You don't show us any code we could use to replicate your problem.

Maybe $file does not contain the file contents?

I recommend saving files as UTF8 and reading them as raw bytes. JSON modules expect either raw bytes (JSON::XS) or can be configured to accept Latin-1 (JSON).

Hi,

Sorry - the code is just being grabbed using wget:

`wget -O/srv/www/site.net/www/cgi-bin/admin/tmp/in.txt 'https://openapi.etsy.com/v2/shops/Syrestria/listings/active?method=GET&api_key=xxxxx&limit=200&includes=MainImage'`;

..and a basic script I wrote, does:



#!/usr/bin/perl

use File::Slurp;
use Encode;

my $file = read_file("./in.txt");

$file =~ s/\\u(....)/chr hex $1/ge;

print "$file\n";
[download]



However, as I explained that does not work well :) (some get encoded, but the vast majority do not)

Are you suggesting I do something like this?



use File::Slurp;
use Encode;
use JSON;
use Unicode::MapUTF8 qw(to_utf8 from_utf8 utf8_supported_charset);

my $file = read_file("./in.txt");

my $json_var = decode_json($file);

foreach (@{$json_var->{results}}) {
    $_->{description} =~ s/([\200-\377]+)/from_utf8({ -string => $1, -
+charset => 'ISO-8859-1'})/eg;
    print "BLA - $_->{description} \n";
}
[download]



Cheers

Andy

No. I'm suggesting that you use a JSON module for loading JSON data. There should be no need at least with the two JSON modules I mentioned to manually convert \uXXXX to their Unicode equivalents.

use JSON;
use Data::Dumper;
$Data::Dumper::Useqq = 1;
my $data = decode_json( $file_content );
warn Dumper $data;
[download]

Note that File::Slurp is horribly broken regarding encodings. Some comments recommend File::Slurper, but I instead roll my own, which isn't rocket surgery either.