in reply to Refactoring challenge.

The final incarnation (unless you know better?), with grateful thanks to all who helped especially dragonchild, hv & Roy Johnson 

#! perl -slw
use strict;
use List::Util qw[ min ];
$|=1;

sub indexBal {
    my( $string, $lookFor, $limit ) = @_;
    my $nesting = 0;
    for( my( $position, $char ) = ( 0, substr $string, 0, 1 );
        $position < min( length( $string ) , $limit );
        $char = substr $string, ++$position, 1
    ) {
        $nesting++ if 1+index '[{', $char;
        $nesting-- if 1+index ']}', $char;
        die 'Unbalanced' if $nesting < 0;
        return $position if $char eq $lookFor and $nesting == 0
    }
    return;
}

my $depth = 0;
my $indent = '    ';
my $string = '';

sub findChunk {
    my( $str, $width ) = @_;
    $str =~ m/([\Q[]{},\E])/g or return ();
    my( $pos, $first ) = ( $+[0], $1 );
    return( $pos +1,  0 ) if $first eq ',';
    return( $pos +1, -1 ) if 1+index ']}', $first;
    if( my $pos2 = indexBal( $str, $first eq '[' ? ']' : '}', $width )
+ ) {
        return( $pos2 +2,  0 );
    }
    return( $pos   , +1 );
}

sub pp {
    my( $nextBit, $width, $EOS ) = @_;
    $string .= $nextBit;
    while( length( $string ) > $width or $EOS ) {
        my( $pos, $adjust ) = findChunk( $string, $width );
        return unless defined $pos;

        $depth-- if $adjust < 0;
        print $indent x $depth, substr( $string, 0, $pos, '' );
        $string =~ s[^\s*][];
        $depth++ if $adjust > 0;
    }
    return;
}

## Test harness only below here
my $data = <DATA>;

my $width = $ARGV[ 0 ]||100;
$indent = $ARGV[ 1 ] if @ARGV == 2;

while( length $data ) {
    my $p = 1+rindex( $data, ',', 1+rand( 100 ) )||length $data;
    my $sub = substr( $data, 0, $p, '' );
    pp( $sub, $width, !length $data );
}

__DATA__
{ a => [ SCALAR(0x18bb45c), { a => b, c => d, e => f, g => h,  }, [ 1,
+ 2, 3, 4, 5, 6, 7, 8, 9, 10,  ], { a => b, c => d, e => f, g => h,  }
+, [ 1, 2, 3, 4, 5, 6, 7, 8, 9, 10,  ], { a => b, c => d, e => f, g =>
+ h,  }, [ 1, 2, 3, 4, 5, 6, 7, 8, 9, 10,  ], { a => b, c => d, e => f
+, g => h,  }, [ 1, 2, 3, 4, 5, 6, 7, 8, 9, 10,  ], { a => b, c => d, 
+e => f, g => h,  }, [ 1, 2, 3, 4, 5, 6, 7, 8, 9, 10,  ],  ], b => [ S
+ELFREF(018bb45c), { a => b, c => d, e => f, g => h,  }, [ 1, 2, 3, 4,
+ 5, 6, 7, 8, 9, 10,  ], { a => b, c => d, e => f, g => h,  }, [ 1, 2,
+ 3, 4, 5, 6, 7, 8, 9, 10,  ], { a => b, c => d, e => f, g => h,  }, [
+ 1, 2, 3, 4, 5, 6, 7, 8, 9, 10,  ], { a => b, c => d, e => f, g => h,
+  }, [ 1, 2, 3, 4, 5, 6, 7, 8, 9, 10,  ], { a => b, c => d, e => f, g
+ => h,  }, [ 1, 2, 3, 4, 5, 6, 7, 8, 9, 10,  ],  ], c => [ SELFREF(01
+8bb45c), { a => b, c => d, e => f, g => h,  }, [ 1, 2, 3, 4, 5, 6, 7,
+ 8, 9, 10,  ], { a => b, c => d, e => f, g => h,  }, [ 1, 2, 3, 4, 5,
+ 6, 7, 8, 9, 10,  ], { a => b, c => d, e => f, g => h,  }, [ 1, 2, 3,
+ 4, 5, 6, 7, 8, 9, 10,  ], { a => b, c => d, e => f, g => h,  }, [ 1,
+ 2, 3, 4, 5, 6, 7, 8, 9, 10,  ], { a => b, c => d, e => f, g => h,  }
+, [ 1, 2, 3, 4, 5, 6, 7, 8, 9, 10,  ],  ], d => [ SELFREF(018bb45c), 
+{ a => b, c => d, e => f, g => h,  }, [ 1, 2, 3, 4, 5, 6, 7, 8, 9, 10
+,  ], { a => b, c => d, e => f, g => h,  }, [ 1, 2, 3, 4, 5, 6, 7, 8,
+ 9, 10,  ], { a => b, c => d, e => f, g => h,  }, [ 1, 2, 3, 4, 5, 6,
+ 7, 8, 9, 10,  ], { a => b, c => d, e => f, g => h,  }, [ 1, 2, 3, 4,
+ 5, 6, 7, 8, 9, 10,  ], { a => b, c => d, e => f, g => h,  }, [ 1, 2,
+ 3, 4, 5, 6, 7, 8, 9, 10,  ],  ], e => [ SELFREF(018bb45c), { a => b,
+ c => d, e => f, g => h,  }, [ 1, 2, 3, 4, 5, 6, 7, 8, 9, 10,  ], { a
+ => b, c => d, e => f, g => h,  }, [ 1, 2, 3, 4, 5, 6, 7, 8, 9, 10,  
+], { a => b, c => d, e => f, g => h,  }, [ 1, 2, 3, 4, 5, 6, 7, 8, 9,
+ 10,  ], { a => b, c => d, e => f, g => h,  }, [ 1, 2, 3, 4, 5, 6, 7,
+ 8, 9, 10,  ], { a => b, c => d, e => f, g => h,  }, [ 1, 2, 3, 4, 5,
+ 6, 7, 8, 9, 10,  ],  ], f => [ SELFREF(018bb45c), { a => b, c => d, 
+e => f, g => h,  }, [ 1, 2, 3, 4, 5, 6, 7, 8, 9, 10,  ], { a => b, c 
+=> d, e => f, g => h,  }, [ 1, 2, 3, 4, 5, 6, 7, 8, 9, 10,  ], { a =>
+ b, c => d, e => f, g => h,  }, [ 1, 2, 3, 4, 5, 6, 7, 8, 9, 10,  ], 
+{ a => b, c => d, e => f, g => h,  }, [ 1, 2, 3, 4, 5, 6, 7, 8, 9, 10
+,  ], { a => b, c => d, e => f, g => h,  }, [ 1, 2, 3, 4, 5, 6, 7, 8,
+ 9, 10,  ],  ], g => [ SELFREF(018bb45c), { a => b, c => d, e => f, g
+ => h,  }, [ 1, 2, 3, 4, 5, 6, 7, 8, 9, 10,  ], { a => b, c => d, e =
+> f, g => h,  }, [ 1, 2, 3, 4, 5, 6, 7, 8, 9, 10,  ], { a => b, c => 
+d, e => f, g => h,  }, [ 1, 2, 3, 4, 5, 6, 7, 8, 9, 10,  ], { a => b,
+ c => d, e => f, g => h,  }, [ 1, 2, 3, 4, 5, 6, 7, 8, 9, 10,  ], { a
+ => b, c => d, e => f, g => h,  }, [ 1, 2, 3, 4, 5, 6, 7, 8, 9, 10,  
+],  ], h => [ SELFREF(018bb45c), { a => b, c => d, e => f, g => h,  }
+, [ 1, 2, 3, 4, 5, 6, 7, 8, 9, 10,  ], { a => b, c => d, e => f, g =>
+ h,  }, [ 1, 2, 3, 4, 5, 6, 7, 8, 9, 10,  ], { a => b, c => d, e => f
+, g => h,  }, [ 1, 2, 3, 4, 5, 6, 7, 8, 9, 10,  ], { a => b, c => d, 
+e => f, g => h,  }, [ 1, 2, 3, 4, 5, 6, 7, 8, 9, 10,  ], { a => b, c 
+=> d, e => f, g => h,  }, [ 1, 2, 3, 4, 5, 6, 7, 8, 9, 10,  ],  ], i 
+=> [ SELFREF(018bb45c), { a => b, c => d, e => f, g => h,  }, [ 1, 2,
+ 3, 4, 5, 6, 7, 8, 9, 10,  ], { a => b, c => d, e => f, g => h,  }, [
+ 1, 2, 3, 4, 5, 6, 7, 8, 9, 10,  ], { a => b, c => d, e => f, g => h,
+  }, [ 1, 2, 3, 4, 5, 6, 7, 8, 9, 10,  ], { a => b, c => d, e => f, g
+ => h,  }, [ 1, 2, 3, 4, 5, 6, 7, 8, 9, 10,  ], { a => b, c => d, e =
+> f, g => h,  }, [ 1, 2, 3, 4, 5, 6, 7, 8, 9, 10,  ],  ], j => [ SELF
+REF(018bb45c), { a => b, c => d, e => f, g => h,  }, [ 1, 2, 3, 4, 5,
+ 6, 7, 8, 9, 10,  ], { a => b, c => d, e => f, g => h,  }, [ 1, 2, 3,
+ 4, 5, 6, 7, 8, 9, 10,  ], { a => b, c => d, e => f, g => h,  }, [ 1,
+ 2, 3, 4, 5, 6, 7, 8, 9, 10,  ], { a => b, c => d, e => f, g => h,  }
+, [ 1, 2, 3, 4, 5, 6, 7, 8, 9, 10,  ], { a => b, c => d, e => f, g =>
+ h,  }, [ 1, 2, 3, 4, 5, 6, 7, 8, 9, 10,  ],  ],  }

[download]
Examine what is said, not who speaks.
Silence betokens consent.
Love the truth but pardon error.

Replies are listed 'Best First'.
Re^2: Refactoring challenge.
by Roy Johnson (Monsignor) on Mar 09, 2005 at 06:57 UTC
    I had to take another shot at it. I noticed that comma was a sequence point, so you basically have the non-comma opening bracket match, and the comma match, which covers both closing brackets and unbracketed terms.

    There may be some refactoring still left to do (update: did some more this morning), but I'm happy with it, and I need to go to bed. 

    sub pp_new {
    my( $nextBit, $width, $EOS ) = @_;
    $string .= $nextBit;
    while( length( $string ) > ($EOS ? 0 : $width) ) {
        for (substr($string, 0, $width)) {
            my $pos;
            my $leftmargin = $indent x $depth;
            if (my ($bracket) = /^[^,]*?([\Q[{\E])/) {
                $bracket =~ tr/[{/]}/;
                if (my $closepos = indexBal($_, $bracket, $width)) {
                    $pos = $closepos + 2;
                }
                else { $pos = $+[0]; ++$depth }
            }
            else {
                $pos = 1 + (/,/ ? $+[0] : $width);
                $leftmargin = $indent x --$depth if substr($_, 0, $pos
+-1) =~ /([\Q]}\E])/;    
            }
            print $leftmargin, substr( $_, 0, $pos, '' );
        }
        $string =~ s/^\s+//;
    }
}

    [download]
    Caution: Contents may have been coded under pressure.
[reply]
[d/l]
Re^2: Refactoring challenge.
by tphyahoo (Vicar) on Mar 08, 2005 at 17:14 UTC
    Maybe it would clearer like this, or at any rate more clear to people that only know perl. (I guess this is a vb/c thing.) 
    #$nesting++ if 1+index '[{', $char;
#$nesting-- if 1+index ']}', $char;
$nesting++ if $char=~/[[{]/;
$nesting-- if $char=~/[\]}]/;  # not sure if I escaped this right

    [download]
    The 1+index idiom baffled me, it requires that you know index returns -1 if no match. Thanks to demerphq for helping me understand.
[reply]
[d/l]

      I was using regex in the earlier versions if you look back, but I personally find the escaped regexes to be most unreadable. I think that the way it is now captures the essence of the tests very well.

      In theory at least, you can use /[[}]/ without any escaping as (most) meta characters have no special meaning inside character classes, but it doesn't work in practice.

      I actually thought that was true for all meta characters until hv showed me /([\Q[]{},\E])/, which is better than the alternative, but still far from transparent.

      I've never used VB, though it's not totally dissimilar to the DEC Basic Plus I used 25 years ago. I still sometimes type instr when I mean index.

      (I think instr is the better name for the function actually)

      Examine what is said, not who speaks.
      Silence betokens consent.
      Love the truth but pardon error.
[reply]
[d/l]
[select]

In Section Seekers of Perl Wisdom