@equationalapplications/core-llm-wiki 5.0.0 → 5.1.1

package/dist/testing.js

@@ -969,6 +969,10 @@ ${JSON.stringify(facts, null, 2)}`
   }
 };
 
+// src/utils/chunkingDefaults.ts
+var DEFAULT_MAX_CHUNK_LENGTH = 12e3;
+var DEFAULT_CHUNK_OVERLAP = 400;
+
 // src/services/IngestionService.ts
 var IngestionService = class {
   constructor(db, prefix, options, entryRepo, searchService, jobManager, embeddingService, promptService, ontologyService) {
@@ -987,10 +991,10 @@ var IngestionService = class {
     if (!sourceRef) throw new Error("Invalid sourceRef");
     const sourceHash = normalizeSourceHash(params.sourceHash);
     if (!sourceHash) throw new Error("Invalid sourceHash (must be 64-char hex string)");
-    const maxChunkLength = params.maxChunkLength ?? this.options.config?.maxChunkLength ?? 12e3;
-    const rawOverlap = params.chunkOverlap ?? this.options.config?.chunkOverlap ?? 400;
+    const maxChunkLength = params.maxChunkLength ?? this.options.config?.maxChunkLength ?? DEFAULT_MAX_CHUNK_LENGTH;
+    const rawOverlap = params.chunkOverlap ?? this.options.config?.chunkOverlap ?? DEFAULT_CHUNK_OVERLAP;
     const chunkOverlap = Math.min(
-      Number.isFinite(rawOverlap) && rawOverlap >= 0 ? Math.floor(rawOverlap) : 400,
+      Number.isFinite(rawOverlap) && rawOverlap >= 0 ? Math.floor(rawOverlap) : DEFAULT_CHUNK_OVERLAP,
       maxChunkLength - 1
     );
     const rawConcurrency = params.chunkConcurrency ?? this.options.config?.chunkConcurrency ?? 1;
@@ -1725,10 +1729,10 @@ var MaintenanceService = class {
     const validNewFacts = newFacts.map(validateFact).filter((f) => f !== null);
     const insertedFacts = [];
     const uniqueDeletedFactIds = Array.from(new Set(safeDeleted));
-    const healFactsForDedupe = (await this.entryRepo.findInferredTitlesByEntityId(entityId)).filter((f) => !safeDeletedSet.has(f.id));
     await this.db.withTransactionAsync(async (tx) => {
       await this.entryRepo.downgradeByIds(safeDowngraded, entityId, tx);
       await this.entryRepo.softDeleteByIds(safeDeleted, entityId, tx);
+      const healFactsForDedupe = await this.entryRepo.findInferredTitlesByEntityId(entityId, tx);
       for (const fact of validNewFacts) {
         const newTokens = titleTokens(fact.title);
         let skip = false;