请教个hash key排序问题

sjdy521

回复 5# jason680


    宽度对齐再进行ascii码排序，思路不错。学习了。不过如果数字和字符串的排序顺序不同呢

jason680

回复 11# sjdy521

给个例吧
想不出是什么问题....
   

sjdy521

回复 12# jason680


    没什么大问题啦，这个思路就很ok，只是想到如果要求数字按从大大小排序，字符按acsii从小到大排序，可能又复杂些

cdtits

貌似不够通用

jason680

回复 13# sjdy521

How about this, assign sort direction by yourself

direction 0: small to big (or A to Z and a to z)
direction 1: big to small (or z to a and Z to A)
    _by_word_number(1,1,0,0)


$ cat sort_word_number.pl

use strict;
use warnings;

my %hash = (
    '1A2a02'=>0.122,
    '1A02b1'=>0.121,
    '1A02a1'=>0.121,
    '1B20a3'=>0.1203,
    '1B21a3'=>0.1213,
    '1A2a1'=>0.121,
    '1Aa2a1'=>0.121,
    '1A3'=>0.5,
    '1A10'=>0.4,
    '1A8'=>0.7,
    '2B1a'=>0.41,
    '2B1'=>0.4,
    '2B3'=>0.2,
    '2B10'=>0.4,
    '2B7'=>0.1,
);
sub _by_word_number{
  my @aDir = @_;
  my @aB = split(/(\d+)/,$b);
  shift @aB if($aB[0] eq "");
  my $sRet;
  #print "$b,$a,@aDir\n";
  foreach(split(/(\d+)/,$a)){
    next if(m/^$/);
    my $sB = shift @aB;
    #print "$_,$sB,@aDir\n";
    if(m/\d+/){
      $sRet = ($_ <=> $sB);
    }
    else{
      $sRet = ($_ cmp $sB);
    }
    # to get the directory by @aDir
    my $sDir = shift @aDir;
    next if($sRet == 0);
    $sRet *= -1 if($sDir);
    return $sRet;
  }
}
foreach( sort{ _by_word_number(1,1,0,0)} keys %hash){
  print "$_=$hash{$_}\n";
}


   

rubyish

jason680 发表于 2012-09-23 12:50
回复 13# sjdy521

How about this, assign sort direction by yourself

学习了。Master!3Q!

wuxiaobo_2009

回复 5# jason680


    map { (my $nwn=$_)=~s/(\d+)/sprintf("%05d",$1)/ge; [$_, $nwn] }

中的$1 代表什么呢？

jason680

回复 17# wuxiaobo_2009

s/(\d+)/sprintf("%05d",$1)/ge