how to convert from gbk encoding to utf-8 encoding in Perl

I have a simple question which I do not know how to solve in Perl. I know how to convert from utf-8 to GBK, for example, from e4b8ad to d6d0. 开发者_如何转开发But I am not sure how to go backward, i.e. given d6d0, how do I know e4b8ad.

Please enlighten me! Many thanks.

When you have hex digits, pack is your friend. Following is a REPL session. Notes:

• To reverse the direction, pack the hex digits into octets, decode from GB octets to character string, encode character string to UTF-8 octets, unpack octets into hex digits.
• GBK is superseded. Use of GB18030 (provided by Encode::HanExtra in Perl) has been mandatory for five years already.

$ use Encode qw(decode encode); use Encode::HanExtra; use Devel::Peek qw(Dump);

$ 'e4b8ad'
e4b8ad                                  # hex digits

$ pack('H*', 'e4b8ad')
中

$ Dump(pack('H*', 'e4b8ad'))
SV = PV(0x3657680) at 0x36b7188
  REFCNT = 1
  FLAGS = (PADTMP,POK,pPOK)
  PV = 0x36c0768 "\344\270\255"\0           # octets of UTF-8 encoded data
  CUR = 3
  LEN = 8

$ decode('UTF-8', pack('H*', 'e4b8ad'))
中

$ Dump(decode('UTF-8', pack('H*', 'e4b8ad')))
SV = PV(0x326c3a0) at 0x36a50c8
  REFCNT = 1
  FLAGS = (TEMP,POK,pPOK,UTF8)
  PV = 0x3698a48 "\344\270\255"\0 [UTF8 "\x{4e2d}"]     # character string
  CUR = 3
  LEN = 8

$ encode('GB18030', decode('UTF-8', pack('H*', 'e4b8ad')))
"\xd6\xd0"

$ Dump(encode('GB18030', decode('UTF-8', pack('H*', 'e4b8ad'))))
SV = PV(0x36a2da0) at 0x36b6d98
  REFCNT = 1
  FLAGS = (TEMP,POK,pPOK)
  PV = 0x36db3e8 "\326\320"\0               # octets of GB18030 encoded data
  CUR = 2
  LEN = 8

$ unpack('H*', encode('GB18030', decode('UTF-8', pack('H*', 'e4b8ad'))))
d6d0                            # hex digits

The answer to the question asked:

use Encode qw( from_to );
my $gbk = "\xD6\xD0";
from_to(my $utf8 = $gbk, 'GB18030', 'UTF-8');  # E4 B8 AD

or

use Encode qw( decode encode );
my $gbk = "\xD6\xD0";
my $utf8 = encode('UTF-8', decode('GB18030', $gbk));  # E4 B8 AD

However, a more normal flow looks like the following:

open(my $fh_in,  '<:encoding(GB18030)', ...) or die ...;
open(my $fh_out, '>:encoding(UTF-8)',   ...) or die ...;

while (<$fh_in>) {
   ...
   print $fh_out ...;
   ...
}

Encode::HanExtra must be installed for Encode to find the encoding.

use Encode qw/encode decode/;
$utf8 = decode("euc-cn", $euc_cn); # ditto

You can also normally specify the encoding when you open or close a FD and it will perform necessary conversions.

Works like a charm:

perl -e 'open(X,">","/tmp/x"); print X chr(0xd6).chr(0xd0);close(X)'
perl -mEncode -e 'open(X,"<","/tmp/x"); $x=<X>; print Encode::decode("euc-cn",$x);' > /tmp/xx