@o,$&while
s#[^(
]$##;chop}if(s/^.//){$_.=$&;push@o,$&while
s#[^(
](?=[-+]$)##||s#[*/](?=[*/]$)##;redo}}$_="@o
"
Fails the first test: Input: 1 - - 2 Output: 1 - 2 -
194.56 Greg Allen Sat Sep 7 17:23:32 2002 
-pl $*='(?:-?\d+(?: (?:[-*\/+]|-?\d+))*)';sub a{$_=$1;0while s/\(([^()]+)\)/&a/e;s&($*(?:[*/]$*)*)([*/])($*)&a(" $3 $2")&eg;$_=a(" $3 $2")if/^(.*(?:\d| .))([-+])($*)$/;"$_@_";}s/\s//g+s/(.*)/a/e
hurrah, some semblance of respectability

Sorry, the segfault persists. The problematic piece of code is: s&($*(?:[*/]$*)*)([*/])($*)&a(" $3 $2")⪚

206.77 Michael Thelen Sun Sep 1 20:49:13 2002 
-ln s/\s//g;/\d/?push@g,$_:/\)/?($f=pop@s)ne'('&&push(@g,$f)&&redo:(/([+-])|[*\/]/&&($z="*/".($1&&"+-"),"@s"=~/[$z]$/)&&push(@g,pop@s)&&redo,push@s,$_)for/(?:(?<![\d()])-)?\d+|./g;push@g,reverse@s;print"@g"
Fails on (-1 + 2)
208.53 Greg Allen Sat Sep 7 10:46:04 2002 
-pl $*='(?:-?\d+(?> (?:[^I*\/+]|-?\d+))*)';sub a{$_=$1;0while s/\(([^()]+)\)/&a/e;s&($t(?:[*/]$*)*)([*/])($*)&a(" $3 $2")&eg;$_=a(" $3 $2")if/^(.*)([+^I])($*)/;"$_@_";}s/\s//g+s/(\d|\))-/$1^I/g+s/(.*)/a/e+y/^I/-/
so much effort, so little progress

Running test 47 ... Program rpn.pl killed by signal (11) consistently dies on test47

228.17 Amir Karger Wed Sep 4 19:59:55 2002 
-pl @a=qw(+- */);
sub e{
    my ($l,$o) = @_;
    if ($l-2) {
^Ido {
^I    e($l+1);
^I    $o && push @c,$o
^I} while $o = s#^[$a[$l]]## && $&
    } else {
^Is/^\(// ? e() + s/^\)// : s/-?\d+// + push @c, $&
    }
}
s/\s//g;
e;s/$/@c/
Get rid of long variable names. Realized that arith/mult. expressions are perfect for do while loops.

doesnt finish fist test.

229.49 Mtv Europe Thu Sep 5 18:33:18 2002 
-p use B;s/-/
-/g;s/\d+/\$$&/g;
for($x=($y=B::svref_2object(eval"sub{$_}"))->START,
@v=($y->PADLIST->ARRAY)[1]->ARRAY;$_=($x=$x->next)->name,!/ea/;){
push@a,/n/?-pop@a:/g/?@v[$x->padix]->NAME:/e/?'/':/y/?'*':/s/?'-':'+'
}$_="@a
"
this B thing crashes too often

Failed with this message : Can't locate object method "padix" via package "B::SVOP" (perhaps you forgot to load "B::SVOP"?)

234.76 Michael Thelen Sun Sep 1 20:11:39 2002 
-ln s/\s//g;/\d/?push@g,$_:/\)/?($f=pop@s)ne'('&&push(@g,$f)&&redo:(/[+-]/?"@s"=~/[\*\/+-]$/&&(push(@g,pop@s),redo):/[*\/]/&&"@s"=~/[\*\/]$/&&(push(@g,pop@s),redo),push@s,$_)for/(?:(?<![\d()])-)?\d+|[(-\/]/g;push@g,reverse@s;print"@g"
Better, but still a long way to go.

Fails on (-1 + 2)

263.69 Greg Allen Thu Sep 5 10:38:22 2002 
-pl sub t{$_=pop;s/\((.*)\)/a($1)/e;"$_@_"}$b=qr/\((?:(?>[^()]+)|(??{$b}))*\)/;$p=qr/$b|-?\d+/;sub p{$_=shift;/(\*|\/)($p)$/?p("$`",'',t(" $1@_",$2)):t@_,$_}sub a{$_=shift;/$p([-+]($p(?:[*\/]$p)*))+$/&&/(.)(\Q$+\E)$/?a("$`",'',p($2," $1@_")):p$_,@_}y/ ^I//d;$_=a$_
guess this approach won't get too close to 104 then...

Segfaults on 5.6.1 on Linux. Probably the (??{}).

337.49 Constantin M Fri Sep 6 21:34:42 2002 
$s='-+*\/';
$d='-?\d+';
$_=<>;
s/\s+//g;
@_=(m/(^$d|(?[$s(]
</td></tr><tr><td colspan="2">
Paste your solution in the textarea above or upload it here:
  
<input type="file" name="file">
  
<span class="error">Please use specified shebang</span>
<br><br>
</td></tr><tr><td>
Comment:
</td><td>
<textarea cols="40" rows="4" name="comment">
seem to have happened something strange when you submitted because your solution have a lot of html in it ..
365.30 Philippe Bricout Sat Sep 7 19:08:28 2002 
-pl y#*/+-#a-d#;
s/\s*([a-d]|[()])\s*/_$1_/g;
s/(^|[a-d(])_*d_*(\d+)/$1_d$2/g;
y/ _/  /s;
&p while &d;
y#_a-d# */+-#s;
s/^\s|\s$//g;
sub a{&p while s/\( (\w+) \)/$1/}
sub b{&p while &a;s/\( (\w+) ([a-d]) (\w+) \)/$1_$3_$2/}
sub c{&p while &b;s/(\w+) ([ab]) (\w+)/$1_$3_$2/}
sub d{&p while &c;s/(\w+) ([cd]) (\w+)/$1_$3_$2/}
sub p{#y#a-d#*/+-#;print;y#*/+-#a-d#;
1};
This time the question is 42, not the answer! Don't know why test 42 failed : problem with my script or problem during evaluation of division and big interger ? Help me ! :)

As you already say, it fails on test 42. We are sorry we cannot help participants, if referees did it the game would be spoiled.

422.98 Taipei.pm Sun Sep 1 13:38:11 2002 
-wle 'print %]'; print '$_/%&__$%%[s/%$@$#//\^#####################################################################################################################################################################################################################################################################################################################################################################################';
818.00 Ton Hospel Sun Sep 1 13:42:20 2002 
#!perl -wle 'print $_$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$