@ninebix/nmt-system 1.0.0

package/dist/services/query.js

@@ -0,0 +1,227 @@
+/**
+ * Query Service - Semantic search and retrieval
+ * @module services/query
+ */
+import { cosineSimilarity } from '../utils/similarity.js';
+/**
+ * Default query embedding provider
+ */
+class DefaultQueryEmbeddingProvider {
+    async embed(text) {
+        const embedding = new Float32Array(384);
+        let hash = 0;
+        for (let i = 0; i < text.length; i++) {
+            hash = ((hash << 5) - hash) + text.charCodeAt(i);
+            hash |= 0;
+        }
+        for (let i = 0; i < 384; i++) {
+            hash = ((hash << 5) - hash) + i;
+            hash |= 0;
+            embedding[i] = (hash % 1000) / 1000 - 0.5;
+        }
+        let norm = 0;
+        for (let i = 0; i < 384; i++) {
+            norm += embedding[i] * embedding[i];
+        }
+        norm = Math.sqrt(norm);
+        for (let i = 0; i < 384; i++) {
+            embedding[i] /= norm;
+        }
+        return embedding;
+    }
+}
+/**
+ * Query Service for semantic search
+ */
+export class QueryService {
+    graphManager;
+    merkleEngine;
+    chunkStore;
+    neuronStore;
+    embeddingProvider;
+    constructor(graphManager, merkleEngine, chunkStore, neuronStore, embeddingProvider) {
+        this.graphManager = graphManager;
+        this.merkleEngine = merkleEngine;
+        this.chunkStore = chunkStore;
+        this.neuronStore = neuronStore;
+        this.embeddingProvider = embeddingProvider ?? new DefaultQueryEmbeddingProvider();
+    }
+    /**
+     * Search by text query
+     */
+    async search(query, options = {}) {
+        const embedding = await this.embeddingProvider.embed(query);
+        return this.searchByEmbedding(embedding, options);
+    }
+    /**
+     * Search by embedding vector
+     */
+    async searchByEmbedding(embedding, options = {}) {
+        const { k = 10, ef, threshold = 0, includeContent = false, includeProof = false, tags, sourceType } = options;
+        // Get similar neurons from graph manager
+        const similar = await this.graphManager.findSimilar(embedding, k * 2, ef);
+        // Filter and transform results
+        const results = [];
+        for (const { neuron, score } of similar) {
+            // Apply filters
+            if (score < threshold)
+                continue;
+            if (tags && !tags.some(t => neuron.metadata.tags.includes(t)))
+                continue;
+            if (sourceType && neuron.metadata.sourceType !== sourceType)
+                continue;
+            const result = {
+                neuron,
+                score
+            };
+            // Include content if requested
+            if (includeContent) {
+                result.content = await this.getContent(neuron);
+            }
+            // Include proof if requested
+            if (includeProof) {
+                const proof = await this.generateProof(neuron, 0);
+                if (proof)
+                    result.proof = proof;
+            }
+            results.push(result);
+            if (results.length >= k)
+                break;
+        }
+        return results;
+    }
+    /**
+     * Search by neuron ID (find similar)
+     */
+    async searchSimilarTo(neuronId, options = {}) {
+        const neuron = await this.graphManager.getNeuron(neuronId);
+        if (!neuron)
+            return [];
+        return this.searchByEmbedding(neuron.embedding, {
+            ...options,
+            k: (options.k ?? 10) + 1 // Add 1 to account for self
+        }).then(results => results.filter(r => r.neuron.id !== neuronId));
+    }
+    /**
+     * Get neuron by ID
+     */
+    async getNeuron(id) {
+        return this.graphManager.getNeuron(id);
+    }
+    /**
+     * Get neuron by Merkle root
+     */
+    async getNeuronByMerkleRoot(merkleRoot) {
+        return this.graphManager.getNeuronByMerkleRoot(merkleRoot);
+    }
+    /**
+     * Get neuron content
+     */
+    async getContent(neuron) {
+        const chunks = [];
+        for (const hash of neuron.chunkHashes) {
+            const chunk = await this.chunkStore.get(hash);
+            if (chunk) {
+                chunks.push(chunk);
+            }
+        }
+        // Sort by index and merge
+        chunks.sort((a, b) => a.index - b.index);
+        const merged = Buffer.concat(chunks.map(c => c.data));
+        return merged.toString('utf-8');
+    }
+    /**
+     * Generate Merkle proof for a chunk
+     */
+    async generateProof(neuron, chunkIndex) {
+        if (chunkIndex < 0 || chunkIndex >= neuron.chunkHashes.length) {
+            return null;
+        }
+        // Rebuild Merkle tree
+        const tree = this.merkleEngine.buildTree(neuron.chunkHashes);
+        // Generate proof
+        return this.merkleEngine.generateProof(tree, chunkIndex);
+    }
+    /**
+     * Compute similarity between two neurons
+     */
+    async computeSimilarity(neuronId1, neuronId2) {
+        const neuron1 = await this.graphManager.getNeuron(neuronId1);
+        const neuron2 = await this.graphManager.getNeuron(neuronId2);
+        if (!neuron1 || !neuron2)
+            return null;
+        return cosineSimilarity(neuron1.embedding, neuron2.embedding);
+    }
+    /**
+     * Get neurons by tags
+     */
+    async getNeuronsByTags(tags) {
+        const allIds = await this.neuronStore.getAllNeuronIds();
+        const results = [];
+        for (const id of allIds) {
+            const neuron = await this.neuronStore.getNeuron(id);
+            if (neuron && tags.some(t => neuron.metadata.tags.includes(t))) {
+                results.push(neuron);
+            }
+        }
+        return results;
+    }
+    /**
+     * Get neurons by source type
+     */
+    async getNeuronsBySourceType(sourceType) {
+        const allIds = await this.neuronStore.getAllNeuronIds();
+        const results = [];
+        for (const id of allIds) {
+            const neuron = await this.neuronStore.getNeuron(id);
+            if (neuron && neuron.metadata.sourceType === sourceType) {
+                results.push(neuron);
+            }
+        }
+        return results;
+    }
+    /**
+     * Get recently accessed neurons
+     */
+    async getRecentlyAccessed(limit = 10) {
+        const allIds = await this.neuronStore.getAllNeuronIds();
+        const neurons = [];
+        for (const id of allIds) {
+            const neuron = await this.neuronStore.getNeuron(id);
+            if (neuron)
+                neurons.push(neuron);
+        }
+        // Sort by last accessed
+        neurons.sort((a, b) => new Date(b.metadata.lastAccessed).getTime() -
+            new Date(a.metadata.lastAccessed).getTime());
+        return neurons.slice(0, limit);
+    }
+    /**
+     * Get most accessed neurons
+     */
+    async getMostAccessed(limit = 10) {
+        const allIds = await this.neuronStore.getAllNeuronIds();
+        const neurons = [];
+        for (const id of allIds) {
+            const neuron = await this.neuronStore.getNeuron(id);
+            if (neuron)
+                neurons.push(neuron);
+        }
+        // Sort by access count
+        neurons.sort((a, b) => b.metadata.accessCount - a.metadata.accessCount);
+        return neurons.slice(0, limit);
+    }
+    /**
+     * Set embedding provider
+     */
+    setEmbeddingProvider(provider) {
+        this.embeddingProvider = provider;
+    }
+}
+/**
+ * Create a QueryService instance
+ */
+export function createQueryService(graphManager, merkleEngine, chunkStore, neuronStore, embeddingProvider) {
+    return new QueryService(graphManager, merkleEngine, chunkStore, neuronStore, embeddingProvider);
+}
+//# sourceMappingURL=query.js.map

package/dist/services/query.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"query.js","sourceRoot":"","sources":["../../src/services/query.ts"],"names":[],"mappings":"AAAA;;;GAGG;AAcH,OAAO,EAAE,gBAAgB,EAAE,MAAM,wBAAwB,CAAC;AAgC1D;;GAEG;AACH,MAAM,6BAA6B;IACjC,KAAK,CAAC,KAAK,CAAC,IAAY;QACtB,MAAM,SAAS,GAAG,IAAI,YAAY,CAAC,GAAG,CAAC,CAAC;QACxC,IAAI,IAAI,GAAG,CAAC,CAAC;QACb,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,IAAI,CAAC,MAAM,EAAE,CAAC,EAAE,EAAE,CAAC;YACrC,IAAI,GAAG,CAAC,CAAC,IAAI,IAAI,CAAC,CAAC,GAAG,IAAI,CAAC,GAAG,IAAI,CAAC,UAAU,CAAC,CAAC,CAAC,CAAC;YACjD,IAAI,IAAI,CAAC,CAAC;QACZ,CAAC;QAED,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,GAAG,EAAE,CAAC,EAAE,EAAE,CAAC;YAC7B,IAAI,GAAG,CAAC,CAAC,IAAI,IAAI,CAAC,CAAC,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC;YAChC,IAAI,IAAI,CAAC,CAAC;YACV,SAAS,CAAC,CAAC,CAAC,GAAG,CAAC,IAAI,GAAG,IAAI,CAAC,GAAG,IAAI,GAAG,GAAG,CAAC;QAC5C,CAAC;QAED,IAAI,IAAI,GAAG,CAAC,CAAC;QACb,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,GAAG,EAAE,CAAC,EAAE,EAAE,CAAC;YAC7B,IAAI,IAAI,SAAS,CAAC,CAAC,CAAC,GAAG,SAAS,CAAC,CAAC,CAAC,CAAC;QACtC,CAAC;QACD,IAAI,GAAG,IAAI,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;QACvB,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,GAAG,EAAE,CAAC,EAAE,EAAE,CAAC;YAC7B,SAAS,CAAC,CAAC,CAAC,IAAI,IAAI,CAAC;QACvB,CAAC;QAED,OAAO,SAAS,CAAC;IACnB,CAAC;CACF;AAED;;GAEG;AACH,MAAM,OAAO,YAAY;IACf,YAAY,CAAqB;IACjC,YAAY,CAAe;IAC3B,UAAU,CAAc;IACxB,WAAW,CAAe;IAC1B,iBAAiB,CAAyB;IAElD,YACE,YAAgC,EAChC,YAA0B,EAC1B,UAAuB,EACvB,WAAyB,EACzB,iBAA0C;QAE1C,IAAI,CAAC,YAAY,GAAG,YAAY,CAAC;QACjC,IAAI,CAAC,YAAY,GAAG,YAAY,CAAC;QACjC,IAAI,CAAC,UAAU,GAAG,UAAU,CAAC;QAC7B,IAAI,CAAC,WAAW,GAAG,WAAW,CAAC;QAC/B,IAAI,CAAC,iBAAiB,GAAG,iBAAiB,IAAI,IAAI,6BAA6B,EAAE,CAAC;IACpF,CAAC;IAED;;OAEG;IACH,KAAK,CAAC,MAAM,CAAC,KAAa,EAAE,UAAyB,EAAE;QACrD,MAAM,SAAS,GAAG,MAAM,IAAI,CAAC,iBAAiB,CAAC,KAAK,CAAC,KAAK,CAAC,CAAC;QAC5D,OAAO,IAAI,CAAC,iBAAiB,CAAC,SAAS,EAAE,OAAO,CAAC,CAAC;IACpD,CAAC;IAED;;OAEG;IACH,KAAK,CAAC,iBAAiB,CACrB,SAAuB,EACvB,UAAyB,EAAE;QAE3B,MAAM,EACJ,CAAC,GAAG,EAAE,EACN,EAAE,EACF,SAAS,GAAG,CAAC,EACb,cAAc,GAAG,KAAK,EACtB,YAAY,GAAG,KAAK,EACpB,IAAI,EACJ,UAAU,EACX,GAAG,OAAO,CAAC;QAEZ,yCAAyC;QACzC,MAAM,OAAO,GAAG,MAAM,IAAI,CAAC,YAAY,CAAC,WAAW,CAAC,SAAS,EAAE,CAAC,GAAG,CAAC,EAAE,EAAE,CAAC,CAAC;QAE1E,+BAA+B;QAC/B,MAAM,OAAO,GAAkB,EAAE,CAAC;QAElC,KAAK,MAAM,EAAE,MAAM,EAAE,KAAK,EAAE,IAAI,OAAO,EAAE,CAAC;YACxC,gBAAgB;YAChB,IAAI,KAAK,GAAG,SAAS;gBAAE,SAAS;YAChC,IAAI,IAAI,IAAI,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,MAAM,CAAC,QAAQ,CAAC,IAAI,CAAC,QAAQ,CAAC,CAAC,CAAC,CAAC;gBAAE,SAAS;YACxE,IAAI,UAAU,IAAI,MAAM,CAAC,QAAQ,CAAC,UAAU,KAAK,UAAU;gBAAE,SAAS;YAEtE,MAAM,MAAM,GAAgB;gBAC1B,MAAM;gBACN,KAAK;aACN,CAAC;YAEF,+BAA+B;YAC/B,IAAI,cAAc,EAAE,CAAC;gBACnB,MAAM,CAAC,OAAO,GAAG,MAAM,IAAI,CAAC,UAAU,CAAC,MAAM,CAAC,CAAC;YACjD,CAAC;YAED,6BAA6B;YAC7B,IAAI,YAAY,EAAE,CAAC;gBACjB,MAAM,KAAK,GAAG,MAAM,IAAI,CAAC,aAAa,CAAC,MAAM,EAAE,CAAC,CAAC,CAAC;gBAClD,IAAI,KAAK;oBAAE,MAAM,CAAC,KAAK,GAAG,KAAK,CAAC;YAClC,CAAC;YAED,OAAO,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC;YAErB,IAAI,OAAO,CAAC,MAAM,IAAI,CAAC;gBAAE,MAAM;QACjC,CAAC;QAED,OAAO,OAAO,CAAC;IACjB,CAAC;IAED;;OAEG;IACH,KAAK,CAAC,eAAe,CACnB,QAAc,EACd,UAAyB,EAAE;QAE3B,MAAM,MAAM,GAAG,MAAM,IAAI,CAAC,YAAY,CAAC,SAAS,CAAC,QAAQ,CAAC,CAAC;QAC3D,IAAI,CAAC,MAAM;YAAE,OAAO,EAAE,CAAC;QAEvB,OAAO,IAAI,CAAC,iBAAiB,CAAC,MAAM,CAAC,SAAS,EAAE;YAC9C,GAAG,OAAO;YACV,CAAC,EAAE,CAAC,OAAO,CAAC,CAAC,IAAI,EAAE,CAAC,GAAG,CAAC,CAAC,4BAA4B;SACtD,CAAC,CAAC,IAAI,CAAC,OAAO,CAAC,EAAE,CAAC,OAAO,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,MAAM,CAAC,EAAE,KAAK,QAAQ,CAAC,CAAC,CAAC;IACpE,CAAC;IAED;;OAEG;IACH,KAAK,CAAC,SAAS,CAAC,EAAQ;QACtB,OAAO,IAAI,CAAC,YAAY,CAAC,SAAS,CAAC,EAAE,CAAC,CAAC;IACzC,CAAC;IAED;;OAEG;IACH,KAAK,CAAC,qBAAqB,CAAC,UAAkB;QAC5C,OAAO,IAAI,CAAC,YAAY,CAAC,qBAAqB,CAAC,UAAU,CAAC,CAAC;IAC7D,CAAC;IAED;;OAEG;IACH,KAAK,CAAC,UAAU,CAAC,MAAkB;QACjC,MAAM,MAAM,GAAY,EAAE,CAAC;QAE3B,KAAK,MAAM,IAAI,IAAI,MAAM,CAAC,WAAW,EAAE,CAAC;YACtC,MAAM,KAAK,GAAG,MAAM,IAAI,CAAC,UAAU,CAAC,GAAG,CAAC,IAAI,CAAC,CAAC;YAC9C,IAAI,KAAK,EAAE,CAAC;gBACV,MAAM,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC;YACrB,CAAC;QACH,CAAC;QAED,0BAA0B;QAC1B,MAAM,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,KAAK,GAAG,CAAC,CAAC,KAAK,CAAC,CAAC;QACzC,MAAM,MAAM,GAAG,MAAM,CAAC,MAAM,CAAC,MAAM,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC;QAEtD,OAAO,MAAM,CAAC,QAAQ,CAAC,OAAO,CAAC,CAAC;IAClC,CAAC;IAED;;OAEG;IACH,KAAK,CAAC,aAAa,CACjB,MAAkB,EAClB,UAAkB;QAElB,IAAI,UAAU,GAAG,CAAC,IAAI,UAAU,IAAI,MAAM,CAAC,WAAW,CAAC,MAAM,EAAE,CAAC;YAC9D,OAAO,IAAI,CAAC;QACd,CAAC;QAED,sBAAsB;QACtB,MAAM,IAAI,GAAG,IAAI,CAAC,YAAY,CAAC,SAAS,CAAC,MAAM,CAAC,WAAW,CAAC,CAAC;QAE7D,iBAAiB;QACjB,OAAO,IAAI,CAAC,YAAY,CAAC,aAAa,CAAC,IAAI,EAAE,UAAU,CAAC,CAAC;IAC3D,CAAC;IAED;;OAEG;IACH,KAAK,CAAC,iBAAiB,CAAC,SAAe,EAAE,SAAe;QACtD,MAAM,OAAO,GAAG,MAAM,IAAI,CAAC,YAAY,CAAC,SAAS,CAAC,SAAS,CAAC,CAAC;QAC7D,MAAM,OAAO,GAAG,MAAM,IAAI,CAAC,YAAY,CAAC,SAAS,CAAC,SAAS,CAAC,CAAC;QAE7D,IAAI,CAAC,OAAO,IAAI,CAAC,OAAO;YAAE,OAAO,IAAI,CAAC;QAEtC,OAAO,gBAAgB,CAAC,OAAO,CAAC,SAAS,EAAE,OAAO,CAAC,SAAS,CAAC,CAAC;IAChE,CAAC;IAED;;OAEG;IACH,KAAK,CAAC,gBAAgB,CAAC,IAAc;QACnC,MAAM,MAAM,GAAG,MAAM,IAAI,CAAC,WAAW,CAAC,eAAe,EAAE,CAAC;QACxD,MAAM,OAAO,GAAiB,EAAE,CAAC;QAEjC,KAAK,MAAM,EAAE,IAAI,MAAM,EAAE,CAAC;YACxB,MAAM,MAAM,GAAG,MAAM,IAAI,CAAC,WAAW,CAAC,SAAS,CAAC,EAAE,CAAC,CAAC;YACpD,IAAI,MAAM,IAAI,IAAI,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,MAAM,CAAC,QAAQ,CAAC,IAAI,CAAC,QAAQ,CAAC,CAAC,CAAC,CAAC,EAAE,CAAC;gBAC/D,OAAO,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC;YACvB,CAAC;QACH,CAAC;QAED,OAAO,OAAO,CAAC;IACjB,CAAC;IAED;;OAEG;IACH,KAAK,CAAC,sBAAsB,CAAC,UAAkB;QAC7C,MAAM,MAAM,GAAG,MAAM,IAAI,CAAC,WAAW,CAAC,eAAe,EAAE,CAAC;QACxD,MAAM,OAAO,GAAiB,EAAE,CAAC;QAEjC,KAAK,MAAM,EAAE,IAAI,MAAM,EAAE,CAAC;YACxB,MAAM,MAAM,GAAG,MAAM,IAAI,CAAC,WAAW,CAAC,SAAS,CAAC,EAAE,CAAC,CAAC;YACpD,IAAI,MAAM,IAAI,MAAM,CAAC,QAAQ,CAAC,UAAU,KAAK,UAAU,EAAE,CAAC;gBACxD,OAAO,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC;YACvB,CAAC;QACH,CAAC;QAED,OAAO,OAAO,CAAC;IACjB,CAAC;IAED;;OAEG;IACH,KAAK,CAAC,mBAAmB,CAAC,QAAgB,EAAE;QAC1C,MAAM,MAAM,GAAG,MAAM,IAAI,CAAC,WAAW,CAAC,eAAe,EAAE,CAAC;QACxD,MAAM,OAAO,GAAiB,EAAE,CAAC;QAEjC,KAAK,MAAM,EAAE,IAAI,MAAM,EAAE,CAAC;YACxB,MAAM,MAAM,GAAG,MAAM,IAAI,CAAC,WAAW,CAAC,SAAS,CAAC,EAAE,CAAC,CAAC;YACpD,IAAI,MAAM;gBAAE,OAAO,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC;QACnC,CAAC;QAED,wBAAwB;QACxB,OAAO,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CACpB,IAAI,IAAI,CAAC,CAAC,CAAC,QAAQ,CAAC,YAAY,CAAC,CAAC,OAAO,EAAE;YAC3C,IAAI,IAAI,CAAC,CAAC,CAAC,QAAQ,CAAC,YAAY,CAAC,CAAC,OAAO,EAAE,CAC5C,CAAC;QAEF,OAAO,OAAO,CAAC,KAAK,CAAC,CAAC,EAAE,KAAK,CAAC,CAAC;IACjC,CAAC;IAED;;OAEG;IACH,KAAK,CAAC,eAAe,CAAC,QAAgB,EAAE;QACtC,MAAM,MAAM,GAAG,MAAM,IAAI,CAAC,WAAW,CAAC,eAAe,EAAE,CAAC;QACxD,MAAM,OAAO,GAAiB,EAAE,CAAC;QAEjC,KAAK,MAAM,EAAE,IAAI,MAAM,EAAE,CAAC;YACxB,MAAM,MAAM,GAAG,MAAM,IAAI,CAAC,WAAW,CAAC,SAAS,CAAC,EAAE,CAAC,CAAC;YACpD,IAAI,MAAM;gBAAE,OAAO,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC;QACnC,CAAC;QAED,uBAAuB;QACvB,OAAO,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,QAAQ,CAAC,WAAW,GAAG,CAAC,CAAC,QAAQ,CAAC,WAAW,CAAC,CAAC;QAExE,OAAO,OAAO,CAAC,KAAK,CAAC,CAAC,EAAE,KAAK,CAAC,CAAC;IACjC,CAAC;IAED;;OAEG;IACH,oBAAoB,CAAC,QAAgC;QACnD,IAAI,CAAC,iBAAiB,GAAG,QAAQ,CAAC;IACpC,CAAC;CACF;AAED;;GAEG;AACH,MAAM,UAAU,kBAAkB,CAChC,YAAgC,EAChC,YAA0B,EAC1B,UAAuB,EACvB,WAAyB,EACzB,iBAA0C;IAE1C,OAAO,IAAI,YAAY,CACrB,YAAY,EACZ,YAAY,EACZ,UAAU,EACV,WAAW,EACX,iBAAiB,CAClB,CAAC;AACJ,CAAC"}

package/dist/services/text-embedding.d.ts

@@ -0,0 +1,183 @@
+/**
+ * Text Embedding Service
+ *
+ * Provides multiple embedding strategies:
+ * 1. Xenova transformers (local neural network) - high quality
+ * 2. External API (OpenAI, Cohere, etc.) - most accurate
+ * 3. TF-IDF + SVD - local, no external dependencies
+ * 4. Semantic hash - fast fallback
+ *
+ * @module services/text-embedding
+ */
+import type { Embedding384 } from '../types/index.js';
+/**
+ * Embedding provider configuration
+ */
+export interface EmbeddingProviderConfig {
+    type: 'xenova' | 'openai' | 'cohere' | 'local' | 'hybrid';
+    apiKey?: string;
+    modelId?: string;
+    baseUrl?: string;
+    dimension?: number;
+}
+/**
+ * Embedding request
+ */
+export interface EmbeddingRequest {
+    text: string;
+    contextEmbedding?: Embedding384;
+    metadata?: Record<string, unknown>;
+}
+/**
+ * Embedding response
+ */
+export interface EmbeddingResponse {
+    embedding: Embedding384;
+    provider: string;
+    model?: string;
+    tokenCount?: number;
+    cached?: boolean;
+}
+/**
+ * Text Embedding Service
+ *
+ * Generates semantic embeddings for text using multiple strategies.
+ */
+export declare class TextEmbeddingService {
+    private config;
+    private vocabulary;
+    private idfValues;
+    private documentCount;
+    private embeddingDimension;
+    private cache;
+    private maxCacheSize;
+    private wordVectors;
+    constructor(config?: Partial<EmbeddingProviderConfig>);
+    /**
+     * Generate embedding for text
+     */
+    embed(request: EmbeddingRequest): Promise<EmbeddingResponse>;
+    /**
+     * Xenova transformers embedding (local neural network)
+     */
+    private embedWithXenova;
+    /**
+     * Batch embed multiple texts
+     */
+    embedBatch(texts: string[]): Promise<EmbeddingResponse[]>;
+    /**
+     * Update vocabulary with new documents (for TF-IDF)
+     */
+    updateVocabulary(documents: string[]): void;
+    /**
+     * Local embedding using TF-IDF + semantic projection
+     */
+    private embedLocal;
+    /**
+     * Hybrid embedding: try best available, fallback to local
+     *
+     * Priority:
+     * 1. OpenAI API (if apiKey configured)
+     * 2. Xenova transformers (local neural network)
+     * 3. TF-IDF + semantic analysis (pure local)
+     */
+    private embedHybrid;
+    /**
+     * OpenAI API embedding
+     */
+    private embedWithOpenAI;
+    /**
+     * Cohere API embedding
+     */
+    private embedWithCohere;
+    /**
+     * Batch API embedding
+     */
+    private batchEmbedAPI;
+    /**
+     * Calculate TF-IDF for tokens
+     */
+    private calculateTFIDF;
+    /**
+     * Weighted average of word vectors
+     */
+    private weightedWordVectorAverage;
+    /**
+     * Character n-gram embedding
+     */
+    private characterNgramEmbedding;
+    /**
+     * Positional encoding (transformer-style)
+     */
+    private positionalEncoding;
+    /**
+     * Semantic category encoding
+     */
+    private semanticCategoryEncoding;
+    /**
+     * Initialize common word vectors
+     */
+    private initializeWordVectors;
+    /**
+     * Generate a semantic vector from direction and spread
+     */
+    private generateSemanticVector;
+    /**
+     * Get or generate word vector
+     */
+    private getWordVector;
+    /**
+     * Tokenize text
+     */
+    private tokenize;
+    /**
+     * Extract character n-grams
+     */
+    private extractNgrams;
+    /**
+     * Hash string to number
+     */
+    private hashString;
+    /**
+     * L2 normalize embedding
+     */
+    private normalize;
+    /**
+     * Resize embedding to target dimension
+     */
+    private resizeEmbedding;
+    /**
+     * Cache key for text
+     */
+    private getCacheKey;
+    /**
+     * Cache embedding
+     */
+    private cacheEmbedding;
+    /**
+     * Clear cache
+     */
+    clearCache(): void;
+    /**
+     * Get statistics
+     */
+    getStats(): {
+        vocabularySize: number;
+        cacheSize: number;
+        wordVectorCount: number;
+        documentCount: number;
+    };
+    /**
+     * Serialize state for persistence
+     */
+    serialize(): object;
+    /**
+     * Load state from serialized data
+     */
+    load(data: any): void;
+}
+/**
+ * Create a TextEmbeddingService instance
+ */
+export declare function createTextEmbeddingService(config?: Partial<EmbeddingProviderConfig>): TextEmbeddingService;
+//# sourceMappingURL=text-embedding.d.ts.map

package/dist/services/text-embedding.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"text-embedding.d.ts","sourceRoot":"","sources":["../../src/services/text-embedding.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;GAUG;AAEH,OAAO,KAAK,EAAE,YAAY,EAAE,MAAM,mBAAmB,CAAC;AAGtD;;GAEG;AACH,MAAM,WAAW,uBAAuB;IACtC,IAAI,EAAE,QAAQ,GAAG,QAAQ,GAAG,QAAQ,GAAG,OAAO,GAAG,QAAQ,CAAC;IAC1D,MAAM,CAAC,EAAE,MAAM,CAAC;IAChB,OAAO,CAAC,EAAE,MAAM,CAAC;IACjB,OAAO,CAAC,EAAE,MAAM,CAAC;IACjB,SAAS,CAAC,EAAE,MAAM,CAAC;CACpB;AAED;;GAEG;AACH,MAAM,WAAW,gBAAgB;IAC/B,IAAI,EAAE,MAAM,CAAC;IACb,gBAAgB,CAAC,EAAE,YAAY,CAAC;IAChC,QAAQ,CAAC,EAAE,MAAM,CAAC,MAAM,EAAE,OAAO,CAAC,CAAC;CACpC;AAED;;GAEG;AACH,MAAM,WAAW,iBAAiB;IAChC,SAAS,EAAE,YAAY,CAAC;IACxB,QAAQ,EAAE,MAAM,CAAC;IACjB,KAAK,CAAC,EAAE,MAAM,CAAC;IACf,UAAU,CAAC,EAAE,MAAM,CAAC;IACpB,MAAM,CAAC,EAAE,OAAO,CAAC;CAClB;AAYD;;;;GAIG;AACH,qBAAa,oBAAoB;IAC/B,OAAO,CAAC,MAAM,CAA0B;IACxC,OAAO,CAAC,UAAU,CAAkC;IACpD,OAAO,CAAC,SAAS,CAAkC;IACnD,OAAO,CAAC,aAAa,CAAa;IAClC,OAAO,CAAC,kBAAkB,CAAS;IAGnC,OAAO,CAAC,KAAK,CAA6C;IAC1D,OAAO,CAAC,YAAY,CAAiB;IAGrC,OAAO,CAAC,WAAW,CAAwC;gBAE/C,MAAM,GAAE,OAAO,CAAC,uBAAuB,CAAM;IAczD;;OAEG;IACG,KAAK,CAAC,OAAO,EAAE,gBAAgB,GAAG,OAAO,CAAC,iBAAiB,CAAC;IAoClE;;OAEG;YACW,eAAe;IAmB7B;;OAEG;IACG,UAAU,CAAC,KAAK,EAAE,MAAM,EAAE,GAAG,OAAO,CAAC,iBAAiB,EAAE,CAAC;IAU/D;;OAEG;IACH,gBAAgB,CAAC,SAAS,EAAE,MAAM,EAAE,GAAG,IAAI;IAyB3C;;OAEG;YACW,UAAU;IAoDxB;;;;;;;OAOG;YACW,WAAW;IAqBzB;;OAEG;YACW,eAAe;IAmC7B;;OAEG;YACW,eAAe;IAwC7B;;OAEG;YACW,aAAa;IAQ3B;;OAEG;IACH,OAAO,CAAC,cAAc;IAmBtB;;OAEG;IACH,OAAO,CAAC,yBAAyB;IA0BjC;;OAEG;IACH,OAAO,CAAC,uBAAuB;IA2B/B;;OAEG;IACH,OAAO,CAAC,kBAAkB;IAyB1B;;OAEG;IACH,OAAO,CAAC,wBAAwB;IA0ChC;;OAEG;IACH,OAAO,CAAC,qBAAqB;IA4C7B;;OAEG;IACH,OAAO,CAAC,sBAAsB;IAc9B;;OAEG;IACH,OAAO,CAAC,aAAa;IAqBrB;;OAEG;IACH,OAAO,CAAC,QAAQ;IAShB;;OAEG;IACH,OAAO,CAAC,aAAa;IAarB;;OAEG;IACH,OAAO,CAAC,UAAU;IAUlB;;OAEG;IACH,OAAO,CAAC,SAAS;IAcjB;;OAEG;IACH,OAAO,CAAC,eAAe;IAiCvB;;OAEG;IACH,OAAO,CAAC,WAAW;IAInB;;OAEG;IACH,OAAO,CAAC,cAAc;IAWtB;;OAEG;IACH,UAAU,IAAI,IAAI;IAIlB;;OAEG;IACH,QAAQ,IAAI;QACV,cAAc,EAAE,MAAM,CAAC;QACvB,SAAS,EAAE,MAAM,CAAC;QAClB,eAAe,EAAE,MAAM,CAAC;QACxB,aAAa,EAAE,MAAM,CAAC;KACvB;IASD;;OAEG;IACH,SAAS,IAAI,MAAM;IASnB;;OAEG;IACH,IAAI,CAAC,IAAI,EAAE,GAAG,GAAG,IAAI;CActB;AAED;;GAEG;AACH,wBAAgB,0BAA0B,CACxC,MAAM,CAAC,EAAE,OAAO,CAAC,uBAAuB,CAAC,GACxC,oBAAoB,CAEtB"}