lost in utf8

[pcouderc]

#!/usr/bin/perl -w use warnings;

use strict;
use Encode;
my $lineo ='azertyuiop';
my $line = encode("utf8",$lineo);
open(TEK,">:utf8","temp.tex") || die("Cannot Open File");
print TEK $line;
print TEK <<EOF;
123456<<<<<
EOF
close TEK;

gives me

0000000   a   z   e   r   t   y   u   i   o   p   1   2   3   4   5   
+6
        61 7a 65 72 74 79 75 69 6f 70 31 32 33 34 35 36
0000020   <   <   <   <   <  nl
        3c 3c 3c 3c 3c 0a

and all my other trials so... How to get utf8 in output? Thank you , o wise emonks..

[almut]

What do you expect? This is UTF-8. Any difference is hard to see, though, as you're only printing ASCII characters... Try printing some non-ASCII characters, and you'll see those characters being encoded as multi-byte sequences.

(See UTF-8 and note that the range U+0000–U+007F (ASCII) encodes as single bytes.)

[pcouderc]

Mmm, yes, thank you, my misundertanding of UTF8...

[ikegami]

By the way, you're double encoding (bad!)

my $line = encode("utf8",$lineo);
open(TEK,">:utf8","temp.tex");
print TEK $line;

should be

my $line = encode("utf8",$lineo);
open(TEK,">","temp.tex");
print TEK $line;

or

my $line = $lineo;
open(TEK,">:utf8","temp.tex");
print TEK $line;

[oha]

Doesn't the second block of code require a binmode() just for compatibility?

[ikegami]

binmode basically disables :crlf and :encoding.

Since neither were used in the second snippet, we're talking about the default layers from the OS, from $ENV{PERLIO} or from use of the open pragma.

While using binmode to remove any default :encoding may be smart, it may accidentally disable the :crlf layer. There's no indication that we want the :crlf status to be different, so it's better to use the last snippet if you're dealing with a text file.

Now, if you're dealing with a binary file, you'd use binmode, and you'd use encode for the text bits. Any LF<->CRLF conversion will have to be handled manually.