use Encode;
use IO::Handle;
STDOUT->autoflush(1);
STDERR->autoflush(1);

my $SRC = "D:\\Youdao\\Source";  #包含离线字典文件的路径
my $DST = $SRC;                         #输出目录和离线字典位置相同

my $YDIC;
my $IDX;
my $IDX_DT;
my $WRT;
my $EXT1;

open $YDIC,  "<:raw", "$SRC\\basicCE.ydic" or die $!;
open $IDX,   "<:raw", "$SRC\\CE.idx" or die $!;
open $IDXDT, "<:raw", "$SRC\\CE.idx" or die $!;
open $EXT1,  "<:raw", "$SRC\\newCE.ydic" or die "$!";

open $WRT,   ">:raw", "$DST\\CE_plaintext.txt";

my $edge = 176382 * 4 * 3;  #单词量 * 12字节索引信息

seek($IDXDT, $edge, 0);
my @idx;

=idx_format
    （以下每个数值为 LONG 型，包含4个字节）
     0  0 22  第一个词汇起点  译文起点  译文长度
    11 22 34  第二个单词起点  译文起点  译文长度
    18 56 28 
    29 84 34 
=cut

seek($IDX, 4, 0);  #第一个单词起点信息skip，以第二个起点作为有效数据
my $prev_w = 0;
my $prev_t = 0;
my $buff;
my $word;
my $trans;

do 
{
    for (0..2)
    {
        read($IDX, $buff, 4, 0);
        $idx[$_] = unpack "L", $buff;
        #printf "%7d ", $idx[$_];
    }
    #printf "\n";

    read($IDXDT, $word, $idx[2] - $prev_w , 0);
    print $WRT word_decode( $word ),"\t";

    seek($YDIC, $idx[0], 0);
    read($YDIC, $trans, $idx[1], 0);
    $trans = get_ex_trans( word_decode( $trans ) );

    #格式化数据结构
    to_perl_struct( \$trans );

    print $WRT $trans ,"\n";
    $prev_w = $idx[2];
}
until ( tell($IDX) >= $edge - 12 );

=decrease
    8094027     121 1621295 
    8094148     121 -1460395742  最后一段译文起点  译文长度  词汇起点内容
    最后一个词汇只有起点信息，没有结束信息，且其本身设计为：和倒数第二词
    汇重复，可以略过
=cut


close $YDIC;
close $IDX;
close $IDXDT;
close $WRT;

#扩展翻译，获取并替换
sub get_ex_trans
{
    my $trans = shift;
    my $buff;
    my $ex_trans;
    my ($info_A, $dict, $offset, $len);

                  #字典代号, offset, 长度
    while ($trans =~/"(\d)#(\d+)#(\d+)"/)   #循环替换，如果出现多个
    {
        ($info_A, $dict, $offset, $len) = ( $&, $1, (sprintf "%08d", $2), $3 );
        seek($EXT1, $offset, 0);
        read($EXT1, $buff, $len, 0);

        $ex_trans = word_decode( $buff );
        $trans=~s/${info_A}/${ex_trans}/;
    }

    ex_format(\$trans);

    return $trans;
}

#清理"ex":{ ... } ，保留花括号内的内容
sub ex_format
{
    my $ref = shift;
    #${$ref}=~s/\"ex\"\:\{(.*)\}/$1/g;
    ${$ref}=~s/##//g;
}


#简单解密
sub word_decode
{
    my $word = shift;
    my $real = "";
    my $code;
    grep
    {
        $code = ord( $_ );
        $real .= chr( $code + !( ($code >> 4) % 2) * 0x20 - 0x10 );
                            #高位 % 2 = 0 或 1 => * 0x20 - 0x10 = 0x10 或 -0x10
    } 
    split("", $word);

    return $real;
}

# 转为Perl数据格式
sub to_perl_struct
{
    my $ref = shift;
    #$$ref =~s/\Q},"ex":{\E/, /g;
    $$ref =~s/":/" => /g;
    $$ref =~s/",/", /g;
    $$ref =~s/\'/\\\'/g;
    $$ref =~s/\$/\\\$/g;
    $$ref =~s/\@/\\\@/g;
}

sub xcode {
    $_[1]='x' if (not defined $_[1]);
    for my $v ( split(//,$_[0]) ) {
        print sprintf ("%l$_[1] ",ord($v)); 
    }
    print "\n\n";
}

C:\>ce.pl 单词
[dān cí]
[语] word
seperate word
single word
discrete word

[dān cí]
◎{语} （词） individual word; word （区别于“词组”）
◎（由一个词素构成的词; 单纯词） single-morpheme word （如“茶”、“走”、“蟋蟀”、“玲珑”等， 区别于“合成词”）
◎{计} （字码） word

=info
    调用参数：输入包含空格的词组时，使用双引号即可
=cut

use Encode;
use YAML 'Dump';

our $PATH = "D:\\Local\\Dict\\Youdao\\Analyse_CE";

open $DICT,"<:raw", "$PATH\\CE_plaintext.txt" or die;

our %hash;
my $word;
my $i = 0;
my $info;
my $dt;
my $first;

$word = defined $ARGV[0] ? gbk_to_utf8( $ARGV[0] ) : "缓";

while ($line = <$DICT>)
{
    if ( $line=~/^$word\t(.*)/ )   #如果要显示大小写不同时的翻译结果，加上/i
    {
        $info = $1;
        eval "\$dt = $info";
        short_print( $dt );
        undef $dt;
    }
}
exit;

sub short_print
{
    my $ref = shift;
    my $indent = "  ";

PRINT:
    printf "[%s]\n", utf8_to_gbk( $ref->{'p'} ) if (exists $ref->{'p'});
    printf "[%s]\n", utf8_to_gbk( $ref->{'w'} ) if (exists $ref->{'w'});
    for my $v ( @{ $ref->{'e'} } )
    {
        print $indent, utf8_to_gbk( $v ),"\n";
    }

    if (exists $ref->{'ex'})
    {
        $ref = $ref->{'ex'};
        $indent .= "  ";
        print "\n";
        goto PRINT;
    }

    return;
}

sub utf8_to_gbk
{
    return encode('gbk', decode('utf8', $_[0]));
}

sub gbk_to_utf8
{
    return encode('utf8', decode('gbk', $_[0]));
}