@maxgfr/codeindex 2.11.1 → 2.13.0

package/scripts/engine.mjs

@@ -14,9 +14,9 @@ var ENGINE_VERSION, SCHEMA_VERSION, EXTRACTOR_VERSION;
 var init_types = __esm({
   "src/types.ts"() {
     "use strict";
-    ENGINE_VERSION = "2.11.1";
+    ENGINE_VERSION = "2.13.0";
     SCHEMA_VERSION = 4;
-    EXTRACTOR_VERSION = 8;
+    EXTRACTOR_VERSION = 10;
   }
 });
 
@@ -320,6 +320,7 @@ function walk(root, opts = {}) {
   const maxFileBytes = opts.maxFileBytes ?? 1024 * 1024;
   const maxFiles = opts.maxFiles ?? DEFAULT_MAX_FILES;
   const useGitignore = opts.gitignore !== false;
+  const ignoreDirs = opts.ignoreDirs ? new Set(opts.ignoreDirs) : IGNORE_DIRS;
   const out2 = [];
   let capped = false;
   let excluded = 0;
@@ -368,7 +369,7 @@ function walk(root, opts = {}) {
         continue;
       }
       if (st.isDirectory()) {
-        if (IGNORE_DIRS.has(name2)) continue;
+        if (ignoreDirs.has(name2)) continue;
         if (isLink) continue;
         if (useGitignore && rules.length && isIgnored(rules, rel, true)) continue;
         stack.push({ dir: abs, rel, rules });
@@ -463,6 +464,7 @@ var init_walk = __esm({
       ".cache",
       "tmp",
       ".ultraindex",
+      ".codeindex",
       "Pods",
       "DerivedData",
       ".terraform",
@@ -737,8 +739,8 @@ function extractReexports(rel, content, localSymbols) {
   const out2 = [];
   const seen = /* @__PURE__ */ new Set();
   const lineAt = (idx) => content.slice(0, idx).split(/\r?\n/).length;
-  const localKindOf = /* @__PURE__ */ new Map();
-  for (const s of localSymbols) if (!localKindOf.has(s.name)) localKindOf.set(s.name, s.kind);
+  const localDeclOf = /* @__PURE__ */ new Map();
+  for (const s of localSymbols) if (!localDeclOf.has(s.name)) localDeclOf.set(s.name, s);
   const named = /export\s*\{([\s\S]*?)\}\s*(?:from\s*['"]([^'"]+)['"])?\s*;?/g;
   let m;
   while ((m = named.exec(content)) && out2.length < 60) {
@@ -750,12 +752,13 @@ function extractReexports(rel, content, localSymbols) {
       const name2 = as ? as[2] : p;
       if (!/^[A-Za-z_$][\w$]*$/.test(name2) || name2 === "default" || seen.has(name2)) continue;
       seen.add(name2);
-      const mirroredKind = !from ? localKindOf.get(orig) : void 0;
+      const decl = !from ? localDeclOf.get(orig) : void 0;
       out2.push({
         name: name2,
-        kind: mirroredKind ?? "reexport",
+        kind: decl?.kind ?? "reexport",
         file: rel,
-        line: lineAt(m.index),
+        line: decl ? decl.line : lineAt(m.index),
+        ...decl?.endLine !== void 0 ? { endLine: decl.endLine } : {},
         signature: from ? `export { ${name2} } from "${from}"` : `export { ${name2} }`,
         exported: true,
         lang
@@ -900,7 +903,7 @@ var init_js_ts = __esm({
     RULES = [
       { re: /^\s*export\s+(?:async\s+)?function\s+(?<name>[\w$]+)/, kind: "function", exported: true },
       { re: /^\s*export\s+default\s+(?:async\s+)?function\s+(?<name>[\w$]+)/, kind: "function", exported: true },
-      { re: /^\s*export\s+default\s+(?:abstract\s+)?class\s+(?<name>[\w$]+)/, kind: "class", exported: true },
+      { re: /^\s*export\s+default\s+(?:abstract\s+)?class\s+(?!extends\b)(?<name>[\w$]+)/, kind: "class", exported: true },
       { re: /^\s*(?:async\s+)?function\s+(?<name>[\w$]+)/, kind: "function", exported: false },
       { re: /^\s*export\s+(?:abstract\s+)?class\s+(?<name>[\w$]+)/, kind: "class", exported: true },
       { re: /^\s*(?:abstract\s+)?class\s+(?<name>[\w$]+)/, kind: "class", exported: false },
@@ -5717,7 +5720,7 @@ function readReceiver(node) {
   const name2 = obj ? readName(obj) : void 0;
   return name2 && /^[A-Za-z_]\w*$/.test(name2) ? name2 : void 0;
 }
-function collectCalls(root, spec) {
+function collectCalls(root, spec, maxCalls = MAX_CALLS) {
   if (!spec.calls) return [];
   const out2 = [];
   const seen = /* @__PURE__ */ new Set();
@@ -5748,7 +5751,7 @@ function collectCalls(root, spec) {
   };
   visit(root);
   out2.sort((a, b) => byStr(a.name, b.name) || a.line - b.line);
-  return out2.slice(0, MAX_CALLS);
+  return out2.slice(0, maxCalls);
 }
 function collectImportedNames(root, spec) {
   if (!spec.imports?.import_statement) return [];
@@ -5775,7 +5778,7 @@ function collectImportedNames(root, spec) {
   visit(root);
   return [...found].sort(byStr).slice(0, MAX_IMPORTED_NAMES);
 }
-function extractAst(rel, ext, content) {
+function extractAst(rel, ext, content, opts = {}) {
   const key = grammarKeyForExt(ext);
   if (!key || !grammarReady(key)) return void 0;
   const spec = SPECS[key];
@@ -5961,7 +5964,7 @@ function extractAst(rel, ext, content) {
     }
     const refs = collectImports(root, spec);
     const idents = collectRefIdents(root, new Set(symbols.map((s) => s.name)));
-    const calls = collectCalls(root, spec);
+    const calls = collectCalls(root, spec, opts.maxCalls);
     const importedNames = collectImportedNames(root, spec);
     let pkg;
     if (spec.lang === "java") {
@@ -6397,12 +6400,12 @@ function extractImports(ext, content) {
   }
   return [...specs].map((spec) => ({ kind: "import", spec }));
 }
-function collectCallsRegex(content, symbols = []) {
+function collectCallsRegex(content, symbols = [], maxCalls = 512) {
   const out2 = /* @__PURE__ */ new Map();
   const ownDefLines = new Set(symbols.map((s) => `${s.name} ${s.line}`));
   const lines = content.split("\n");
   const CALL_RE = /(?:\bnew\s+)?(?:([A-Za-z_$][\w$]*)\s*\.\s*)?([A-Za-z_$][\w$]*)\s*\(/g;
-  for (let i2 = 0; i2 < lines.length && out2.size < 512; i2++) {
+  for (let i2 = 0; i2 < lines.length && out2.size < maxCalls; i2++) {
     const line = lines[i2];
     const trimmed = line.trimStart();
     if (trimmed.startsWith("//") || trimmed.startsWith("#") || trimmed.startsWith("*")) continue;
@@ -6417,7 +6420,7 @@ function collectCallsRegex(content, symbols = []) {
     CALL_RE.lastIndex = 0;
     let m;
     const fallbackExcluded = /* @__PURE__ */ new Set();
-    while ((m = CALL_RE.exec(line)) !== null && out2.size < 512) {
+    while ((m = CALL_RE.exec(line)) !== null && out2.size < maxCalls) {
       const receiver = m[1];
       const name2 = m[2];
       if (name2.length < 2 || CALL_KEYWORDS.has(name2)) continue;
@@ -6434,8 +6437,8 @@ function collectCallsRegex(content, symbols = []) {
   }
   return [...out2.values()].sort((a, b) => a.name < b.name ? -1 : a.name > b.name ? 1 : a.line - b.line);
 }
-function extractCode(rel, ext, content) {
-  const ast = extractAst(rel, ext, content);
+function extractCode(rel, ext, content, opts = {}) {
+  const ast = extractAst(rel, ext, content, { maxCalls: opts.maxCallsPerFile });
   const symbols = (ast ? ast.symbols : extractSymbols(rel, ext, content)).slice(0, 400);
   const known = new Set(symbols.map((s) => s.name));
   const reexports = extractReexports(rel, content, symbols).filter((s) => !known.has(s.name));
@@ -6451,7 +6454,7 @@ function extractCode(rel, ext, content) {
     // collector otherwise, so caller indexes exist without the wasm sidecar.
     // `symbols` (this file's own regex-extracted defs) lets the collector
     // exclude a definition's own name+line from its call candidates.
-    calls: ast ? ast.calls : collectCallsRegex(content, symbols),
+    calls: ast ? ast.calls : collectCallsRegex(content, symbols, opts.maxCallsPerFile),
     importedNames: ast?.importedNames
   };
 }
@@ -6523,7 +6526,8 @@ function scanRepo(root, opts = {}) {
   const { files: walked, capped, excluded } = walk(root, {
     maxFileBytes: opts.maxBytes,
     maxFiles: opts.maxFiles,
-    gitignore: opts.gitignore
+    gitignore: opts.gitignore,
+    ignoreDirs: opts.ignoreDirs
   });
   const outPrefix = opts.out ? opts.out.replace(/\/+$/, "") + "/" : null;
   const files = [];
@@ -6572,7 +6576,7 @@ function scanRepo(root, opts = {}) {
       } else if (kind === "doc") {
         record.title = basename(f.rel);
       } else if (kind === "code") {
-        const code = extractCode(f.rel, f.ext, content);
+        const code = extractCode(f.rel, f.ext, content, { maxCallsPerFile: opts.maxCallsPerFile });
         record.title = basename(f.rel);
         record.summary = code.summary;
         record.symbols = code.symbols;
@@ -7731,8 +7735,11 @@ function buildCallerIndex(scan2, importPairs, opts = {}) {
 function enclosingSymbol(scan2, file, line) {
   const f = scan2.files.find((x) => x.rel === file);
   if (!f?.symbols.length) return void 0;
+  return enclosingAmong(f.symbols, line);
+}
+function enclosingAmong(symbols, line) {
   let best;
-  for (const s of f.symbols) {
+  for (const s of symbols) {
     if (REFERENCE_KINDS3.has(s.kind)) continue;
     if (s.line > line) continue;
     if (s.endLine !== void 0 && line > s.endLine) continue;
@@ -7742,6 +7749,29 @@ function enclosingSymbol(scan2, file, line) {
   }
   return best;
 }
+function buildRawCallerIndex(scan2) {
+  const byName = /* @__PURE__ */ new Map();
+  for (const f of scan2.files) {
+    if (!f.calls?.length) continue;
+    const symbols = f.symbols.filter((s) => !REFERENCE_KINDS3.has(s.kind));
+    for (const c2 of f.calls) {
+      const site = { file: f.rel, line: c2.line };
+      if (c2.receiver !== void 0) site.receiver = c2.receiver;
+      const enc = enclosingAmong(symbols, c2.line);
+      if (enc) site.enclosingSymbol = enc;
+      let arr = byName.get(c2.name);
+      if (!arr) byName.set(c2.name, arr = []);
+      arr.push(site);
+    }
+  }
+  const index = /* @__PURE__ */ new Map();
+  for (const name2 of [...byName.keys()].sort(byStr)) {
+    const sites = byName.get(name2);
+    sites.sort((a, b) => byStr(a.file, b.file) || a.line - b.line);
+    index.set(name2, sites);
+  }
+  return index;
+}
 var REFERENCE_KINDS3;
 var init_callers = __esm({
   "src/callers.ts"() {
@@ -9374,6 +9404,7 @@ var init_bm25 = __esm({
 });
 
 // src/embed/model.ts
+import { createHash as createHash2 } from "crypto";
 import { existsSync as existsSync3, readFileSync as readFileSync5 } from "fs";
 import { join as join10 } from "path";
 function resolveEmbedModelDir(repo) {
@@ -9390,16 +9421,12 @@ function resolveEmbedModelDir(repo) {
 function hasEmbedModel(repo) {
   return resolveEmbedModelDir(repo) !== void 0;
 }
-function loadEmbedModel(dir) {
-  if (!dir) return void 0;
-  const path = join10(dir, "model.json");
-  if (!existsSync3(path)) return void 0;
-  const raw = JSON.parse(readFileSync5(path, "utf8"));
-  const { modelId, dim, vocab, weights } = raw;
-  if (typeof modelId !== "string" || !modelId) throw new Error(`embed model: missing modelId in ${path}`);
-  if (!Number.isInteger(dim) || dim <= 0) throw new Error(`embed model: bad dim ${dim} in ${path}`);
+function parseEmbedModel(raw, source) {
+  const { modelId, dim, vocab, weights, unk: rawUnk } = raw ?? {};
+  if (typeof modelId !== "string" || !modelId) throw new Error(`embed model: missing modelId in ${source}`);
+  if (!Number.isInteger(dim) || dim <= 0) throw new Error(`embed model: bad dim ${dim} in ${source}`);
   if (!Array.isArray(vocab) || !Array.isArray(weights) || vocab.length !== weights.length) {
-    throw new Error(`embed model: vocab/weights length mismatch in ${path}`);
+    throw new Error(`embed model: vocab/weights length mismatch in ${source}`);
   }
   const vocabSize = vocab.length;
   const flat = new Float64Array(vocabSize * dim);
@@ -9414,20 +9441,42 @@ function loadEmbedModel(dir) {
     }
     for (let d = 0; d < dim; d++) flat[i2 * dim + d] = Number(row[d]);
   }
-  const unk = typeof raw.unk === "string" ? raw.unk : "[UNK]";
+  const unk = typeof rawUnk === "string" ? rawUnk : "[UNK]";
   const unkId = vmap.has(unk) ? vmap.get(unk) : -1;
   return { modelId, dim, unk, unkId, vocabSize, vocab: vmap, weights: flat };
 }
+function loadEmbedModel(dir) {
+  if (!dir) return void 0;
+  const path = join10(dir, "model.json");
+  if (!existsSync3(path)) return void 0;
+  const raw = JSON.parse(readFileSync5(path, "utf8"));
+  return parseEmbedModel(raw, path);
+}
 function resolveEmbedPullUrl() {
-  const url = process.env.CODEINDEX_EMBED_URL;
-  return url && url.trim() ? url.trim() : void 0;
+  const env = process.env.CODEINDEX_EMBED_URL;
+  if (env && env.trim()) return { url: env.trim() };
+  return { url: DEFAULT_EMBED_URL, sha256: EMBED_ASSET_SHA256 };
 }
-var EMBED_VERSION, DEFAULT_EMBED_DIRNAME;
+async function fetchEmbedModel(url, expectedSha256) {
+  const res = await fetch(url);
+  if (!res.ok) throw new Error(`HTTP ${res.status} from ${url}`);
+  const body2 = await res.text();
+  if (expectedSha256) {
+    const got = createHash2("sha256").update(body2).digest("hex");
+    if (got !== expectedSha256) {
+      throw new Error(`sha256 mismatch: expected ${expectedSha256}, got ${got}`);
+    }
+  }
+  return body2;
+}
+var EMBED_VERSION, DEFAULT_EMBED_DIRNAME, DEFAULT_EMBED_URL, EMBED_ASSET_SHA256;
 var init_model = __esm({
   "src/embed/model.ts"() {
     "use strict";
     EMBED_VERSION = 1;
     DEFAULT_EMBED_DIRNAME = "models";
+    DEFAULT_EMBED_URL = "https://github.com/maxgfr/codeindex/releases/download/embed-model-v1/model.json";
+    EMBED_ASSET_SHA256 = "163ad053eab4e9a80d421ed4164f32292c83290f02fbbe6fe4b9b1cd6ea18d34";
   }
 });
 
@@ -10129,8 +10178,13 @@ var init_viz = __esm({
 // src/mcp.ts
 var mcp_exports = {};
 __export(mcp_exports, {
-  runMcpServer: () => runMcpServer
+  memoizedEmbedModel: () => memoizedEmbedModel,
+  memoizedEmbeddingIndex: () => memoizedEmbeddingIndex,
+  runMcpServer: () => runMcpServer,
+  scanFingerprint: () => scanFingerprint
 });
+import { statSync as statSync4 } from "fs";
+import { join as join12 } from "path";
 import { createInterface } from "readline";
 function str(v) {
   return typeof v === "string" && v ? v : void 0;
@@ -10138,6 +10192,32 @@ function str(v) {
 function strArray(v) {
   return Array.isArray(v) && v.every((x) => typeof x === "string") && v.length ? v : void 0;
 }
+function errMessage(e) {
+  return e instanceof Error ? e.message : String(e);
+}
+function scanFingerprint(scan2) {
+  return sha1(scan2.files.map((f) => `${f.rel}:${f.hash}`).join("\n"));
+}
+async function memoizedEmbeddingIndex(key, build) {
+  const cacheKey = `${key.mode}:${key.identity}:${scanFingerprint(key.scan)}`;
+  if (embeddingIndexCache && embeddingIndexCache.key === cacheKey) return embeddingIndexCache.index;
+  const index = await build();
+  embeddingIndexCache = { key: cacheKey, index };
+  return index;
+}
+function memoizedEmbedModel(modelDir) {
+  let stat;
+  try {
+    stat = statSync4(join12(modelDir, "model.json"));
+  } catch {
+    return void 0;
+  }
+  const key = `${modelDir}:${stat.mtimeMs}:${stat.size}`;
+  if (embedModelCache && embedModelCache.key === key) return embedModelCache.model;
+  const model = loadEmbedModel(modelDir);
+  if (model) embedModelCache = { key, model };
+  return model;
+}
 async function callTool(name2, args2) {
   const repo = str(args2.repo);
   if (!repo) throw new Error("`repo` is required (absolute path to the repository root)");
@@ -10278,25 +10358,41 @@ async function callTool(name2, args2) {
       const endpoint = resolveEmbedEndpoint();
       if (endpoint) {
         try {
-          const index = await buildEndpointIndex(scan2);
+          const index = await memoizedEmbeddingIndex({ mode: "endpoint", identity: endpoint, scan: scan2 }, () => buildEndpointIndex(scan2));
           const queryVec = await encodeQueryViaEndpoint(query);
-          return JSON.stringify(searchSemantic(scan2, query, index, { queryVec, limit, fuzzy }), null, 2);
-        } catch {
-          return JSON.stringify(searchIndex(scan2, query, { limit, fuzzy }), null, 2);
+          const results2 = searchSemantic(scan2, query, index, { queryVec, limit, fuzzy });
+          return JSON.stringify({ results: results2, tier: "endpoint" }, null, 2);
+        } catch (e) {
+          const results2 = searchIndex(scan2, query, { limit, fuzzy });
+          return JSON.stringify(
+            { results: results2, tier: "lexical", degradedReason: `embedding endpoint failed: ${errMessage(e)}` },
+            null,
+            2
+          );
         }
       }
       const modelDir = resolveEmbedModelDir(repo);
-      const model = modelDir ? loadEmbedModel(modelDir) : void 0;
+      const model = modelDir ? memoizedEmbedModel(modelDir) : void 0;
       if (model) {
-        const index = buildEmbeddingIndex(scan2, model);
-        return JSON.stringify(searchSemantic(scan2, query, index, { model, limit, fuzzy }), null, 2);
+        const index = await memoizedEmbeddingIndex(
+          { mode: "static", identity: `${modelDir}#${model.modelId}`, scan: scan2 },
+          () => buildEmbeddingIndex(scan2, model)
+        );
+        const results2 = searchSemantic(scan2, query, index, { model, limit, fuzzy });
+        return JSON.stringify({ results: results2, tier: "static" }, null, 2);
       }
+      const results = searchIndex(scan2, query, { limit, fuzzy });
+      return JSON.stringify(
+        { results, tier: "lexical", degradedReason: "no embedding endpoint or static model configured \u2014 see embed_status" },
+        null,
+        2
+      );
     }
     return JSON.stringify(searchIndex(scan2, query, { limit, fuzzy }), null, 2);
   }
   if (name2 === "embed_status") {
     const modelDir = resolveEmbedModelDir(repo);
-    const model = modelDir ? loadEmbedModel(modelDir) : void 0;
+    const model = modelDir ? memoizedEmbedModel(modelDir) : void 0;
     const endpoint = resolveEmbedEndpoint();
     const mode = endpoint ? "endpoint" : model ? "static" : "none";
     const status = {
@@ -10371,7 +10467,7 @@ async function runMcpServer() {
     }
   }
 }
-var repoProp, scopeProps, TOOLS;
+var repoProp, scopeProps, TOOLS, embeddingIndexCache, embedModelCache;
 var init_mcp = __esm({
   "src/mcp.ts"() {
     "use strict";
@@ -10398,6 +10494,7 @@ var init_mcp = __esm({
     init_embed();
     init_search();
     init_endpoint();
+    init_hash();
     repoProp = { repo: { type: "string", description: "Absolute path to the repository root" } };
     scopeProps = {
       scope: { type: "string", description: "Restrict to one directory (repo-relative)" },
@@ -10610,7 +10707,7 @@ var init_mcp = __esm({
       },
       {
         name: "search",
-        description: 'Natural-language-ish lexical search: BM25 ranking (k1=1.2, b=0.75) over symbol names (camelCase/snake_case subtokens), file path segments, markdown headings and summary lines. NOT embeddings by default \u2014 deterministic, diacritic-folded, zero API keys. Answers "where is auth handled?"-style queries with ranked files, matched terms and top symbols. Query terms with zero document frequency get a deterministic trigram-fuzzy fallback (typo-tolerant) unless `fuzzy: false`. Set `semantic: true` to RRF-fuse the deterministic static-embedding tier when a model asset is present (degrades to lexical otherwise \u2014 see embed_status).',
+        description: 'Natural-language-ish lexical search: BM25 ranking (k1=1.2, b=0.75) over symbol names (camelCase/snake_case subtokens), file path segments, markdown headings and summary lines. NOT embeddings by default \u2014 deterministic, diacritic-folded, zero API keys. Answers "where is auth handled?"-style queries with ranked files, matched terms and top symbols. Query terms with zero document frequency get a deterministic trigram-fuzzy fallback (typo-tolerant) unless `fuzzy: false`. Set `semantic: true` to RRF-fuse an embedding tier (HTTP endpoint, else a local static model) with lexical \u2014 the response then wraps the ranked list as `{ results, tier, degradedReason? }`, `tier` being "endpoint"/"static" when fusion happened or "lexical" (with `degradedReason`) when it did not (see embed_status). Without `semantic`, the response is the bare ranked array, unchanged.',
         inputSchema: {
           type: "object",
           properties: {
@@ -10624,7 +10721,7 @@ var init_mcp = __esm({
             },
             semantic: {
               type: "boolean",
-              description: "RRF-fuse an embedding tier with lexical (default false). Precedence: the HTTP endpoint (CODEINDEX_EMBED_ENDPOINT) if set, else a local static model. Degrades silently to lexical-only when neither is available/reachable \u2014 see embed_status."
+              description: 'RRF-fuse an embedding tier with lexical (default false). Precedence: the HTTP endpoint (CODEINDEX_EMBED_ENDPOINT) if set, else a local static model. The response reports the effective tier as a top-level `tier` field ("endpoint"/"static" on success, "lexical" plus `degradedReason` when neither is available/reachable) instead of degrading silently \u2014 see embed_status.'
             }
           },
           required: ["repo", "query"]
@@ -11073,7 +11170,7 @@ init_pipeline();
 init_graph_json();
 init_symbols_json();
 import { existsSync as existsSync4, mkdirSync as mkdirSync2, readFileSync as readFileSync6, writeFileSync as writeFileSync3 } from "fs";
-import { join as join12, resolve } from "path";
+import { join as join13, resolve } from "path";
 init_scan();
 init_callers();
 init_workspaces();
@@ -11116,8 +11213,9 @@ Commands:
                 embed status   Effective mode (none/static/endpoint), model +
                                EMBED_VERSION, and endpoint reachability (JSON)
                 embed build    Write embeddings.bin into --out <dir> (static tier)
-                embed pull     Fetch the model asset into CODEINDEX_EMBED_DIR (or
-                               <repo>/.codeindex/models/) \u2014 needs CODEINDEX_EMBED_URL
+                embed pull     Fetch the official model asset into CODEINDEX_EMBED_DIR
+                               (or <repo>/.codeindex/models/); sha256-verified. Override
+                               the source with CODEINDEX_EMBED_URL
                 embed serve    Print (or --run) the docker command that starts the
                                containerized embedding server (rich tier)
   rules       Architecture rules (forbidden edges, cycles, orphans) validated
@@ -11140,8 +11238,11 @@ Flags:
   --exclude <glob>    Exclude matching paths (repeatable)
   --scope <dir>       Restrict to one directory (sugar for --include '<dir>/**')
   --no-gitignore      Do not honor .gitignore files (default: honored)
+  --ignore-dir <name> Directory names to skip (repeatable) \u2014 REPLACES the
+                      default ignored-directory set, never merges with it
   --max-files <n>     Cap walked files (default 20000)
   --max-bytes <n>     Skip files above this size (default 1 MiB)
+  --max-calls <n>     Per-file call-site cap for extraction (default 512)
   --no-ast            Skip tree-sitter grammars even when present (regex tier)
   --config <file>     Rules config for \`rules\` (JSON: [{name, from, to, \u2026}])
   --limit <n>         Max results for \`search\` (default 20)
@@ -11156,7 +11257,7 @@ Flags:
                       each site corroborated|unique-name
 `;
 function parseFlags(args2) {
-  const flags2 = { repo: process.cwd(), include: [], exclude: [], gitignore: true, noAst: false, fuzzy: true, semantic: false };
+  const flags2 = { repo: process.cwd(), include: [], exclude: [], gitignore: true, ignoreDirs: [], noAst: false, fuzzy: true, semantic: false };
   for (let i2 = 0; i2 < args2.length; i2++) {
     const a = args2[i2];
     const next = () => {
@@ -11179,8 +11280,10 @@ function parseFlags(args2) {
     else if (a === "--exclude") flags2.exclude.push(next());
     else if (a === "--scope") flags2.scope = next();
     else if (a === "--no-gitignore") flags2.gitignore = false;
+    else if (a === "--ignore-dir") flags2.ignoreDirs.push(next());
     else if (a === "--max-files") flags2.maxFiles = num();
     else if (a === "--max-bytes") flags2.maxBytes = num();
+    else if (a === "--max-calls") flags2.maxCalls = num();
     else if (a === "--ignore-case") flags2.ignoreCase = true;
     else if (a === "--max-hits") flags2.maxHits = num();
     else if (a === "--budget-tokens") flags2.budgetTokens = num();
@@ -11207,8 +11310,10 @@ function scanOptions(flags2) {
     exclude: flags2.exclude.length ? flags2.exclude : void 0,
     scope: flags2.scope,
     gitignore: flags2.gitignore,
+    ignoreDirs: flags2.ignoreDirs.length ? flags2.ignoreDirs : void 0,
     maxFiles: flags2.maxFiles,
-    maxBytes: flags2.maxBytes
+    maxBytes: flags2.maxBytes,
+    maxCallsPerFile: flags2.maxCalls
   };
 }
 async function runCli(argv) {
@@ -11233,7 +11338,7 @@ async function runCli(argv) {
     if (!flags2.out) throw new Error("index needs --out <dir>");
     const outDir = flags2.out;
     mkdirSync2(outDir, { recursive: true });
-    const cachePath = join12(outDir, "cache.json");
+    const cachePath = join13(outDir, "cache.json");
     let cache;
     try {
       const parsed = JSON.parse(readFileSync6(cachePath, "utf8"));
@@ -11243,8 +11348,8 @@ async function runCli(argv) {
     } catch {
     }
     const { scan: scan2, graph, symbols } = buildIndexArtifacts(flags2.repo, { ...scanOptions(flags2), cache, out: outDir });
-    writeFileSync3(join12(outDir, "graph.json"), renderGraphJson(graph));
-    writeFileSync3(join12(outDir, "symbols.json"), renderSymbolsJson(symbols));
+    writeFileSync3(join13(outDir, "graph.json"), renderGraphJson(graph));
+    writeFileSync3(join13(outDir, "symbols.json"), renderSymbolsJson(symbols));
     const files = {};
     for (const f of scan2.files) {
       const entry = { hash: f.hash, record: f, size: f.size };
@@ -11261,7 +11366,7 @@ async function runCli(argv) {
     const model = modelDir ? loadEmbedModel(modelDir) : void 0;
     if (model) {
       const index = buildEmbeddingIndex(scan2, model);
-      writeFileSync3(join12(outDir, "embeddings.bin"), serializeEmbeddings(index));
+      writeFileSync3(join13(outDir, "embeddings.bin"), serializeEmbeddings(index));
       embedNote = ` + embeddings.bin (${index.records.length} records, model ${model.modelId})`;
     }
     process.stderr.write(`codeindex: ${scan2.files.length} files \u2192 ${outDir}/graph.json + symbols.json${embedNote}${scan2.capped ? " (capped)" : ""}
@@ -11393,39 +11498,36 @@ async function runCli(argv) {
       mkdirSync2(flags2.out, { recursive: true });
       const scan2 = scanRepo(flags2.repo, scanOptions(flags2));
       const index = buildEmbeddingIndex(scan2, model);
-      writeFileSync3(join12(flags2.out, "embeddings.bin"), serializeEmbeddings(index));
+      writeFileSync3(join13(flags2.out, "embeddings.bin"), serializeEmbeddings(index));
       process.stderr.write(`codeindex: ${index.records.length} embedding records \u2192 ${flags2.out}/embeddings.bin (model ${model.modelId})
 `);
     } else if (sub === "pull") {
-      const url = resolveEmbedPullUrl();
-      if (!url) {
-        process.stderr.write(
-          "codeindex: no model URL configured. The official static-embedding asset is not published yet.\nSet CODEINDEX_EMBED_URL to a model.json URL (optionally CODEINDEX_EMBED_DIR as the destination), then re-run `codeindex embed pull`.\n"
-        );
-        process.exitCode = 1;
-        return;
-      }
-      const destDir = process.env.CODEINDEX_EMBED_DIR ?? join12(flags2.repo, ".codeindex", "models");
+      const { url, sha256 } = resolveEmbedPullUrl();
+      const destDir = process.env.CODEINDEX_EMBED_DIR ?? join13(flags2.repo, ".codeindex", "models");
       mkdirSync2(destDir, { recursive: true });
-      process.stderr.write(`codeindex: fetching model from ${url} \u2192 ${join12(destDir, "model.json")}
+      process.stderr.write(`codeindex: fetching model from ${url} \u2192 ${join13(destDir, "model.json")}
 `);
-      const res = await fetch(url);
-      if (!res.ok) {
-        process.stderr.write(`codeindex: pull failed \u2014 HTTP ${res.status} from ${url}
+      let body2;
+      try {
+        body2 = await fetchEmbedModel(url, sha256);
+      } catch (e) {
+        process.stderr.write(`codeindex: pull failed \u2014 ${e instanceof Error ? e.message : String(e)} (nothing written)
 `);
         process.exitCode = 1;
         return;
       }
-      const body2 = await res.text();
       try {
-        JSON.parse(body2);
-      } catch {
-        process.stderr.write("codeindex: pull failed \u2014 response is not a valid model.json (expected JSON)\n");
+        parseEmbedModel(JSON.parse(body2), url);
+      } catch (e) {
+        process.stderr.write(
+          `codeindex: pull failed \u2014 response is not a valid model.json (${e instanceof Error ? e.message : String(e)}) (nothing written)
+`
+        );
         process.exitCode = 1;
         return;
       }
-      writeFileSync3(join12(destDir, "model.json"), body2);
-      process.stderr.write(`codeindex: model written to ${join12(destDir, "model.json")}
+      writeFileSync3(join13(destDir, "model.json"), body2);
+      process.stderr.write(`codeindex: model written to ${join13(destDir, "model.json")}
 `);
     } else {
       throw new Error("embed needs a subcommand: status | build | pull | serve");
@@ -11508,6 +11610,7 @@ export {
   buildGraph,
   buildIndexArtifacts,
   buildModules,
+  buildRawCallerIndex,
   buildResolveContext,
   buildSymbolIndex,
   byKey,

package/src/ast/extract.ts

@@ -399,8 +399,9 @@ function readReceiver(node: TSNode | null): string | undefined {
 // dedicated member-call node's `name`; "constructor" reads the constructed type.
 // A qualified call also carries the immediate `receiver` name (see readReceiver).
 // Names are filtered to plausible identifiers (≥ 2 chars), deduped by name+line,
-// sorted, and capped, so the set stays small and deterministic.
-function collectCalls(root: TSNode, spec: LangSpec): { name: string; line: number; receiver?: string }[] {
+// sorted, and capped (default MAX_CALLS; overridable via maxCalls), so the set
+// stays small and deterministic.
+function collectCalls(root: TSNode, spec: LangSpec, maxCalls: number = MAX_CALLS): { name: string; line: number; receiver?: string }[] {
   if (!spec.calls) return [];
   const out: { name: string; line: number; receiver?: string }[] = [];
   const seen = new Set<string>();
@@ -439,7 +440,7 @@ function collectCalls(root: TSNode, spec: LangSpec): { name: string; line: numbe
   };
   visit(root);
   out.sort((a, b) => byStr(a.name, b.name) || a.line - b.line);
-  return out.slice(0, MAX_CALLS);
+  return out.slice(0, maxCalls);
 }
 
 // Collect JS/TS named-import bindings: `import { a, b as c } from "x"` →
@@ -476,7 +477,8 @@ function collectImportedNames(root: TSNode, spec: LangSpec): string[] {
 // Extract declared symbols from one file via its committed grammar. Returns
 // undefined when no grammar is loaded for the extension (caller falls back to the
 // regex extractor). Walks top-level declarations plus one level of nested members.
-export function extractAst(rel: string, ext: string, content: string): AstResult | undefined {
+// `opts.maxCalls` overrides the per-file call-site cap (default MAX_CALLS).
+export function extractAst(rel: string, ext: string, content: string, opts: { maxCalls?: number } = {}): AstResult | undefined {
   const key = grammarKeyForExt(ext);
   if (!key || !grammarReady(key)) return undefined;
   const spec = SPECS[key];
@@ -697,7 +699,7 @@ export function extractAst(rel: string, ext: string, content: string): AstResult
 
     const refs = collectImports(root, spec);
     const idents = collectRefIdents(root, new Set(symbols.map((s) => s.name)));
-    const calls = collectCalls(root, spec);
+    const calls = collectCalls(root, spec, opts.maxCalls);
     const importedNames = collectImportedNames(root, spec);
     let pkg: string | undefined;
     if (spec.lang === "java") {