last step
This commit is contained in:
team2
@@ -4,69 +4,24 @@ declare(strict_types=1);
|
||||
|
||||
namespace App\Config;
|
||||
|
||||
/**
|
||||
* YAML-backed shop search-repair configuration.
|
||||
*
|
||||
* This class intentionally has no PHP fallback values. Missing or invalid
|
||||
* configuration must be fixed in config/retriex/search_repair.yaml and
|
||||
* config/retriex/vocabulary.yaml.
|
||||
*/
|
||||
final class SearchRepairConfig
|
||||
{
|
||||
private const GENERIC_CANDIDATE_TOKENS = [
|
||||
'wasser',
|
||||
'messgerät',
|
||||
'messgeraet',
|
||||
'produkt',
|
||||
'geräte',
|
||||
'geraete',
|
||||
'gerät',
|
||||
'geraet',
|
||||
'resthärte',
|
||||
'resthaerte',
|
||||
'preis',
|
||||
'infos',
|
||||
'wissen',
|
||||
];
|
||||
|
||||
private const ACCESSORY_CANDIDATE_TERMS = [
|
||||
'indikator',
|
||||
'indicator',
|
||||
'reagenz',
|
||||
'reagent',
|
||||
'kit',
|
||||
'set',
|
||||
];
|
||||
|
||||
private const ACCESSORY_OR_BUNDLE_TERMS = [
|
||||
'passend',
|
||||
'passende',
|
||||
'zubehor',
|
||||
'zubehör',
|
||||
'dazu',
|
||||
'zusatz',
|
||||
'erganzung',
|
||||
'ergänzung',
|
||||
'indikator',
|
||||
'reagenz',
|
||||
'kit',
|
||||
'set',
|
||||
'auch\s+das',
|
||||
'mit\s+preis\s+und\s+allen\s+infos',
|
||||
];
|
||||
|
||||
private const SPECIFICITY_BOOST_TERMS = [
|
||||
'indikator',
|
||||
'indicator',
|
||||
'testomat',
|
||||
'tritromat',
|
||||
'titromat',
|
||||
'reagenz',
|
||||
'reagent',
|
||||
];
|
||||
|
||||
/**
|
||||
* @param array<string, mixed> $config
|
||||
*/
|
||||
public function __construct(
|
||||
private readonly bool $enabled = true,
|
||||
private readonly int $maxRepairQueries = 3,
|
||||
private readonly int $minPrimaryResultsWithoutRepair = 2,
|
||||
private readonly array $config = [],
|
||||
private readonly ?DomainVocabularyConfig $vocabulary = null,
|
||||
private readonly bool $enabled,
|
||||
private readonly int $maxRepairQueries,
|
||||
private readonly int $minPrimaryResultsWithoutRepair,
|
||||
private readonly array $config,
|
||||
private readonly DomainVocabularyConfig $vocabulary,
|
||||
) {
|
||||
}
|
||||
|
||||
@@ -87,257 +42,259 @@ final class SearchRepairConfig
|
||||
|
||||
public function shouldRestrictRequestedAccessoryCodeRepair(): bool
|
||||
{
|
||||
return $this->bool('strict_requested_accessory_code_repair', true);
|
||||
return $this->requiredBool('strict_requested_accessory_code_repair');
|
||||
}
|
||||
|
||||
public function shouldPreferPromptAnchoredModelForRequestedAccessoryCode(): bool
|
||||
{
|
||||
return $this->bool('prefer_prompt_anchored_model_for_requested_accessory_code', true);
|
||||
return $this->requiredBool('prefer_prompt_anchored_model_for_requested_accessory_code');
|
||||
}
|
||||
|
||||
/** @return string[] */
|
||||
public function getRequestedAccessoryCodeFallbackQueryTemplates(): array
|
||||
{
|
||||
return $this->stringList(
|
||||
'requested_accessory_code_fallback_query_templates',
|
||||
['{term} {code}']
|
||||
);
|
||||
return $this->requiredStringList('requested_accessory_code_fallback_query_templates');
|
||||
}
|
||||
|
||||
/** @return string[] */
|
||||
public function getRequestedAccessoryCodeFallbackTerms(): array
|
||||
{
|
||||
return $this->stringList(
|
||||
'requested_accessory_code_fallback_terms',
|
||||
$this->getAccessoryCandidateTerms()
|
||||
);
|
||||
return $this->requiredStringList('requested_accessory_code_fallback_terms');
|
||||
}
|
||||
|
||||
/** @return string[] */
|
||||
public function getRequestedAccessoryCodeContextPrefixTerms(): array
|
||||
{
|
||||
return $this->stringList(
|
||||
'requested_accessory_code_context_prefix_terms',
|
||||
$this->getAccessoryCandidateTerms()
|
||||
);
|
||||
return $this->requiredStringList('requested_accessory_code_context_prefix_terms');
|
||||
}
|
||||
|
||||
public function getRequestedAccessoryCodeProximityWindow(): int
|
||||
{
|
||||
return $this->int('requested_accessory_code_proximity_window', 1600);
|
||||
return $this->requiredPositiveInt('requested_accessory_code_proximity_window');
|
||||
}
|
||||
|
||||
/** @return string[] */
|
||||
public function getSpecificModelCandidatePatterns(): array
|
||||
{
|
||||
return $this->stringList(
|
||||
'specific_model_candidate_patterns',
|
||||
[$this->getModelLikePattern()]
|
||||
);
|
||||
return $this->requiredStringList('specific_model_candidate_patterns');
|
||||
}
|
||||
|
||||
/** @return string[] */
|
||||
public function getModelCandidateExcludeTerms(): array
|
||||
{
|
||||
return $this->stringList(
|
||||
'model_candidate_exclude_terms',
|
||||
array_merge($this->getAccessoryCandidateTerms(), ['verfuegbarkeit', 'verfügbarkeit', 'shop'])
|
||||
);
|
||||
return $this->requiredStringList('model_candidate_exclude_terms');
|
||||
}
|
||||
|
||||
public function getTopProductLogLimit(): int
|
||||
{
|
||||
return 3;
|
||||
return $this->requiredNonNegativeInt('limits.top_product_log_limit');
|
||||
}
|
||||
|
||||
public function getModelCandidatePattern(): string
|
||||
{
|
||||
return '/\b([A-Za-zÄÖÜäöüß][A-Za-zÄÖÜäöüß®\-]*(?:\s+[A-Za-zÄÖÜäöüß][A-Za-zÄÖÜäöüß®\-]*){0,2}\s+\d{2,5}[A-Za-z0-9\-]*)\b/u';
|
||||
return $this->requiredString('patterns.model_candidate');
|
||||
}
|
||||
|
||||
public function getAccessoryCandidatePattern(): string
|
||||
{
|
||||
return '/\b((?:' . implode('|', $this->getAccessoryCandidateTerms()) . ')\s+\d{1,5}[A-Za-z0-9\-]*)\b/iu';
|
||||
return $this->renderPatternTemplate(
|
||||
'patterns.accessory_candidate_template',
|
||||
['terms' => $this->patternAlternation($this->getAccessoryCandidateTerms())]
|
||||
);
|
||||
}
|
||||
|
||||
public function getRequestedAccessoryCodePattern(): string
|
||||
{
|
||||
$fallbackTerms = array_map(
|
||||
static fn(string $term): string => preg_quote($term, '/'),
|
||||
$this->getRequestedAccessoryCodeContextPrefixTerms()
|
||||
);
|
||||
$fallbackTerms = array_filter($fallbackTerms, static fn(string $term): bool => $term !== '');
|
||||
|
||||
$fallbackPattern = $fallbackTerms === []
|
||||
? '/\b([A-Za-z]{0,3}\s*\d{1,5}[A-Za-z0-9\-]*)\b/iu'
|
||||
: '/\b(?:' . implode('|', $fallbackTerms) . ')\s*([A-Za-z]{0,3}\s*\d{1,5}[A-Za-z0-9\-]*)\b/iu';
|
||||
|
||||
return $this->string('requested_accessory_code_pattern', $fallbackPattern);
|
||||
return $this->requiredString('patterns.requested_accessory_code');
|
||||
}
|
||||
|
||||
public function getAccessoryOrBundlePattern(): string
|
||||
{
|
||||
return '/\b(' . implode('|', $this->getAccessoryOrBundleTerms()) . ')\b/iu';
|
||||
return $this->renderPatternTemplate(
|
||||
'patterns.accessory_or_bundle_template',
|
||||
['terms' => $this->patternAlternation($this->getAccessoryOrBundleTerms())]
|
||||
);
|
||||
}
|
||||
|
||||
public function getModelLikePattern(): string
|
||||
{
|
||||
return '/\b[A-Za-zÄÖÜäöüß][A-Za-zÄÖÜäöüß®\-]*(?:\s+[A-Za-zÄÖÜäöüß][A-Za-zÄÖÜäöüß®\-]*){0,2}\s+\d{2,5}[A-Za-z0-9\-]*\b/u';
|
||||
return $this->requiredString('patterns.model_like');
|
||||
}
|
||||
|
||||
public function getSpecificityBoostPattern(): string
|
||||
{
|
||||
return '/\b(?:' . implode('|', $this->getSpecificityBoostTerms()) . ')\b/iu';
|
||||
return $this->renderPatternTemplate(
|
||||
'patterns.specificity_boost_template',
|
||||
['terms' => $this->patternAlternation($this->getSpecificityBoostTerms())]
|
||||
);
|
||||
}
|
||||
|
||||
/** @return string[] */
|
||||
public function getGenericCandidateTokens(): array
|
||||
{
|
||||
return $this->stringList(
|
||||
return $this->configOrVocabularyStringList(
|
||||
'generic_candidate_tokens',
|
||||
$this->vocabularyView('search_repair.generic_candidate_tokens', self::GENERIC_CANDIDATE_TOKENS)
|
||||
'search_repair.generic_candidate_tokens'
|
||||
);
|
||||
}
|
||||
|
||||
public function getSanitizeTrimCharacters(): string
|
||||
{
|
||||
return " \t\n\r\0\x0B\"'`.,;:-";
|
||||
return implode('', array_map(
|
||||
static fn (int $code): string => chr($code),
|
||||
$this->requiredCharacterCodes('sanitize_trim_character_codes')
|
||||
));
|
||||
}
|
||||
|
||||
public function getContainsDigitPattern(): string
|
||||
{
|
||||
return '/\d/u';
|
||||
return $this->requiredString('patterns.contains_digit');
|
||||
}
|
||||
|
||||
public function getWhitespaceCollapsePattern(): string
|
||||
{
|
||||
return '/\s+/u';
|
||||
return $this->requiredString('patterns.whitespace_collapse');
|
||||
}
|
||||
|
||||
public function getTokenizeCleanupPattern(): string
|
||||
{
|
||||
return '/[^\p{L}\p{N}\s\-]+/u';
|
||||
return $this->requiredString('patterns.tokenize_cleanup');
|
||||
}
|
||||
|
||||
public function getProductKeySeparator(): string
|
||||
{
|
||||
return '|';
|
||||
return $this->requiredString('product_key_separator');
|
||||
}
|
||||
|
||||
public function getCandidateDigitScore(): int
|
||||
{
|
||||
return 4;
|
||||
return $this->requiredInt('scores.candidate_digit');
|
||||
}
|
||||
|
||||
public function getCandidateWordCountCap(): int
|
||||
{
|
||||
return 4;
|
||||
return $this->requiredPositiveInt('scores.candidate_word_count_cap');
|
||||
}
|
||||
|
||||
public function getSpecificityBoostScore(): int
|
||||
{
|
||||
return 3;
|
||||
return $this->requiredInt('scores.specificity_boost');
|
||||
}
|
||||
|
||||
public function getPrimaryQueryOverlapThreshold(): float
|
||||
{
|
||||
return 0.9;
|
||||
return $this->requiredFloat('scores.primary_query_overlap_threshold');
|
||||
}
|
||||
|
||||
public function getPromptMatchWeight(): int
|
||||
{
|
||||
return 3;
|
||||
return $this->requiredInt('scores.prompt_match_weight');
|
||||
}
|
||||
|
||||
public function getPrimaryQueryMatchWeight(): int
|
||||
{
|
||||
return 2;
|
||||
return $this->requiredInt('scores.primary_query_match_weight');
|
||||
}
|
||||
|
||||
public function getRepairSignalMatchWeight(): int
|
||||
{
|
||||
return 4;
|
||||
return $this->requiredInt('scores.repair_signal_match_weight');
|
||||
}
|
||||
|
||||
public function getPrimaryResultOrderBonus(): int
|
||||
{
|
||||
return 1;
|
||||
return $this->requiredInt('scores.primary_result_order_bonus');
|
||||
}
|
||||
|
||||
public function getTokenIntersectionScore(): int
|
||||
{
|
||||
return 2;
|
||||
return $this->requiredInt('scores.token_intersection_score');
|
||||
}
|
||||
|
||||
public function getNumericTokenMatchScore(): int
|
||||
{
|
||||
return 4;
|
||||
return $this->requiredInt('scores.numeric_token_match_score');
|
||||
}
|
||||
|
||||
/** @return string[] */
|
||||
public function getAccessoryCandidateTerms(): array
|
||||
{
|
||||
return $this->stringList(
|
||||
return $this->configOrVocabularyStringList(
|
||||
'accessory_candidate_terms',
|
||||
$this->vocabularyView('search_repair.accessory_candidate_terms', self::ACCESSORY_CANDIDATE_TERMS)
|
||||
'search_repair.accessory_candidate_terms'
|
||||
);
|
||||
}
|
||||
|
||||
/** @return string[] */
|
||||
public function getAccessoryOrBundleTerms(): array
|
||||
{
|
||||
return $this->stringList(
|
||||
return $this->configOrVocabularyStringList(
|
||||
'accessory_or_bundle_terms',
|
||||
$this->vocabularyView('search_repair.accessory_or_bundle_terms', self::ACCESSORY_OR_BUNDLE_TERMS)
|
||||
'search_repair.accessory_or_bundle_terms'
|
||||
);
|
||||
}
|
||||
|
||||
/** @return string[] */
|
||||
public function getSpecificityBoostTerms(): array
|
||||
{
|
||||
return $this->stringList(
|
||||
return $this->configOrVocabularyStringList(
|
||||
'specificity_boost_terms',
|
||||
$this->vocabularyView('search_repair.specificity_boost_terms', self::SPECIFICITY_BOOST_TERMS)
|
||||
'search_repair.specificity_boost_terms'
|
||||
);
|
||||
}
|
||||
|
||||
/** @return string[] */
|
||||
private function vocabularyView(string $path, array $fallback): array
|
||||
private function configOrVocabularyStringList(string $configKey, string $vocabularyPath): array
|
||||
{
|
||||
return $this->vocabulary?->view($path, $fallback) ?? $fallback;
|
||||
}
|
||||
|
||||
private function string(string $key, string $default): string
|
||||
{
|
||||
$value = $this->config[$key] ?? $default;
|
||||
|
||||
if (!is_scalar($value)) {
|
||||
return $default;
|
||||
if (array_key_exists($configKey, $this->config)) {
|
||||
return $this->requiredStringList($configKey);
|
||||
}
|
||||
|
||||
$value = trim((string) $value);
|
||||
|
||||
return $value !== '' ? $value : $default;
|
||||
}
|
||||
|
||||
private function int(string $key, int $default): int
|
||||
{
|
||||
$value = $this->config[$key] ?? $default;
|
||||
|
||||
if (is_int($value)) {
|
||||
return $value;
|
||||
$items = $this->vocabulary->view($vocabularyPath, []);
|
||||
if ($items === []) {
|
||||
throw new \InvalidArgumentException(sprintf(
|
||||
'Missing required RetrieX search repair vocabulary view "%s".',
|
||||
$vocabularyPath
|
||||
));
|
||||
}
|
||||
|
||||
if (is_numeric($value)) {
|
||||
return (int) $value;
|
||||
}
|
||||
|
||||
return $default;
|
||||
return $items;
|
||||
}
|
||||
|
||||
private function bool(string $key, bool $default): bool
|
||||
/** @param array<string, string> $variables */
|
||||
private function renderPatternTemplate(string $path, array $variables): string
|
||||
{
|
||||
$value = $this->config[$key] ?? $default;
|
||||
$template = $this->requiredString($path);
|
||||
|
||||
foreach ($variables as $key => $value) {
|
||||
$template = str_replace('{' . $key . '}', $value, $template);
|
||||
}
|
||||
|
||||
if (preg_match('/\{[A-Za-z_][A-Za-z0-9_]*\}/', $template) === 1) {
|
||||
throw new \InvalidArgumentException(sprintf('RetrieX search repair pattern template "%s" contains unresolved placeholders.', $path));
|
||||
}
|
||||
|
||||
return $template;
|
||||
}
|
||||
|
||||
/** @param string[] $terms */
|
||||
private function patternAlternation(array $terms): string
|
||||
{
|
||||
$terms = array_values(array_filter(
|
||||
array_map(static fn (string $term): string => trim($term), $terms),
|
||||
static fn (string $term): bool => $term !== ''
|
||||
));
|
||||
|
||||
if ($terms === []) {
|
||||
throw new \InvalidArgumentException('RetrieX search repair pattern alternation requires at least one term.');
|
||||
}
|
||||
|
||||
return implode('|', $terms);
|
||||
}
|
||||
|
||||
private function requiredBool(string $path): bool
|
||||
{
|
||||
$value = $this->requiredValue($path);
|
||||
|
||||
if (is_bool($value)) {
|
||||
return $value;
|
||||
@@ -357,15 +314,80 @@ final class SearchRepairConfig
|
||||
}
|
||||
}
|
||||
|
||||
return $default;
|
||||
throw new \InvalidArgumentException(sprintf('RetrieX search repair config key "%s" must be boolean.', $path));
|
||||
}
|
||||
|
||||
private function requiredString(string $path): string
|
||||
{
|
||||
$value = $this->requiredValue($path);
|
||||
|
||||
if (!is_scalar($value)) {
|
||||
throw new \InvalidArgumentException(sprintf('RetrieX search repair config key "%s" must be a string.', $path));
|
||||
}
|
||||
|
||||
$value = trim((string) $value);
|
||||
if ($value === '') {
|
||||
throw new \InvalidArgumentException(sprintf('RetrieX search repair config key "%s" must not be empty.', $path));
|
||||
}
|
||||
|
||||
return $value;
|
||||
}
|
||||
|
||||
private function requiredInt(string $path): int
|
||||
{
|
||||
$value = $this->requiredValue($path);
|
||||
|
||||
if (is_int($value)) {
|
||||
return $value;
|
||||
}
|
||||
|
||||
if (is_string($value) && preg_match('/^-?\d+$/', trim($value)) === 1) {
|
||||
return (int) trim($value);
|
||||
}
|
||||
|
||||
throw new \InvalidArgumentException(sprintf('RetrieX search repair config key "%s" must be an integer.', $path));
|
||||
}
|
||||
|
||||
private function requiredNonNegativeInt(string $path): int
|
||||
{
|
||||
$value = $this->requiredInt($path);
|
||||
|
||||
if ($value < 0) {
|
||||
throw new \InvalidArgumentException(sprintf('RetrieX search repair config key "%s" must be greater than or equal to 0.', $path));
|
||||
}
|
||||
|
||||
return $value;
|
||||
}
|
||||
|
||||
private function requiredPositiveInt(string $path): int
|
||||
{
|
||||
$value = $this->requiredInt($path);
|
||||
|
||||
if ($value <= 0) {
|
||||
throw new \InvalidArgumentException(sprintf('RetrieX search repair config key "%s" must be greater than 0.', $path));
|
||||
}
|
||||
|
||||
return $value;
|
||||
}
|
||||
|
||||
private function requiredFloat(string $path): float
|
||||
{
|
||||
$value = $this->requiredValue($path);
|
||||
|
||||
if (is_int($value) || is_float($value) || (is_string($value) && is_numeric(trim($value)))) {
|
||||
return (float) $value;
|
||||
}
|
||||
|
||||
throw new \InvalidArgumentException(sprintf('RetrieX search repair config key "%s" must be numeric.', $path));
|
||||
}
|
||||
|
||||
/** @return string[] */
|
||||
private function stringList(string $key, array $default): array
|
||||
private function requiredStringList(string $path): array
|
||||
{
|
||||
$value = $this->config[$key] ?? $default;
|
||||
$value = $this->requiredValue($path);
|
||||
|
||||
if (!is_array($value)) {
|
||||
return $default;
|
||||
throw new \InvalidArgumentException(sprintf('RetrieX search repair config key "%s" must be a list.', $path));
|
||||
}
|
||||
|
||||
$out = [];
|
||||
@@ -382,6 +404,57 @@ final class SearchRepairConfig
|
||||
$out[] = $item;
|
||||
}
|
||||
|
||||
return $out !== [] ? $out : $default;
|
||||
if ($out === []) {
|
||||
throw new \InvalidArgumentException(sprintf('RetrieX search repair config key "%s" must not be empty.', $path));
|
||||
}
|
||||
|
||||
return $out;
|
||||
}
|
||||
|
||||
/** @return int[] */
|
||||
private function requiredCharacterCodes(string $path): array
|
||||
{
|
||||
$value = $this->requiredValue($path);
|
||||
|
||||
if (!is_array($value)) {
|
||||
throw new \InvalidArgumentException(sprintf('RetrieX search repair config key "%s" must be a list of character codes.', $path));
|
||||
}
|
||||
|
||||
$codes = [];
|
||||
foreach ($value as $item) {
|
||||
if (is_int($item)) {
|
||||
$code = $item;
|
||||
} elseif (is_string($item) && preg_match('/^\d+$/', trim($item)) === 1) {
|
||||
$code = (int) trim($item);
|
||||
} else {
|
||||
throw new \InvalidArgumentException(sprintf('RetrieX search repair config key "%s" contains a non-integer character code.', $path));
|
||||
}
|
||||
|
||||
if ($code < 0 || $code > 255) {
|
||||
throw new \InvalidArgumentException(sprintf('RetrieX search repair config key "%s" contains an invalid character code.', $path));
|
||||
}
|
||||
|
||||
$codes[] = $code;
|
||||
}
|
||||
|
||||
if ($codes === []) {
|
||||
throw new \InvalidArgumentException(sprintf('RetrieX search repair config key "%s" must not be empty.', $path));
|
||||
}
|
||||
|
||||
return $codes;
|
||||
}
|
||||
|
||||
private function requiredValue(string $path): mixed
|
||||
{
|
||||
$current = $this->config;
|
||||
foreach (explode('.', $path) as $segment) {
|
||||
if (!is_array($current) || !array_key_exists($segment, $current)) {
|
||||
throw new \InvalidArgumentException(sprintf('Missing required RetrieX search repair config key "%s".', $path));
|
||||
}
|
||||
|
||||
$current = $current[$segment];
|
||||
}
|
||||
|
||||
return $current;
|
||||
}
|
||||
}
|
||||
|
||||
Reference in New Issue
Block a user