使用Perl对具有特定字母顺序的字符串进行排序

以下可能是最快的[1]：

sub my_compare($$) {
    $_[0] =~ tr{abdrtwsuiopqe987654}{abcdefghijklmnopqrs}r
       cmp
    $_[1] =~ tr{abdrtwsuiopqe987654}{abcdefghijklmnopqrs}r
}

my @sorted = sort my_compare @unsorted;

或者如果你想要更动态的东西,以下可能是最快的[2]：

@H_154_14@my @syms = split //,'abdrtwsuiopqe987654'; my @map; $map[ord($syms[$_])] = $_ for 0..$#syms; sub my_compare($$) { (pack 'C*',map $map[ord($_)],unpack 'C*',$_[0]) cmp (pack 'C*',$_[1]) } my @sorted = sort my_compare @unsorted;

我们可以逐字逐句地比较,但这会慢得多.

use List::Util qw( min );

my @syms = split //,'abdrtwsuiopqe987654';
my @map; $map[ord($syms[$_])] = $_ for 0..$#syms;

sub my_compare($$) {
    my $l0 = length($_[0]);
    my $l1 = length($_[1]);
    for (0..min($l0,$l1)) {
       my $ch0 = $map[ord(substr($_[0],$_,1))];
       my $ch1 = $map[ord(substr($_[1],1))];
       return -1 if $ch0 < $ch1;
       return +1 if $ch0 > $ch1;
    }

    return -1 if $l0 < $l1;
    return +1 if $l0 > $l1;
    return 0;
}

my @sorted = sort my_compare @unsorted;

>从技术上讲,使用GRT可以更快地完成.

@H_154_14@my @sorted = map /\0(.*)/s,sort map { tr{abdrtwsuiopqe987654}{abcdefghijklmnopqrs}r . "\0" . $_ } @unsorted;

>从技术上讲,sort map { ( pack 'C*',$_ ) . "\0" . $_ } @unsorted;

cmp由scmp运算符实现.

$perl -MO=Concise,-exec -e'$x cmp $y'
1  <0> enter
2  <;> nextstate(main 1 -e:1) v:{
3  <#> gvsv[*x] s
4  <#> gvsv[*y] s
5  <2> scmp[t3] vK/2
6  <@> leave[1 ref] vKP/REFC

scmp操作符由pp.c中的pp_scmp函数实现,当使用locale时,它实际上只是sv.c中sv_cmp_flags的包装器;没有效果. sv_cmp_flags使用C库函数memcmp或UTF-8识别版本(取决于标量的类型).