68 lines
1.5 KiB
PHP
68 lines
1.5 KiB
PHP
<?php
|
|
|
|
declare(strict_types=1);
|
|
|
|
namespace App\Config;
|
|
|
|
/**
|
|
* YAML-backed stop-word configuration.
|
|
*
|
|
* This class intentionally has no PHP fallback list. The complete list lives in
|
|
* config/retriex/language.yaml.
|
|
*/
|
|
final class StopWordsConfig
|
|
{
|
|
/**
|
|
* @param array<string, mixed> $config
|
|
*/
|
|
public function __construct(private readonly array $config)
|
|
{
|
|
}
|
|
|
|
/**
|
|
* @return string[]
|
|
*/
|
|
public function getStopWords(): array
|
|
{
|
|
return $this->requiredStringList('words');
|
|
}
|
|
|
|
/**
|
|
* @return string[]
|
|
*/
|
|
private function requiredStringList(string $key): array
|
|
{
|
|
if (!array_key_exists($key, $this->config)) {
|
|
throw new \InvalidArgumentException(sprintf('Missing required RetrieX stopwords config key "%s".', $key));
|
|
}
|
|
|
|
$value = $this->config[$key];
|
|
|
|
if (!is_array($value)) {
|
|
throw new \InvalidArgumentException(sprintf('RetrieX stopwords config key "%s" must be a list.', $key));
|
|
}
|
|
|
|
$out = [];
|
|
foreach ($value as $item) {
|
|
if (!is_scalar($item)) {
|
|
continue;
|
|
}
|
|
|
|
$item = trim((string) $item);
|
|
if ($item === '') {
|
|
continue;
|
|
}
|
|
|
|
if (!in_array($item, $out, true)) {
|
|
$out[] = $item;
|
|
}
|
|
}
|
|
|
|
if ($out === []) {
|
|
throw new \InvalidArgumentException(sprintf('RetrieX stopwords config key "%s" must not be empty.', $key));
|
|
}
|
|
|
|
return $out;
|
|
}
|
|
}
|