use strict;
use warnings;
use utf8;
use Future::AsyncAwait;
use Encode qw(decode_utf8 encode_utf8);
# =============================================================================
# PAGI UTF-8 Round-Trip Test Demo
# =============================================================================
#
# Tests UTF-8 handling across all PAGI input vectors:
#
# 1. PATH TEST: /echo/{utf8_string}
# - $scope->{path}: decoded Unicode characters (per PAGI spec)
# - $scope->{raw_path}: percent-encoded bytes from wire
#
# 2. QUERY STRING TEST: ?text={utf8_string}
# - $scope->{query_string}: percent-encoded bytes (app decodes)
#
# 3. POST BODY TEST: form submit with textarea
# - Request body: percent-encoded bytes (application/x-www-form-urlencoded)
# - App must: percent-decode → decode_utf8 → characters
#
# 4. RESPONSE TEST: UTF-8 literals in source
# - App must: encode_utf8 → bytes for wire
# - Content-Length must be byte count, not character count
#
# =============================================================================
my $app = async sub {
my ($scope, $receive, $send) = @_;
die "Unsupported scope type: $scope->{type}" if $scope->{type} ne 'http';
# Collect request body
my $body = '';
while (1) {
my $message = await $receive->();
$body .= $message->{body} // '';
last unless $message->{more};
}
my $echo = '';
my $source = '';
# 1. PATH: Check for /echo/{value}
if ($scope->{path} =~ m{^/echo/(.+)$}) {
$echo = $1; # Already decoded per PAGI spec
$source = 'path';
}
# 2. QUERY STRING: ?text={value}
elsif (length($scope->{query_string} // '') && $scope->{query_string} =~ /text=([^&]*)/) {
$echo = _uri_decode($1);
$source = 'query string';
}
# 3. POST BODY: text={value}
elsif (length $body && $body =~ /text=(.*)/) {
$echo = _uri_decode($1);
$source = 'POST body';
}
# Build echo section
my $echo_section = '';
if (length $echo) {
my $chars = length($echo);
my $bytes = length(encode_utf8($echo));
my $codepoints = join ' ', map { sprintf 'U+%04X', ord($_) } split //, $echo;
$echo_section = <<"ECHO";
Echo (from $source):
$echo
Characters: $chars | UTF-8 bytes: $bytes
Codepoints: $codepoints
ECHO
}
# Build path test links
my @path_samples = ('λ', '🔥', '中文', '♥', 'café');
my $path_links = join ' | ', map {
my $encoded = _uri_encode($_);
qq{$_}
} @path_samples;
my $html = <<"HTML";
PAGI UTF-8
PAGI UTF-8 Round-Trip Test
Samples: λ (Greek) | 🔥 (Emoji) | 中文 (CJK) | ♥ (Symbol) | café (Accented)
$echo_section
1. Path Test
Click: $path_links
Tests \$scope->{path} (decoded) vs \$scope->{raw_path} (bytes)
2. Query String Test
Tests \$scope->{query_string} (percent-encoded bytes)
3. POST Body Test
Tests request body (application/x-www-form-urlencoded)
HTML
my $bytes = encode_utf8($html);
await $send->({
type => 'http.response.start',
status => 200,
headers => [
['content-type', 'text/html; charset=utf-8'],
['content-length', length($bytes)],
],
});
await $send->({
type => 'http.response.body',
body => $bytes,
more => 0,
});
};
sub _uri_decode {
my ($str) = @_;
$str =~ s/\+/ /g;
$str =~ s/%([0-9A-Fa-f]{2})/chr(hex($1))/eg;
return decode_utf8($str);
}
sub _uri_encode {
my ($str) = @_;
my $bytes = encode_utf8($str);
$bytes =~ s/([^A-Za-z0-9\-._~])/sprintf("%%%02X", ord($1))/eg;
return $bytes;
}
$app;