#!/usr/bin/perl -w
use warnings;
use strict;
$|++;
#
# Connection script between the Autshumato ITE and Apertium
#
#-- IMPORTS --
use CGI;
use CGI::Carp qw/fatalsToBrowser/;
use Encode;
use WWW::Curl;
use WWW::Curl::Easy;
use JSON;
use lib 'lib';
# API Key: dEQF38hrmBbupvSPcjLJwbg0/FI
my @APERTIUM_ADDRESS = ("api.apertium.org:80");
# clear the path
$ENV{PATH} = '';
# get the String to be translated
# The URL from Autshumato looks like this:
#
# http://129.242.4.42:80/cgi-bin/moses/ite.cgi?tl=nb&sl=se&url=Mun%20lean%20Gogol
#
#my $sentence = decode('UTF-8', CGI->new->param ('url')); # change the variable in html
my $sentence = CGI->new->param ('url'); # change the variable in html
my $target_l = CGI->new->param ('tl'); # target language from Autshumato
my $source_l = CGI->new->param ('sl'); # source language from Autshumato
# This is the string sent to the official Apertium server (for language pairs other than the UiT ones):
#curl 'http://api.apertium.org/json/translate?q=hello%20world&q=bye&langpair=en%7Ces&markUnknown=no'
# %7C = | (vertical bar)
my $string = urlencode($sentence);;
my $pair = $source_l."%7C".$target_l;
my $unknown = "no";
my $curl = new WWW::Curl::Easy;
my $response_body;
my $output = "Dummy text!";
$CGITempFile::TMPDIRECTORY = "/tmp";
my %dir = (
"sme-smj", "/home/fran/source/apertium-sme-smj/",
"fin-sme", "/home/fran/source/apertium-sme-fin/",
"sme-sma", "/home/fran/source/apertium-sme-sma/",
"sme-nob", "/home/fran/source/apertium-sme-nob/",
"nn-nb", "/home/fran/source/apertium-nn-nb/",
"nb-nn", "/home/fran/source/apertium-nn-nb/"
);
# Autshumato ITE is sending off only two-letter codes if they are available:
if( ($pair eq "fi%7Cse") ||
($pair eq "fin%7Csme") ||
($pair eq "se%7Cnb") ||
($pair eq "sme%7Cnob") ||
($pair eq "se%7Csmj") ||
($pair eq "se%7Csma") ||
($pair eq "sme%7Csmj") ||
($pair eq "sme%7Csma") ||
($pair eq "nb%7Cnn") ||
($pair eq "nn%7Cnb") ||
($pair eq "nob%7Cnon") ||
($pair eq "nno%7Cnob") ) {
## victorio stuff
$output = "Victorio answer";
# map to the appropriate language pairs as used in the apertium dirs:
if ($source_l eq "se" && $target_l eq "nb" ) { $pair = "sme-nob"; }
elsif ($source_l eq "se" && $target_l eq "smj") { $pair = "sme-smj"; }
elsif ($source_l eq "se" && $target_l eq "sma") { $pair = "sme-sma"; }
elsif ($source_l eq "fi" && $target_l eq "se" ) { $pair = "fin-sme"; }
else { $pair = "$source_l-$target_l"; }
# else { $output = "No translation supported for the language pair $source_l-$target_l"; }
my $prefix = "PATH=/home/fran/local/bin:/usr/bin:/bin:\$PATH PKG_CONFIG_PATH=/home/fran/local/lib/pkgconfig LD_LIBRARY_PATH=/home/fran/local/lib ";
my $cmd = "echo \"$sentence\" | " . $prefix . "LANG=no_NO.UTF-8 /home/fran/local/bin/apertium -d" . $dir{$pair} . " -u " . $pair;
#print $cmd; # DEBUG
if($sentence ne '') {
$output = decode('UTF-8', `$cmd`);
}
} else {
$curl->setopt(CURLOPT_URL, "http://api.apertium.org/json/translate?q=" . $string . "&langpair=" . $pair . "&markUnknown=" . $unknown );
open (my $fileb, ">", \$response_body);
$curl->setopt(CURLOPT_WRITEDATA,$fileb);
my $retcode = $curl->perform;
if ($retcode == 0) {
# print("Transfer went ok\n");
my $response_code = $curl->getinfo(CURLINFO_HTTP_CODE);
# judge result and next action based on $response_code
# print("Received response: $response_body\n");
} else {
print("An error happened: ".$curl->strerror($retcode)." ($retcode)\n");
}
#translate the string
my $output_arr = from_json($response_body, {utf8 => 1});
$output = $output_arr->{"responseData"}{"translatedText"};
}
# return the string
print "Content-Type: text/html; charset=UTF-8\n\n";
print encode ('UTF-8',
"
Apertium Output
$output
");
sub urlencode {
my $theURL = $_[0];
$theURL =~ s/([\W])/"%" . uc(sprintf("%2.2x",ord($1)))/eg;
return $theURL;
}