move intent an config value into config files

2026-04-23 21:49:54 +02:00
parent 87417febf4
commit fce44e971d
17 changed files with 1937 additions and 1133 deletions
--- a/src/Config/SearchRepairConfig.php
+++ b/src/Config/SearchRepairConfig.php
@@ -0,0 +1,204 @@
+<?php
+
+declare(strict_types=1);
+
+namespace App\Config;
+
+final class SearchRepairConfig
+{
+    public function isEnabled(): bool
+    {
+        return true;
+    }
+
+    public function getMaxRepairQueries(): int
+    {
+        return 3;
+    }
+
+    public function getMinPrimaryResultsWithoutRepair(): int
+    {
+        return 2;
+    }
+
+    public function getTopProductLogLimit(): int
+    {
+        return 3;
+    }
+
+    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';
+    }
+
+    public function getAccessoryCandidatePattern(): string
+    {
+        return '/\b((?:' . implode('|', $this->getAccessoryCandidateTerms()) . ')\s+\d{1,5}[A-Za-z0-9\-]*)\b/iu';
+    }
+
+    public function getAccessoryOrBundlePattern(): string
+    {
+        return '/\b(' . implode('|', $this->getAccessoryOrBundleTerms()) . ')\b/iu';
+    }
+
+    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';
+    }
+
+    public function getSpecificityBoostPattern(): string
+    {
+        return '/\b(?:' . implode('|', $this->getSpecificityBoostTerms()) . ')\b/iu';
+    }
+
+    /**
+     * @return string[]
+     */
+    public function getGenericCandidateTokens(): array
+    {
+        return [
+            'wasser',
+            'messgerät',
+            'messgeraet',
+            'produkt',
+            'geräte',
+            'geraete',
+            'gerät',
+            'geraet',
+            'resthärte',
+            'resthaerte',
+            'preis',
+            'infos',
+            'wissen',
+        ];
+    }
+
+    public function getSanitizeTrimCharacters(): string
+    {
+        return " \t\n\r\0\x0B\"'`.,;:-";
+    }
+
+    public function getContainsDigitPattern(): string
+    {
+        return '/\d/u';
+    }
+
+    public function getWhitespaceCollapsePattern(): string
+    {
+        return '/\s+/u';
+    }
+
+    public function getTokenizeCleanupPattern(): string
+    {
+        return '/[^\p{L}\p{N}\s\-]+/u';
+    }
+
+    public function getProductKeySeparator(): string
+    {
+        return '|';
+    }
+
+    public function getCandidateDigitScore(): int
+    {
+        return 4;
+    }
+
+    public function getCandidateWordCountCap(): int
+    {
+        return 4;
+    }
+
+    public function getSpecificityBoostScore(): int
+    {
+        return 3;
+    }
+
+    public function getPrimaryQueryOverlapThreshold(): float
+    {
+        return 0.9;
+    }
+
+    public function getPromptMatchWeight(): int
+    {
+        return 3;
+    }
+
+    public function getPrimaryQueryMatchWeight(): int
+    {
+        return 2;
+    }
+
+    public function getRepairSignalMatchWeight(): int
+    {
+        return 4;
+    }
+
+    public function getPrimaryResultOrderBonus(): int
+    {
+        return 1;
+    }
+
+    public function getTokenIntersectionScore(): int
+    {
+        return 2;
+    }
+
+    public function getNumericTokenMatchScore(): int
+    {
+        return 4;
+    }
+
+    /**
+     * @return string[]
+     */
+    public function getAccessoryCandidateTerms(): array
+    {
+        return [
+            'indikator',
+            'indicator',
+            'reagenz',
+            'reagent',
+            'kit',
+            'set',
+        ];
+    }
+
+    /**
+     * @return string[]
+     */
+    public function getAccessoryOrBundleTerms(): array
+    {
+        return [
+            'passend',
+            'passende',
+            'zubehor',
+            'zubehör',
+            'dazu',
+            'zusatz',
+            'erganzung',
+            'ergänzung',
+            'indikator',
+            'reagenz',
+            'kit',
+            'set',
+            'auch\s+das',
+            'mit\s+preis\s+und\s+allen\s+infos',
+        ];
+    }
+
+    /**
+     * @return string[]
+     */
+    public function getSpecificityBoostTerms(): array
+    {
+        return [
+            'indikator',
+            'indicator',
+            'testomat',
+            'tritromat',
+            'titromat',
+            'reagenz',
+            'reagent',
+        ];
+    }
+}