Re: [tlug] utf form problems: solution

[steven smith <sjs@example.com>]

David Shanahan wrote:
> Sounds like perl might be treating your string as bytes rather than a
> unicode string,
> then muching them on output.
>
> You should read the docs on the Encode module;
> perldoc Encode
>
> You may need to do
> $string = decode("utf8", param("stuff"));
> to tell perl the %E5%8A%A9 bytes from the form is actually utf8 string
>
> also look at the docs on the -C option ...

I needed -CO for printing to stdout, decode to tell perl 
string was utf8 and 'use utf8' because file contained utf8. 
 This was my minimal debug script for testing the post form.

#!.../perl -CO
# output unicode to stdout.  Input is ascii
use utf8; # file contained utf-8 characters.
use Encode;
use URI::Escape;
my $input;
read(STDIN,$input,$ENV{CONTENT_LENGTH},0);
# Tell perl this is a utf-8 string.
my $unencoded = decode('utf8', uri_unescape($input));
print  "content-type: text/plain charset=UTF-8\n\n",
    "今日は世界\n",  # 日本語と"hello world"
    $unencoded;

Next I work on reading stdin safely.

Thanks everyone!

Steve S.