CLD: Keep the original detected language array

This commit is contained in:
Michael 2023-10-11 18:49:42 +00:00
parent fbafa80815
commit 73c6a0ff0c

View file

@ -69,10 +69,9 @@ function cld_detect_languages(array &$data)
if (!$result['is_reliable']) { if (!$result['is_reliable']) {
Logger::debug('Unreliable detection', ['uri-id' => $data['uri-id'], 'original' => $original, 'detected' => $detected, 'name' => $result['language_name'], 'probability' => $result['language_probability'], 'text' => $data['text']]); Logger::debug('Unreliable detection', ['uri-id' => $data['uri-id'], 'original' => $original, 'detected' => $detected, 'name' => $result['language_name'], 'probability' => $result['language_probability'], 'text' => $data['text']]);
return; if (($original == $detected) && ($data['detected'][$original] < $result['language_probability'] / 100)) {
} $data['detected'][$original] = $result['language_probability'] / 100;
}
if ($original == $detected) {
return; return;
} }
@ -83,6 +82,15 @@ function cld_detect_languages(array &$data)
return; return;
} }
Logger::debug('Detected different language', ['uri-id' => $data['uri-id'], 'original' => $original, 'detected' => $detected, 'name' => $result['language_name'], 'probability' => $result['language_probability'], 'text' => $data['text']]); if ($original != $detected) {
$data['detected'] = [$detected => $result['language_probability'] / 100]; Logger::debug('Detected different language', ['uri-id' => $data['uri-id'], 'original' => $original, 'detected' => $detected, 'name' => $result['language_name'], 'probability' => $result['language_probability'], 'text' => $data['text']]);
}
$length = count($data['detected']);
if ($length > 0) {
unset($data['detected'][$detected]);
$data['detected'] = array_merge([$detected => $result['language_probability'] / 100], array_slice($data['detected'], 0, $length - 1));
} else {
$data['detected'] = [$detected => $result['language_probability'] / 100];
}
} }