vemora 0.1.0-alpha.15

package/dist/search/mmr.js

@@ -0,0 +1,95 @@
+"use strict";
+Object.defineProperty(exports, "__esModule", { value: true });
+exports.applyMMR = applyMMR;
+const vector_1 = require("./vector");
+/**
+ * Maximal Marginal Relevance (MMR) reranking.
+ *
+ * Selects up to `topK` results from `candidates` by iteratively picking
+ * the candidate that maximises:
+ *
+ *   MMR(d) = lambda * relevance(d) - (1 - lambda) * max_sim(d, selected)
+ *
+ * where:
+ *   - relevance(d) is the original retrieval score (normalised to [0, 1])
+ *   - max_sim(d, selected) is the maximum cosine similarity between d and
+ *     any already-selected result (requires embeddings in the cache)
+ *
+ * lambda=1.0 → pure relevance ordering (identical to original ranking)
+ * lambda=0.5 → balanced relevance / diversity (default)
+ * lambda=0.0 → maximum diversity, ignores relevance
+ *
+ * Falls back to returning `candidates.slice(0, topK)` when embeddings
+ * are unavailable (keyword-only mode).
+ */
+function applyMMR(candidates, cache, topK, lambda = 0.5) {
+    if (candidates.length <= 1)
+        return candidates.slice(0, topK);
+    // Without vector data MMR cannot compute inter-result similarity.
+    if (!cache?.vectors || !cache.chunkIds || !cache.dimensions) {
+        return candidates.slice(0, topK);
+    }
+    const { vectors, chunkIds, dimensions } = cache;
+    // Build a fast chunk-id → flat-buffer-index map.
+    const idToIdx = new Map();
+    chunkIds.forEach((id, i) => idToIdx.set(id, i));
+    // Normalise relevance scores to [0, 1] so lambda is meaningful regardless
+    // of whether scores come from cosine similarity (already ~[0,1]) or BM25.
+    // Use reduce instead of Math.max(...spread) to avoid call-stack overflow on large arrays.
+    let maxScore = -Infinity;
+    let minScore = Infinity;
+    for (const r of candidates) {
+        if (r.score > maxScore)
+            maxScore = r.score;
+        if (r.score < minScore)
+            minScore = r.score;
+    }
+    const range = maxScore - minScore || 1;
+    const relNorm = candidates.map((r) => (r.score - minScore) / range);
+    // Pre-extract each candidate's vector as number[] so the inner loop avoids
+    // repeated Array.from() allocations (O(n² * d) → O(n * d) allocations).
+    const candidateVecs = candidates.map((r) => {
+        const idx = idToIdx.get(r.chunk.id);
+        if (idx === undefined)
+            return null;
+        const offset = idx * dimensions;
+        return Array.from(vectors.subarray(offset, offset + dimensions));
+    });
+    const selected = [];
+    // Flat-buffer indices of already-selected results (for similarity queries).
+    const selectedVecIdx = [];
+    // Track remaining candidates as index positions into `candidates`.
+    const remaining = candidates.map((_, i) => i);
+    while (selected.length < topK && remaining.length > 0) {
+        let bestPos = -1; // position in `remaining`
+        let bestMMR = -Infinity;
+        for (let pos = 0; pos < remaining.length; pos++) {
+            const candIdx = remaining[pos];
+            const relevance = relNorm[candIdx];
+            let maxSim = 0;
+            if (selectedVecIdx.length > 0) {
+                const vecA = candidateVecs[candIdx];
+                if (vecA !== null) {
+                    for (const selIdx of selectedVecIdx) {
+                        const sim = (0, vector_1.cosineSimilarityBinary)(vecA, vectors, selIdx * dimensions, dimensions);
+                        if (sim > maxSim)
+                            maxSim = sim;
+                    }
+                }
+            }
+            const mmrScore = lambda * relevance - (1 - lambda) * maxSim;
+            if (mmrScore > bestMMR) {
+                bestMMR = mmrScore;
+                bestPos = pos;
+            }
+        }
+        const chosenIdx = remaining[bestPos];
+        selected.push(candidates[chosenIdx]);
+        const vecIdx = idToIdx.get(candidates[chosenIdx].chunk.id);
+        if (vecIdx !== undefined)
+            selectedVecIdx.push(vecIdx);
+        remaining.splice(bestPos, 1);
+    }
+    return selected;
+}
+//# sourceMappingURL=mmr.js.map

package/dist/search/mmr.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"mmr.js","sourceRoot":"","sources":["../../src/search/mmr.ts"],"names":[],"mappings":";;AAuBA,4BAwFC;AA9GD,qCAAkD;AAElD;;;;;;;;;;;;;;;;;;;GAmBG;AACH,SAAgB,QAAQ,CACtB,UAA0B,EAC1B,KAA4B,EAC5B,IAAY,EACZ,MAAM,GAAG,GAAG;IAEZ,IAAI,UAAU,CAAC,MAAM,IAAI,CAAC;QAAE,OAAO,UAAU,CAAC,KAAK,CAAC,CAAC,EAAE,IAAI,CAAC,CAAC;IAE7D,kEAAkE;IAClE,IAAI,CAAC,KAAK,EAAE,OAAO,IAAI,CAAC,KAAK,CAAC,QAAQ,IAAI,CAAC,KAAK,CAAC,UAAU,EAAE,CAAC;QAC5D,OAAO,UAAU,CAAC,KAAK,CAAC,CAAC,EAAE,IAAI,CAAC,CAAC;IACnC,CAAC;IAED,MAAM,EAAE,OAAO,EAAE,QAAQ,EAAE,UAAU,EAAE,GAAG,KAAK,CAAC;IAEhD,iDAAiD;IACjD,MAAM,OAAO,GAAG,IAAI,GAAG,EAAkB,CAAC;IAC1C,QAAQ,CAAC,OAAO,CAAC,CAAC,EAAE,EAAE,CAAC,EAAE,EAAE,CAAC,OAAO,CAAC,GAAG,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,CAAC;IAEhD,0EAA0E;IAC1E,0EAA0E;IAC1E,0FAA0F;IAC1F,IAAI,QAAQ,GAAG,CAAC,QAAQ,CAAC;IACzB,IAAI,QAAQ,GAAG,QAAQ,CAAC;IACxB,KAAK,MAAM,CAAC,IAAI,UAAU,EAAE,CAAC;QAC3B,IAAI,CAAC,CAAC,KAAK,GAAG,QAAQ;YAAE,QAAQ,GAAG,CAAC,CAAC,KAAK,CAAC;QAC3C,IAAI,CAAC,CAAC,KAAK,GAAG,QAAQ;YAAE,QAAQ,GAAG,CAAC,CAAC,KAAK,CAAC;IAC7C,CAAC;IACD,MAAM,KAAK,GAAG,QAAQ,GAAG,QAAQ,IAAI,CAAC,CAAC;IACvC,MAAM,OAAO,GAAG,UAAU,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,CAAC,KAAK,GAAG,QAAQ,CAAC,GAAG,KAAK,CAAC,CAAC;IAEpE,2EAA2E;IAC3E,wEAAwE;IACxE,MAAM,aAAa,GAA2B,UAAU,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE;QACjE,MAAM,GAAG,GAAG,OAAO,CAAC,GAAG,CAAC,CAAC,CAAC,KAAK,CAAC,EAAE,CAAC,CAAC;QACpC,IAAI,GAAG,KAAK,SAAS;YAAE,OAAO,IAAI,CAAC;QACnC,MAAM,MAAM,GAAG,GAAG,GAAG,UAAU,CAAC;QAChC,OAAO,KAAK,CAAC,IAAI,CAAC,OAAO,CAAC,QAAQ,CAAC,MAAM,EAAE,MAAM,GAAG,UAAU,CAAC,CAAC,CAAC;IACnE,CAAC,CAAC,CAAC;IAEH,MAAM,QAAQ,GAAmB,EAAE,CAAC;IACpC,4EAA4E;IAC5E,MAAM,cAAc,GAAa,EAAE,CAAC;IAEpC,mEAAmE;IACnE,MAAM,SAAS,GAAG,UAAU,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,CAAC;IAE9C,OAAO,QAAQ,CAAC,MAAM,GAAG,IAAI,IAAI,SAAS,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC;QACtD,IAAI,OAAO,GAAG,CAAC,CAAC,CAAC,CAAC,0BAA0B;QAC5C,IAAI,OAAO,GAAG,CAAC,QAAQ,CAAC;QAExB,KAAK,IAAI,GAAG,GAAG,CAAC,EAAE,GAAG,GAAG,SAAS,CAAC,MAAM,EAAE,GAAG,EAAE,EAAE,CAAC;YAChD,MAAM,OAAO,GAAG,SAAS,CAAC,GAAG,CAAC,CAAC;YAC/B,MAAM,SAAS,GAAG,OAAO,CAAC,OAAO,CAAC,CAAC;YAEnC,IAAI,MAAM,GAAG,CAAC,CAAC;YACf,IAAI,cAAc,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC;gBAC9B,MAAM,IAAI,GAAG,aAAa,CAAC,OAAO,CAAC,CAAC;gBACpC,IAAI,IAAI,KAAK,IAAI,EAAE,CAAC;oBAClB,KAAK,MAAM,MAAM,IAAI,cAAc,EAAE,CAAC;wBACpC,MAAM,GAAG,GAAG,IAAA,+BAAsB,EAChC,IAAI,EACJ,OAAO,EACP,MAAM,GAAG,UAAU,EACnB,UAAU,CACX,CAAC;wBACF,IAAI,GAAG,GAAG,MAAM;4BAAE,MAAM,GAAG,GAAG,CAAC;oBACjC,CAAC;gBACH,CAAC;YACH,CAAC;YAED,MAAM,QAAQ,GAAG,MAAM,GAAG,SAAS,GAAG,CAAC,CAAC,GAAG,MAAM,CAAC,GAAG,MAAM,CAAC;YAC5D,IAAI,QAAQ,GAAG,OAAO,EAAE,CAAC;gBACvB,OAAO,GAAG,QAAQ,CAAC;gBACnB,OAAO,GAAG,GAAG,CAAC;YAChB,CAAC;QACH,CAAC;QAED,MAAM,SAAS,GAAG,SAAS,CAAC,OAAO,CAAC,CAAC;QACrC,QAAQ,CAAC,IAAI,CAAC,UAAU,CAAC,SAAS,CAAC,CAAC,CAAC;QAErC,MAAM,MAAM,GAAG,OAAO,CAAC,GAAG,CAAC,UAAU,CAAC,SAAS,CAAC,CAAC,KAAK,CAAC,EAAE,CAAC,CAAC;QAC3D,IAAI,MAAM,KAAK,SAAS;YAAE,cAAc,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC;QAEtD,SAAS,CAAC,MAAM,CAAC,OAAO,EAAE,CAAC,CAAC,CAAC;IAC/B,CAAC;IAED,OAAO,QAAQ,CAAC;AAClB,CAAC"}

package/dist/search/rerank.d.ts

@@ -0,0 +1,12 @@
+import type { RerankConfig, SearchResult } from "../core/types";
+/**
+ * Re-scores search results using the configured reranker.
+ *
+ * @param query       The user's natural language query
+ * @param results     Initial search results (from vector or keyword search)
+ * @param topK        Number of results to return after reranking
+ * @param config      Reranker config (default: xenova cross-encoder)
+ * @param fallbackModel  Model name to use when config.model is not set (ollama only)
+ */
+export declare function rerankResults(query: string, results: SearchResult[], topK?: number, config?: RerankConfig, fallbackModel?: string): Promise<SearchResult[]>;
+//# sourceMappingURL=rerank.d.ts.map

package/dist/search/rerank.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"rerank.d.ts","sourceRoot":"","sources":["../../src/search/rerank.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,YAAY,EAAE,YAAY,EAAE,MAAM,eAAe,CAAC;AA0HhE;;;;;;;;GAQG;AACH,wBAAsB,aAAa,CACjC,KAAK,EAAE,MAAM,EACb,OAAO,EAAE,YAAY,EAAE,EACvB,IAAI,SAAK,EACT,MAAM,CAAC,EAAE,YAAY,EACrB,aAAa,CAAC,EAAE,MAAM,GACrB,OAAO,CAAC,YAAY,EAAE,CAAC,CAezB"}

package/dist/search/rerank.js

@@ -0,0 +1,113 @@
+"use strict";
+Object.defineProperty(exports, "__esModule", { value: true });
+exports.rerankResults = rerankResults;
+const ollama_1 = require("../llm/ollama");
+// ─── Xenova cross-encoder (original implementation) ──────────────────────────
+// biome-ignore lint/suspicious/noExplicitAny: @xenova/transformers has no exported TS types
+let xenovaModel = null;
+// biome-ignore lint/suspicious/noExplicitAny: @xenova/transformers has no exported TS types
+let xenovaTokenizer = null;
+async function initXenovaReranker() {
+    if (!xenovaModel) {
+        // biome-ignore lint/suspicious/noExplicitAny: dynamic optional dependency
+        let transformers;
+        try {
+            transformers = require("@xenova/transformers");
+        }
+        catch {
+            throw new Error("The xenova reranker requires @xenova/transformers.\n" +
+                "Install it with: npm install @xenova/transformers\n" +
+                'Or switch to the ollama reranker: set reranker.provider = "ollama" in .vemora/config.json');
+        }
+        const { AutoModelForSequenceClassification, AutoTokenizer } = transformers;
+        xenovaModel = await AutoModelForSequenceClassification.from_pretrained("Xenova/ms-marco-MiniLM-L-6-v2", { quantized: false });
+        xenovaTokenizer = await AutoTokenizer.from_pretrained("Xenova/ms-marco-MiniLM-L-6-v2");
+    }
+}
+async function rerankXenova(query, results, topK) {
+    await initXenovaReranker();
+    const reranked = [];
+    const candidates = results.slice(0, 25);
+    for (const res of candidates) {
+        try {
+            const inputs = await xenovaTokenizer(query, {
+                text_pair: res.chunk.content,
+                truncation: true,
+                padding: true,
+            });
+            const { logits } = await xenovaModel(inputs);
+            reranked.push({ ...res, score: logits.data[0] });
+        }
+        catch {
+            reranked.push(res);
+        }
+    }
+    reranked.sort((a, b) => b.score - a.score);
+    return reranked.slice(0, topK);
+}
+// ─── Ollama LLM reranker ──────────────────────────────────────────────────────
+async function rerankOllama(query, results, topK, config, fallbackModel) {
+    const candidates = results.slice(0, 25);
+    if (candidates.length === 0)
+        return [];
+    const model = config.model ?? fallbackModel ?? "gemma4:e4b";
+    const ollama = new ollama_1.OllamaProvider(config.baseUrl ?? "http://localhost:11434");
+    const chunkList = candidates
+        .map((r, i) => {
+        const snippet = r.chunk.content.slice(0, 300).replace(/\n/g, " ");
+        return `[${i}] ${r.chunk.file}:${r.chunk.start}\n${snippet}`;
+    })
+        .join("\n\n");
+    const prompt = `Rank the following code chunks by relevance to the query. ` +
+        `Return ONLY a JSON array of 0-based indices, most relevant first. ` +
+        `Example for 3 chunks: [2, 0, 1]\n\n` +
+        `Query: "${query}"\n\n` +
+        `Chunks:\n${chunkList}`;
+    try {
+        const response = await ollama.chat([{ role: "user", content: prompt }], { model, temperature: 0, maxTokens: 100 });
+        const raw = response.content.trim();
+        // Extract JSON array from response (model may wrap it in prose)
+        const match = raw.match(/\[[\d,\s]+\]/);
+        if (!match)
+            return results.slice(0, topK);
+        const indices = JSON.parse(match[0]);
+        const reranked = indices
+            .filter((i) => i >= 0 && i < candidates.length)
+            .map((i) => candidates[i]);
+        // Append any candidates not mentioned by the model, preserving original order
+        const seen = new Set(indices);
+        for (let i = 0; i < candidates.length; i++) {
+            if (!seen.has(i))
+                reranked.push(candidates[i]);
+        }
+        return reranked.slice(0, topK);
+    }
+    catch {
+        // Graceful fallback: return original order
+        return results.slice(0, topK);
+    }
+}
+// ─── Public API ───────────────────────────────────────────────────────────────
+/**
+ * Re-scores search results using the configured reranker.
+ *
+ * @param query       The user's natural language query
+ * @param results     Initial search results (from vector or keyword search)
+ * @param topK        Number of results to return after reranking
+ * @param config      Reranker config (default: xenova cross-encoder)
+ * @param fallbackModel  Model name to use when config.model is not set (ollama only)
+ */
+async function rerankResults(query, results, topK = 10, config, fallbackModel) {
+    if (results.length === 0)
+        return [];
+    const provider = config?.provider ?? "xenova";
+    if (provider === "none") {
+        return results.slice(0, topK);
+    }
+    if (provider === "ollama") {
+        return rerankOllama(query, results, topK, config, fallbackModel);
+    }
+    // Default: xenova
+    return rerankXenova(query, results, topK);
+}
+//# sourceMappingURL=rerank.js.map

package/dist/search/rerank.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"rerank.js","sourceRoot":"","sources":["../../src/search/rerank.ts"],"names":[],"mappings":";;AAmIA,sCAqBC;AAvJD,0CAA+C;AAE/C,gFAAgF;AAEhF,4FAA4F;AAC5F,IAAI,WAAW,GAAQ,IAAI,CAAC;AAC5B,4FAA4F;AAC5F,IAAI,eAAe,GAAQ,IAAI,CAAC;AAEhC,KAAK,UAAU,kBAAkB;IAC/B,IAAI,CAAC,WAAW,EAAE,CAAC;QACjB,0EAA0E;QAC1E,IAAI,YAAiB,CAAC;QACtB,IAAI,CAAC;YACH,YAAY,GAAG,OAAO,CAAC,sBAAsB,CAAC,CAAC;QACjD,CAAC;QAAC,MAAM,CAAC;YACP,MAAM,IAAI,KAAK,CACb,sDAAsD;gBACpD,qDAAqD;gBACrD,2FAA2F,CAC9F,CAAC;QACJ,CAAC;QACD,MAAM,EAAE,kCAAkC,EAAE,aAAa,EAAE,GAAG,YAAY,CAAC;QAC3E,WAAW,GAAG,MAAM,kCAAkC,CAAC,eAAe,CACpE,+BAA+B,EAC/B,EAAE,SAAS,EAAE,KAAK,EAAE,CACrB,CAAC;QACF,eAAe,GAAG,MAAM,aAAa,CAAC,eAAe,CACnD,+BAA+B,CAChC,CAAC;IACJ,CAAC;AACH,CAAC;AAED,KAAK,UAAU,YAAY,CACzB,KAAa,EACb,OAAuB,EACvB,IAAY;IAEZ,MAAM,kBAAkB,EAAE,CAAC;IAE3B,MAAM,QAAQ,GAAmB,EAAE,CAAC;IACpC,MAAM,UAAU,GAAG,OAAO,CAAC,KAAK,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC;IAExC,KAAK,MAAM,GAAG,IAAI,UAAU,EAAE,CAAC;QAC7B,IAAI,CAAC;YACH,MAAM,MAAM,GAAG,MAAM,eAAe,CAAC,KAAK,EAAE;gBAC1C,SAAS,EAAE,GAAG,CAAC,KAAK,CAAC,OAAO;gBAC5B,UAAU,EAAE,IAAI;gBAChB,OAAO,EAAE,IAAI;aACd,CAAC,CAAC;YACH,MAAM,EAAE,MAAM,EAAE,GAAG,MAAM,WAAW,CAAC,MAAM,CAAC,CAAC;YAC7C,QAAQ,CAAC,IAAI,CAAC,EAAE,GAAG,GAAG,EAAE,KAAK,EAAE,MAAM,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC;QACnD,CAAC;QAAC,MAAM,CAAC;YACP,QAAQ,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;QACrB,CAAC;IACH,CAAC;IAED,QAAQ,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,KAAK,GAAG,CAAC,CAAC,KAAK,CAAC,CAAC;IAC3C,OAAO,QAAQ,CAAC,KAAK,CAAC,CAAC,EAAE,IAAI,CAAC,CAAC;AACjC,CAAC;AAED,iFAAiF;AAEjF,KAAK,UAAU,YAAY,CACzB,KAAa,EACb,OAAuB,EACvB,IAAY,EACZ,MAAoB,EACpB,aAAsB;IAEtB,MAAM,UAAU,GAAG,OAAO,CAAC,KAAK,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC;IACxC,IAAI,UAAU,CAAC,MAAM,KAAK,CAAC;QAAE,OAAO,EAAE,CAAC;IAEvC,MAAM,KAAK,GAAG,MAAM,CAAC,KAAK,IAAI,aAAa,IAAI,YAAY,CAAC;IAC5D,MAAM,MAAM,GAAG,IAAI,uBAAc,CAAC,MAAM,CAAC,OAAO,IAAI,wBAAwB,CAAC,CAAC;IAE9E,MAAM,SAAS,GAAG,UAAU;SACzB,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE;QACZ,MAAM,OAAO,GAAG,CAAC,CAAC,KAAK,CAAC,OAAO,CAAC,KAAK,CAAC,CAAC,EAAE,GAAG,CAAC,CAAC,OAAO,CAAC,KAAK,EAAE,GAAG,CAAC,CAAC;QAClE,OAAO,IAAI,CAAC,KAAK,CAAC,CAAC,KAAK,CAAC,IAAI,IAAI,CAAC,CAAC,KAAK,CAAC,KAAK,KAAK,OAAO,EAAE,CAAC;IAC/D,CAAC,CAAC;SACD,IAAI,CAAC,MAAM,CAAC,CAAC;IAEhB,MAAM,MAAM,GACV,4DAA4D;QAC5D,oEAAoE;QACpE,qCAAqC;QACrC,WAAW,KAAK,OAAO;QACvB,YAAY,SAAS,EAAE,CAAC;IAE1B,IAAI,CAAC;QACH,MAAM,QAAQ,GAAG,MAAM,MAAM,CAAC,IAAI,CAChC,CAAC,EAAE,IAAI,EAAE,MAAM,EAAE,OAAO,EAAE,MAAM,EAAE,CAAC,EACnC,EAAE,KAAK,EAAE,WAAW,EAAE,CAAC,EAAE,SAAS,EAAE,GAAG,EAAE,CAC1C,CAAC;QAEF,MAAM,GAAG,GAAG,QAAQ,CAAC,OAAO,CAAC,IAAI,EAAE,CAAC;QACpC,gEAAgE;QAChE,MAAM,KAAK,GAAG,GAAG,CAAC,KAAK,CAAC,cAAc,CAAC,CAAC;QACxC,IAAI,CAAC,KAAK;YAAE,OAAO,OAAO,CAAC,KAAK,CAAC,CAAC,EAAE,IAAI,CAAC,CAAC;QAE1C,MAAM,OAAO,GAAa,IAAI,CAAC,KAAK,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC;QAC/C,MAAM,QAAQ,GAAG,OAAO;aACrB,MAAM,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,IAAI,CAAC,IAAI,CAAC,GAAG,UAAU,CAAC,MAAM,CAAC;aAC9C,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,UAAU,CAAC,CAAC,CAAC,CAAC,CAAC;QAE7B,8EAA8E;QAC9E,MAAM,IAAI,GAAG,IAAI,GAAG,CAAC,OAAO,CAAC,CAAC;QAC9B,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,UAAU,CAAC,MAAM,EAAE,CAAC,EAAE,EAAE,CAAC;YAC3C,IAAI,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC;gBAAE,QAAQ,CAAC,IAAI,CAAC,UAAU,CAAC,CAAC,CAAC,CAAC,CAAC;QACjD,CAAC;QAED,OAAO,QAAQ,CAAC,KAAK,CAAC,CAAC,EAAE,IAAI,CAAC,CAAC;IACjC,CAAC;IAAC,MAAM,CAAC;QACP,2CAA2C;QAC3C,OAAO,OAAO,CAAC,KAAK,CAAC,CAAC,EAAE,IAAI,CAAC,CAAC;IAChC,CAAC;AACH,CAAC;AAED,iFAAiF;AAEjF;;;;;;;;GAQG;AACI,KAAK,UAAU,aAAa,CACjC,KAAa,EACb,OAAuB,EACvB,IAAI,GAAG,EAAE,EACT,MAAqB,EACrB,aAAsB;IAEtB,IAAI,OAAO,CAAC,MAAM,KAAK,CAAC;QAAE,OAAO,EAAE,CAAC;IAEpC,MAAM,QAAQ,GAAG,MAAM,EAAE,QAAQ,IAAI,QAAQ,CAAC;IAE9C,IAAI,QAAQ,KAAK,MAAM,EAAE,CAAC;QACxB,OAAO,OAAO,CAAC,KAAK,CAAC,CAAC,EAAE,IAAI,CAAC,CAAC;IAChC,CAAC;IAED,IAAI,QAAQ,KAAK,QAAQ,EAAE,CAAC;QAC1B,OAAO,YAAY,CAAC,KAAK,EAAE,OAAO,EAAE,IAAI,EAAE,MAAO,EAAE,aAAa,CAAC,CAAC;IACpE,CAAC;IAED,kBAAkB;IAClB,OAAO,YAAY,CAAC,KAAK,EAAE,OAAO,EAAE,IAAI,CAAC,CAAC;AAC5C,CAAC"}

package/dist/search/signature.d.ts

@@ -0,0 +1,42 @@
+/**
+ * Signature extraction for code chunks.
+ *
+ * Given the raw content of a chunk, returns the declaration signature —
+ * the part that describes *what* exists without the implementation body.
+ * This is used in the query output for medium-relevance results.
+ *
+ * Design philosophy:
+ *   - Interfaces and type aliases ARE their signature → show them in full (compact)
+ *   - Functions and classes → extract up to the opening brace, replace body with { … }
+ *   - Arrow functions with expression bodies → show up to =>
+ *   - Everything else → show first meaningful lines
+ */
+/**
+ * Extracts the declaration signature from a chunk's content.
+ *
+ * Examples:
+ *
+ *   Input:  "export async function connect(\n  host: string,\n): Promise<void> {\n  ..."
+ *   Output: "export async function connect(\n  host: string,\n): Promise<void> { … }"
+ *
+ *   Input:  "export interface ImapConfig {\n  host: string;\n}"
+ *   Output: (same — interfaces are returned as-is, they are compact)
+ *
+ *   Input:  "export const send = async (msg: Email): Promise<void> =>\n  smtp.send(msg);"
+ *   Output: "export const send = async (msg: Email): Promise<void> => …"
+ */
+export declare function extractSignature(content: string): string;
+export type DisplayTier = "high" | "med" | "low";
+/**
+ * Determines how much of a chunk to show based on its rank in results.
+ *
+ *   high (rank 1-3)  → full code block (capped at MAX_HIGH_LINES automatically)
+ *   med  (rank 4-7)  → signature only (declaration without body)
+ *   low  (rank 8+)   → file + symbol + score only (no code)
+ *
+ * The --show-code flag overrides all tiers to show full code.
+ */
+export declare function getDisplayTier(rank: number): DisplayTier;
+/** Lines shown automatically for high-tier results (without --show-code) */
+export declare const HIGH_CODE_LINES = 30;
+//# sourceMappingURL=signature.d.ts.map

package/dist/search/signature.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"signature.d.ts","sourceRoot":"","sources":["../../src/search/signature.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;GAYG;AAMH;;;;;;;;;;;;;GAaG;AACH,wBAAgB,gBAAgB,CAAC,OAAO,EAAE,MAAM,GAAG,MAAM,CAmExD;AAID,MAAM,MAAM,WAAW,GAAG,MAAM,GAAG,KAAK,GAAG,KAAK,CAAC;AAEjD;;;;;;;;GAQG;AACH,wBAAgB,cAAc,CAAC,IAAI,EAAE,MAAM,GAAG,WAAW,CAIxD;AAED,4EAA4E;AAC5E,eAAO,MAAM,eAAe,KAAK,CAAC"}

package/dist/search/signature.js

@@ -0,0 +1,112 @@
+"use strict";
+/**
+ * Signature extraction for code chunks.
+ *
+ * Given the raw content of a chunk, returns the declaration signature —
+ * the part that describes *what* exists without the implementation body.
+ * This is used in the query output for medium-relevance results.
+ *
+ * Design philosophy:
+ *   - Interfaces and type aliases ARE their signature → show them in full (compact)
+ *   - Functions and classes → extract up to the opening brace, replace body with { … }
+ *   - Arrow functions with expression bodies → show up to =>
+ *   - Everything else → show first meaningful lines
+ */
+Object.defineProperty(exports, "__esModule", { value: true });
+exports.HIGH_CODE_LINES = void 0;
+exports.extractSignature = extractSignature;
+exports.getDisplayTier = getDisplayTier;
+const MAX_SIG_LINES = 10;
+// ─── Public API ───────────────────────────────────────────────────────────────
+/**
+ * Extracts the declaration signature from a chunk's content.
+ *
+ * Examples:
+ *
+ *   Input:  "export async function connect(\n  host: string,\n): Promise<void> {\n  ..."
+ *   Output: "export async function connect(\n  host: string,\n): Promise<void> { … }"
+ *
+ *   Input:  "export interface ImapConfig {\n  host: string;\n}"
+ *   Output: (same — interfaces are returned as-is, they are compact)
+ *
+ *   Input:  "export const send = async (msg: Email): Promise<void> =>\n  smtp.send(msg);"
+ *   Output: "export const send = async (msg: Email): Promise<void> => …"
+ */
+function extractSignature(content) {
+    const lines = content.split("\n");
+    const firstMeaningful = lines.find((l) => l.trim().length > 0) ?? "";
+    // ── Interfaces and type aliases ──────────────────────────────────────────
+    // These are inherently compact — the full declaration IS the signature.
+    // Show up to 20 lines (rare for them to be longer).
+    if (/^\s*(export\s+)?(type\s+\w|interface\s+\w)/.test(firstMeaningful)) {
+        const slice = lines.slice(0, 20);
+        if (lines.length > 20)
+            slice.push("  …");
+        return slice.join("\n").trim();
+    }
+    // ── File header / import blocks ──────────────────────────────────────────
+    // Chunks that are just import statements — show first 5 lines.
+    if (/^\s*import\s/.test(firstMeaningful)) {
+        const slice = lines.slice(0, 5);
+        if (lines.length > 5)
+            slice.push(`  … (${lines.length - 5} more lines)`);
+        return slice.join("\n").trim();
+    }
+    // ── Function / class / method declarations ───────────────────────────────
+    const result = [];
+    for (let i = 0; i < lines.length; i++) {
+        const line = lines[i];
+        const trimmed = line.trimEnd();
+        // Line ends with opening brace → body starts here.
+        // Captures patterns like:
+        //   ): Promise<void> {
+        //   export class Foo extends Bar {
+        //   } catch (e) {          (skip mid-function braces — result.length guard)
+        if (/\{\s*(\/\/[^\n]*)?\s*$/.test(trimmed)) {
+            // Strip the brace (and any trailing comment) and append { … }
+            const withoutBrace = trimmed
+                .replace(/\s*\{\s*(\/\/[^\n]*)?\s*$/, "")
+                .trimEnd();
+            result.push((withoutBrace !== ""
+                ? withoutBrace
+                : trimmed.replace(/\{.*$/, "").trim()) + " { … }");
+            break;
+        }
+        // Arrow function with expression body:
+        //   const f = (x: number) =>
+        //     x * 2
+        if (trimmed.endsWith("=>") && !trimmed.endsWith("=>>")) {
+            result.push(trimmed + " …");
+            break;
+        }
+        result.push(line);
+        // Single-line declarations ending with `;` (e.g. `declare function f(): void;`)
+        if (result.length === 1 && trimmed.endsWith(";"))
+            break;
+        // Safety cap
+        if (result.length >= MAX_SIG_LINES) {
+            result.push("  …");
+            break;
+        }
+    }
+    return result.join("\n").trim();
+}
+/**
+ * Determines how much of a chunk to show based on its rank in results.
+ *
+ *   high (rank 1-3)  → full code block (capped at MAX_HIGH_LINES automatically)
+ *   med  (rank 4-7)  → signature only (declaration without body)
+ *   low  (rank 8+)   → file + symbol + score only (no code)
+ *
+ * The --show-code flag overrides all tiers to show full code.
+ */
+function getDisplayTier(rank) {
+    if (rank <= 3)
+        return "high";
+    if (rank <= 7)
+        return "med";
+    return "low";
+}
+/** Lines shown automatically for high-tier results (without --show-code) */
+exports.HIGH_CODE_LINES = 30;
+//# sourceMappingURL=signature.js.map

package/dist/search/signature.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"signature.js","sourceRoot":"","sources":["../../src/search/signature.ts"],"names":[],"mappings":";AAAA;;;;;;;;;;;;GAYG;;;AAoBH,4CAmEC;AAeD,wCAIC;AAxGD,MAAM,aAAa,GAAG,EAAE,CAAC;AAEzB,iFAAiF;AAEjF;;;;;;;;;;;;;GAaG;AACH,SAAgB,gBAAgB,CAAC,OAAe;IAC9C,MAAM,KAAK,GAAG,OAAO,CAAC,KAAK,CAAC,IAAI,CAAC,CAAC;IAClC,MAAM,eAAe,GAAG,KAAK,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC,MAAM,GAAG,CAAC,CAAC,IAAI,EAAE,CAAC;IAErE,4EAA4E;IAC5E,wEAAwE;IACxE,oDAAoD;IACpD,IAAI,4CAA4C,CAAC,IAAI,CAAC,eAAe,CAAC,EAAE,CAAC;QACvE,MAAM,KAAK,GAAG,KAAK,CAAC,KAAK,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC;QACjC,IAAI,KAAK,CAAC,MAAM,GAAG,EAAE;YAAE,KAAK,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC;QACzC,OAAO,KAAK,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC,IAAI,EAAE,CAAC;IACjC,CAAC;IAED,4EAA4E;IAC5E,+DAA+D;IAC/D,IAAI,cAAc,CAAC,IAAI,CAAC,eAAe,CAAC,EAAE,CAAC;QACzC,MAAM,KAAK,GAAG,KAAK,CAAC,KAAK,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC;QAChC,IAAI,KAAK,CAAC,MAAM,GAAG,CAAC;YAAE,KAAK,CAAC,IAAI,CAAC,QAAQ,KAAK,CAAC,MAAM,GAAG,CAAC,cAAc,CAAC,CAAC;QACzE,OAAO,KAAK,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC,IAAI,EAAE,CAAC;IACjC,CAAC;IAED,4EAA4E;IAC5E,MAAM,MAAM,GAAa,EAAE,CAAC;IAE5B,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,KAAK,CAAC,MAAM,EAAE,CAAC,EAAE,EAAE,CAAC;QACtC,MAAM,IAAI,GAAG,KAAK,CAAC,CAAC,CAAC,CAAC;QACtB,MAAM,OAAO,GAAG,IAAI,CAAC,OAAO,EAAE,CAAC;QAE/B,mDAAmD;QACnD,0BAA0B;QAC1B,uBAAuB;QACvB,mCAAmC;QACnC,4EAA4E;QAC5E,IAAI,wBAAwB,CAAC,IAAI,CAAC,OAAO,CAAC,EAAE,CAAC;YAC3C,8DAA8D;YAC9D,MAAM,YAAY,GAAG,OAAO;iBACzB,OAAO,CAAC,2BAA2B,EAAE,EAAE,CAAC;iBACxC,OAAO,EAAE,CAAC;YACb,MAAM,CAAC,IAAI,CACT,CAAC,YAAY,KAAK,EAAE;gBAClB,CAAC,CAAC,YAAY;gBACd,CAAC,CAAC,OAAO,CAAC,OAAO,CAAC,OAAO,EAAE,EAAE,CAAC,CAAC,IAAI,EAAE,CAAC,GAAG,QAAQ,CACpD,CAAC;YACF,MAAM;QACR,CAAC;QAED,uCAAuC;QACvC,6BAA6B;QAC7B,YAAY;QACZ,IAAI,OAAO,CAAC,QAAQ,CAAC,IAAI,CAAC,IAAI,CAAC,OAAO,CAAC,QAAQ,CAAC,KAAK,CAAC,EAAE,CAAC;YACvD,MAAM,CAAC,IAAI,CAAC,OAAO,GAAG,IAAI,CAAC,CAAC;YAC5B,MAAM;QACR,CAAC;QAED,MAAM,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;QAElB,gFAAgF;QAChF,IAAI,MAAM,CAAC,MAAM,KAAK,CAAC,IAAI,OAAO,CAAC,QAAQ,CAAC,GAAG,CAAC;YAAE,MAAM;QAExD,aAAa;QACb,IAAI,MAAM,CAAC,MAAM,IAAI,aAAa,EAAE,CAAC;YACnC,MAAM,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC;YACnB,MAAM;QACR,CAAC;IACH,CAAC;IAED,OAAO,MAAM,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC,IAAI,EAAE,CAAC;AAClC,CAAC;AAMD;;;;;;;;GAQG;AACH,SAAgB,cAAc,CAAC,IAAY;IACzC,IAAI,IAAI,IAAI,CAAC;QAAE,OAAO,MAAM,CAAC;IAC7B,IAAI,IAAI,IAAI,CAAC;QAAE,OAAO,KAAK,CAAC;IAC5B,OAAO,KAAK,CAAC;AACf,CAAC;AAED,4EAA4E;AAC/D,QAAA,eAAe,GAAG,EAAE,CAAC"}

package/dist/search/vector.d.ts

@@ -0,0 +1,41 @@
+import type { Chunk, EmbeddingCache, SearchResult, SymbolIndex } from "../core/types";
+/**
+ * Computes cosine similarity between two vectors.
+ * Returns a value in [-1, 1]. For normalized embeddings (OpenAI, Ollama)
+ * this is effectively in [0, 1] for semantically related pairs.
+ */
+export declare function cosineSimilarity(a: number[], b: number[]): number;
+/**
+ * Fast cosine similarity between a number[] and a Float32Array slice.
+ */
+export declare function cosineSimilarityBinary(query: number[], vectors: Float32Array, offset: number, dims: number): number;
+/**
+ * Performs an exhaustive nearest-neighbor search over all cached embeddings.
+ *
+ * Uses HNSW (O(log N)) when the index is available in the cache.
+ * Falls back to exhaustive O(n * d) cosine search on the contiguous
+ * Float32Array buffer if the HNSW index is missing or fails to load.
+ */
+export declare function vectorSearch(queryEmbedding: number[], chunks: Chunk[], cache: EmbeddingCache, symbols: SymbolIndex, topK?: number): SearchResult[];
+/**
+ * Direct symbol lookup — bypasses embedding when the query matches a known
+ * symbol name. Returns chunks for matched symbols scored by match quality.
+ *
+ * Match tiers (in order of precedence):
+ *   1.0 — exact match: query === symbolName (case-insensitive)
+ *   0.95 — single-word match: one word in query === symbolName
+ *   0.80 — prefix match: symbolName starts with a query word (or vice versa)
+ *
+ * Returns [] if no symbol matches (caller should fall through to vector/BM25).
+ * Only activates for narrow queries (≤5 matches) to avoid false positives on
+ * generic names like "connect" or "init".
+ */
+export declare function symbolLookup(query: string, chunks: Chunk[], symbols: SymbolIndex): SearchResult[];
+/**
+ * TF-based keyword search — used as fallback when embeddings are not available.
+ *
+ * Scores each chunk by term frequency normalized by content length.
+ * Not as powerful as semantic search but requires no embedding service.
+ */
+export declare function keywordSearch(query: string, chunks: Chunk[], symbols: SymbolIndex, topK?: number): SearchResult[];
+//# sourceMappingURL=vector.d.ts.map

package/dist/search/vector.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"vector.d.ts","sourceRoot":"","sources":["../../src/search/vector.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EACV,KAAK,EACL,cAAc,EACd,YAAY,EACZ,WAAW,EACZ,MAAM,eAAe,CAAC;AAIvB;;;;GAIG;AACH,wBAAgB,gBAAgB,CAAC,CAAC,EAAE,MAAM,EAAE,EAAE,CAAC,EAAE,MAAM,EAAE,GAAG,MAAM,CAejE;AAED;;GAEG;AACH,wBAAgB,sBAAsB,CACpC,KAAK,EAAE,MAAM,EAAE,EACf,OAAO,EAAE,YAAY,EACrB,MAAM,EAAE,MAAM,EACd,IAAI,EAAE,MAAM,GACX,MAAM,CAcR;AAID;;;;;;GAMG;AACH,wBAAgB,YAAY,CAC1B,cAAc,EAAE,MAAM,EAAE,EACxB,MAAM,EAAE,KAAK,EAAE,EACf,KAAK,EAAE,cAAc,EACrB,OAAO,EAAE,WAAW,EACpB,IAAI,SAAK,GACR,YAAY,EAAE,CA4DhB;AAID;;;;;;;;;;;;GAYG;AACH,wBAAgB,YAAY,CAC1B,KAAK,EAAE,MAAM,EACb,MAAM,EAAE,KAAK,EAAE,EACf,OAAO,EAAE,WAAW,GACnB,YAAY,EAAE,CA+BhB;AAID;;;;;GAKG;AACH,wBAAgB,aAAa,CAC3B,KAAK,EAAE,MAAM,EACb,MAAM,EAAE,KAAK,EAAE,EACf,OAAO,EAAE,WAAW,EACpB,IAAI,SAAK,GACR,YAAY,EAAE,CAyChB"}

package/dist/search/vector.js

@@ -0,0 +1,185 @@
+"use strict";
+Object.defineProperty(exports, "__esModule", { value: true });
+exports.cosineSimilarity = cosineSimilarity;
+exports.cosineSimilarityBinary = cosineSimilarityBinary;
+exports.vectorSearch = vectorSearch;
+exports.symbolLookup = symbolLookup;
+exports.keywordSearch = keywordSearch;
+// ─── Cosine Similarity ────────────────────────────────────────────────────────
+/**
+ * Computes cosine similarity between two vectors.
+ * Returns a value in [-1, 1]. For normalized embeddings (OpenAI, Ollama)
+ * this is effectively in [0, 1] for semantically related pairs.
+ */
+function cosineSimilarity(a, b) {
+    if (a.length === 0 || a.length !== b.length)
+        return 0;
+    let dot = 0;
+    let normA = 0;
+    let normB = 0;
+    for (let i = 0; i < a.length; i++) {
+        dot += a[i] * b[i];
+        normA += a[i] * a[i];
+        normB += b[i] * b[i];
+    }
+    const denom = Math.sqrt(normA) * Math.sqrt(normB);
+    return denom === 0 ? 0 : dot / denom;
+}
+/**
+ * Fast cosine similarity between a number[] and a Float32Array slice.
+ */
+function cosineSimilarityBinary(query, vectors, offset, dims) {
+    let dot = 0;
+    let normA = 0;
+    let normB = 0;
+    for (let i = 0; i < dims; i++) {
+        const valB = vectors[offset + i];
+        dot += query[i] * valB;
+        normA += query[i] * query[i];
+        normB += valB * valB;
+    }
+    const denom = Math.sqrt(normA) * Math.sqrt(normB);
+    return denom === 0 ? 0 : dot / denom;
+}
+// ─── Vector Search ────────────────────────────────────────────────────────────
+/**
+ * Performs an exhaustive nearest-neighbor search over all cached embeddings.
+ *
+ * Uses HNSW (O(log N)) when the index is available in the cache.
+ * Falls back to exhaustive O(n * d) cosine search on the contiguous
+ * Float32Array buffer if the HNSW index is missing or fails to load.
+ */
+function vectorSearch(queryEmbedding, chunks, cache, symbols, topK = 10) {
+    if (queryEmbedding.length === 0)
+        return [];
+    const scored = [];
+    const { vectors, chunkIds, dimensions, hnswIndex } = cache;
+    if (!vectors || !chunkIds) {
+        return [];
+    }
+    if (hnswIndex) {
+        try {
+            const { HNSW } = require("hnsw");
+            const index = HNSW.fromJSON(hnswIndex);
+            // Increased efSearch for better accuracy
+            const hnswResults = index.searchKNN(queryEmbedding, Math.max(topK * 2, 20), { efSearch: 64 });
+            const chunkById = new Map(chunks.map(c => [c.id, c]));
+            const chunkResults = [];
+            for (const res of hnswResults) {
+                const chunkId = chunkIds[res.id];
+                const chunk = chunkById.get(chunkId);
+                if (chunk) {
+                    const symbol = chunk.symbol ? symbols[chunk.symbol] : undefined;
+                    chunkResults.push({ chunk, score: res.score, symbol });
+                }
+            }
+            chunkResults.sort((a, b) => b.score - a.score);
+            return chunkResults.slice(0, topK);
+        }
+        catch (e) {
+            console.warn("HNSW search failed, falling back to exhaustive search:", e);
+        }
+    }
+    // Fallback: Optimized path: binary search over contiguous buffer
+    // Map chunkId to index for fast lookup
+    const idToIndex = new Map();
+    chunkIds.forEach((id, i) => idToIndex.set(id, i));
+    for (const chunk of chunks) {
+        const idx = idToIndex.get(chunk.id);
+        if (idx === undefined)
+            continue;
+        const score = cosineSimilarityBinary(queryEmbedding, vectors, idx * dimensions, dimensions);
+        const symbol = chunk.symbol ? symbols[chunk.symbol] : undefined;
+        scored.push({ chunk, score, symbol });
+    }
+    scored.sort((a, b) => b.score - a.score);
+    return scored.slice(0, topK);
+}
+// ─── Symbol Lookup ────────────────────────────────────────────────────────────
+/**
+ * Direct symbol lookup — bypasses embedding when the query matches a known
+ * symbol name. Returns chunks for matched symbols scored by match quality.
+ *
+ * Match tiers (in order of precedence):
+ *   1.0 — exact match: query === symbolName (case-insensitive)
+ *   0.95 — single-word match: one word in query === symbolName
+ *   0.80 — prefix match: symbolName starts with a query word (or vice versa)
+ *
+ * Returns [] if no symbol matches (caller should fall through to vector/BM25).
+ * Only activates for narrow queries (≤5 matches) to avoid false positives on
+ * generic names like "connect" or "init".
+ */
+function symbolLookup(query, chunks, symbols) {
+    const q = query.trim().toLowerCase();
+    const words = q.split(/[\s\W]+/).filter((w) => w.length >= 2);
+    const candidates = [];
+    for (const name of Object.keys(symbols)) {
+        const lower = name.toLowerCase();
+        if (lower === q) {
+            candidates.push({ name, score: 1.0 });
+        }
+        else if (words.some((w) => lower === w)) {
+            candidates.push({ name, score: 0.95 });
+        }
+        else if (words.some((w) => lower.startsWith(w) || w.startsWith(lower))) {
+            candidates.push({ name, score: 0.8 });
+        }
+    }
+    if (candidates.length === 0 || candidates.length > 5)
+        return [];
+    candidates.sort((a, b) => b.score - a.score);
+    const results = [];
+    for (const { name, score } of candidates) {
+        const sym = symbols[name];
+        const chunk = chunks.find((c) => c.symbol === name && c.file === sym.file);
+        if (chunk) {
+            results.push({ chunk, score, symbol: sym });
+        }
+    }
+    return results;
+}
+// ─── Keyword Search ───────────────────────────────────────────────────────────
+/**
+ * TF-based keyword search — used as fallback when embeddings are not available.
+ *
+ * Scores each chunk by term frequency normalized by content length.
+ * Not as powerful as semantic search but requires no embedding service.
+ */
+function keywordSearch(query, chunks, symbols, topK = 10) {
+    // Split query into meaningful terms (skip stop words shorter than 3 chars)
+    const terms = query
+        .toLowerCase()
+        .split(/[\s\W]+/)
+        .filter((t) => t.length >= 3);
+    if (terms.length === 0)
+        return [];
+    const scored = chunks.map((chunk) => {
+        const content = chunk.content.toLowerCase();
+        let score = 0;
+        for (const term of terms) {
+            // Count occurrences
+            let pos = 0;
+            let count = 0;
+            while ((pos = content.indexOf(term, pos)) !== -1) {
+                count++;
+                pos += term.length;
+            }
+            if (count > 0) {
+                // TF normalized by content length (log-scaled to dampen large files)
+                score += count / Math.log(content.length + 2);
+            }
+        }
+        // Bonus: symbol name matches the query terms
+        if (chunk.symbol) {
+            const symLower = chunk.symbol.toLowerCase();
+            if (terms.some((t) => symLower.includes(t))) {
+                score *= 2;
+            }
+        }
+        const symbol = chunk.symbol ? symbols[chunk.symbol] : undefined;
+        return { chunk, score, symbol };
+    });
+    scored.sort((a, b) => b.score - a.score);
+    return scored.filter((r) => r.score > 0).slice(0, topK);
+}
+//# sourceMappingURL=vector.js.map

package/dist/search/vector.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"vector.js","sourceRoot":"","sources":["../../src/search/vector.ts"],"names":[],"mappings":";;AAcA,4CAeC;AAKD,wDAmBC;AAWD,oCAkEC;AAiBD,oCAmCC;AAUD,sCA8CC;AAvOD,iFAAiF;AAEjF;;;;GAIG;AACH,SAAgB,gBAAgB,CAAC,CAAW,EAAE,CAAW;IACvD,IAAI,CAAC,CAAC,MAAM,KAAK,CAAC,IAAI,CAAC,CAAC,MAAM,KAAK,CAAC,CAAC,MAAM;QAAE,OAAO,CAAC,CAAC;IAEtD,IAAI,GAAG,GAAG,CAAC,CAAC;IACZ,IAAI,KAAK,GAAG,CAAC,CAAC;IACd,IAAI,KAAK,GAAG,CAAC,CAAC;IAEd,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,CAAC,CAAC,MAAM,EAAE,CAAC,EAAE,EAAE,CAAC;QAClC,GAAG,IAAI,CAAC,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC;QACnB,KAAK,IAAI,CAAC,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC;QACrB,KAAK,IAAI,CAAC,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC;IACvB,CAAC;IAED,MAAM,KAAK,GAAG,IAAI,CAAC,IAAI,CAAC,KAAK,CAAC,GAAG,IAAI,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC;IAClD,OAAO,KAAK,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,GAAG,GAAG,KAAK,CAAC;AACvC,CAAC;AAED;;GAEG;AACH,SAAgB,sBAAsB,CACpC,KAAe,EACf,OAAqB,EACrB,MAAc,EACd,IAAY;IAEZ,IAAI,GAAG,GAAG,CAAC,CAAC;IACZ,IAAI,KAAK,GAAG,CAAC,CAAC;IACd,IAAI,KAAK,GAAG,CAAC,CAAC;IAEd,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,IAAI,EAAE,CAAC,EAAE,EAAE,CAAC;QAC9B,MAAM,IAAI,GAAG,OAAO,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC;QACjC,GAAG,IAAI,KAAK,CAAC,CAAC,CAAC,GAAG,IAAI,CAAC;QACvB,KAAK,IAAI,KAAK,CAAC,CAAC,CAAC,GAAG,KAAK,CAAC,CAAC,CAAC,CAAC;QAC7B,KAAK,IAAI,IAAI,GAAG,IAAI,CAAC;IACvB,CAAC;IAED,MAAM,KAAK,GAAG,IAAI,CAAC,IAAI,CAAC,KAAK,CAAC,GAAG,IAAI,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC;IAClD,OAAO,KAAK,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,GAAG,GAAG,KAAK,CAAC;AACvC,CAAC;AAED,iFAAiF;AAEjF;;;;;;GAMG;AACH,SAAgB,YAAY,CAC1B,cAAwB,EACxB,MAAe,EACf,KAAqB,EACrB,OAAoB,EACpB,IAAI,GAAG,EAAE;IAET,IAAI,cAAc,CAAC,MAAM,KAAK,CAAC;QAAE,OAAO,EAAE,CAAC;IAE3C,MAAM,MAAM,GAAmB,EAAE,CAAC;IAElC,MAAM,EAAE,OAAO,EAAE,QAAQ,EAAE,UAAU,EAAE,SAAS,EAAE,GAAG,KAAK,CAAC;IAE3D,IAAI,CAAC,OAAO,IAAI,CAAC,QAAQ,EAAE,CAAC;QAC1B,OAAO,EAAE,CAAC;IACZ,CAAC;IACD,IAAI,SAAS,EAAE,CAAC;QACd,IAAI,CAAC;YACH,MAAM,EAAE,IAAI,EAAE,GAAG,OAAO,CAAC,MAAM,CAAC,CAAC;YACjC,MAAM,KAAK,GAAG,IAAI,CAAC,QAAQ,CAAC,SAAS,CAAC,CAAC;YACvC,yCAAyC;YACzC,MAAM,WAAW,GAAG,KAAK,CAAC,SAAS,CACjC,cAAc,EACd,IAAI,CAAC,GAAG,CAAC,IAAI,GAAG,CAAC,EAAE,EAAE,CAAC,EACtB,EAAE,QAAQ,EAAE,EAAE,EAAE,CACjB,CAAC;YAEA,MAAM,SAAS,GAAG,IAAI,GAAG,CAAC,MAAM,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,CAAC,CAAC;YACtD,MAAM,YAAY,GAAmB,EAAE,CAAC;YACxC,KAAK,MAAM,GAAG,IAAI,WAAW,EAAE,CAAC;gBAC9B,MAAM,OAAO,GAAG,QAAQ,CAAC,GAAG,CAAC,EAAE,CAAC,CAAC;gBACjC,MAAM,KAAK,GAAG,SAAS,CAAC,GAAG,CAAC,OAAO,CAAC,CAAC;gBACrC,IAAI,KAAK,EAAE,CAAC;oBACV,MAAM,MAAM,GAAG,KAAK,CAAC,MAAM,CAAC,CAAC,CAAC,OAAO,CAAC,KAAK,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC,SAAS,CAAC;oBAChE,YAAY,CAAC,IAAI,CAAC,EAAE,KAAK,EAAE,KAAK,EAAE,GAAG,CAAC,KAAK,EAAE,MAAM,EAAE,CAAC,CAAC;gBACzD,CAAC;YACH,CAAC;YAEH,YAAY,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,KAAK,GAAG,CAAC,CAAC,KAAK,CAAC,CAAC;YAC/C,OAAO,YAAY,CAAC,KAAK,CAAC,CAAC,EAAE,IAAI,CAAC,CAAC;QACrC,CAAC;QAAC,OAAO,CAAC,EAAE,CAAC;YACX,OAAO,CAAC,IAAI,CAAC,wDAAwD,EAAE,CAAC,CAAC,CAAC;QAC5E,CAAC;IACH,CAAC;IAED,iEAAiE;IACjE,uCAAuC;IACvC,MAAM,SAAS,GAAG,IAAI,GAAG,EAAkB,CAAC;IAC5C,QAAQ,CAAC,OAAO,CAAC,CAAC,EAAE,EAAE,CAAC,EAAE,EAAE,CAAC,SAAS,CAAC,GAAG,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,CAAC;IAElD,KAAK,MAAM,KAAK,IAAI,MAAM,EAAE,CAAC;QAC3B,MAAM,GAAG,GAAG,SAAS,CAAC,GAAG,CAAC,KAAK,CAAC,EAAE,CAAC,CAAC;QACpC,IAAI,GAAG,KAAK,SAAS;YAAE,SAAS;QAEhC,MAAM,KAAK,GAAG,sBAAsB,CAClC,cAAc,EACd,OAAQ,EACR,GAAG,GAAG,UAAU,EAChB,UAAU,CACX,CAAC;QACF,MAAM,MAAM,GAAG,KAAK,CAAC,MAAM,CAAC,CAAC,CAAC,OAAO,CAAC,KAAK,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC,SAAS,CAAC;QAChE,MAAM,CAAC,IAAI,CAAC,EAAE,KAAK,EAAE,KAAK,EAAE,MAAM,EAAE,CAAC,CAAC;IACxC,CAAC;IAED,MAAM,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,KAAK,GAAG,CAAC,CAAC,KAAK,CAAC,CAAC;IACzC,OAAO,MAAM,CAAC,KAAK,CAAC,CAAC,EAAE,IAAI,CAAC,CAAC;AAC/B,CAAC;AAED,iFAAiF;AAEjF;;;;;;;;;;;;GAYG;AACH,SAAgB,YAAY,CAC1B,KAAa,EACb,MAAe,EACf,OAAoB;IAEpB,MAAM,CAAC,GAAG,KAAK,CAAC,IAAI,EAAE,CAAC,WAAW,EAAE,CAAC;IACrC,MAAM,KAAK,GAAG,CAAC,CAAC,KAAK,CAAC,SAAS,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,MAAM,IAAI,CAAC,CAAC,CAAC;IAE9D,MAAM,UAAU,GAA2C,EAAE,CAAC;IAE9D,KAAK,MAAM,IAAI,IAAI,MAAM,CAAC,IAAI,CAAC,OAAO,CAAC,EAAE,CAAC;QACxC,MAAM,KAAK,GAAG,IAAI,CAAC,WAAW,EAAE,CAAC;QACjC,IAAI,KAAK,KAAK,CAAC,EAAE,CAAC;YAChB,UAAU,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,KAAK,EAAE,GAAG,EAAE,CAAC,CAAC;QACxC,CAAC;aAAM,IAAI,KAAK,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,KAAK,KAAK,CAAC,CAAC,EAAE,CAAC;YAC1C,UAAU,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,KAAK,EAAE,IAAI,EAAE,CAAC,CAAC;QACzC,CAAC;aAAM,IAAI,KAAK,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,KAAK,CAAC,UAAU,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,UAAU,CAAC,KAAK,CAAC,CAAC,EAAE,CAAC;YACzE,UAAU,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,KAAK,EAAE,GAAG,EAAE,CAAC,CAAC;QACxC,CAAC;IACH,CAAC;IAED,IAAI,UAAU,CAAC,MAAM,KAAK,CAAC,IAAI,UAAU,CAAC,MAAM,GAAG,CAAC;QAAE,OAAO,EAAE,CAAC;IAEhE,UAAU,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,KAAK,GAAG,CAAC,CAAC,KAAK,CAAC,CAAC;IAE7C,MAAM,OAAO,GAAmB,EAAE,CAAC;IACnC,KAAK,MAAM,EAAE,IAAI,EAAE,KAAK,EAAE,IAAI,UAAU,EAAE,CAAC;QACzC,MAAM,GAAG,GAAG,OAAO,CAAC,IAAI,CAAC,CAAC;QAC1B,MAAM,KAAK,GAAG,MAAM,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,MAAM,KAAK,IAAI,IAAI,CAAC,CAAC,IAAI,KAAK,GAAG,CAAC,IAAI,CAAC,CAAC;QAC3E,IAAI,KAAK,EAAE,CAAC;YACV,OAAO,CAAC,IAAI,CAAC,EAAE,KAAK,EAAE,KAAK,EAAE,MAAM,EAAE,GAAG,EAAE,CAAC,CAAC;QAC9C,CAAC;IACH,CAAC;IAED,OAAO,OAAO,CAAC;AACjB,CAAC;AAED,iFAAiF;AAEjF;;;;;GAKG;AACH,SAAgB,aAAa,CAC3B,KAAa,EACb,MAAe,EACf,OAAoB,EACpB,IAAI,GAAG,EAAE;IAET,2EAA2E;IAC3E,MAAM,KAAK,GAAG,KAAK;SAChB,WAAW,EAAE;SACb,KAAK,CAAC,SAAS,CAAC;SAChB,MAAM,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,MAAM,IAAI,CAAC,CAAC,CAAC;IAEhC,IAAI,KAAK,CAAC,MAAM,KAAK,CAAC;QAAE,OAAO,EAAE,CAAC;IAElC,MAAM,MAAM,GAAmB,MAAM,CAAC,GAAG,CAAC,CAAC,KAAK,EAAE,EAAE;QAClD,MAAM,OAAO,GAAG,KAAK,CAAC,OAAO,CAAC,WAAW,EAAE,CAAC;QAC5C,IAAI,KAAK,GAAG,CAAC,CAAC;QAEd,KAAK,MAAM,IAAI,IAAI,KAAK,EAAE,CAAC;YACzB,oBAAoB;YACpB,IAAI,GAAG,GAAG,CAAC,CAAC;YACZ,IAAI,KAAK,GAAG,CAAC,CAAC;YACd,OAAO,CAAC,GAAG,GAAG,OAAO,CAAC,OAAO,CAAC,IAAI,EAAE,GAAG,CAAC,CAAC,KAAK,CAAC,CAAC,EAAE,CAAC;gBACjD,KAAK,EAAE,CAAC;gBACR,GAAG,IAAI,IAAI,CAAC,MAAM,CAAC;YACrB,CAAC;YACD,IAAI,KAAK,GAAG,CAAC,EAAE,CAAC;gBACd,qEAAqE;gBACrE,KAAK,IAAI,KAAK,GAAG,IAAI,CAAC,GAAG,CAAC,OAAO,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC;YAChD,CAAC;QACH,CAAC;QAED,6CAA6C;QAC7C,IAAI,KAAK,CAAC,MAAM,EAAE,CAAC;YACjB,MAAM,QAAQ,GAAG,KAAK,CAAC,MAAM,CAAC,WAAW,EAAE,CAAC;YAC5C,IAAI,KAAK,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,QAAQ,CAAC,QAAQ,CAAC,CAAC,CAAC,CAAC,EAAE,CAAC;gBAC5C,KAAK,IAAI,CAAC,CAAC;YACb,CAAC;QACH,CAAC;QAED,MAAM,MAAM,GAAG,KAAK,CAAC,MAAM,CAAC,CAAC,CAAC,OAAO,CAAC,KAAK,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC,SAAS,CAAC;QAChE,OAAO,EAAE,KAAK,EAAE,KAAK,EAAE,MAAM,EAAE,CAAC;IAClC,CAAC,CAAC,CAAC;IAEH,MAAM,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,KAAK,GAAG,CAAC,CAAC,KAAK,CAAC,CAAC;IACzC,OAAO,MAAM,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,KAAK,GAAG,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,EAAE,IAAI,CAAC,CAAC;AAC1D,CAAC"}