@takara-ai/miru-code 0.4.4 → 0.4.5

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@takara-ai/miru-code",
-  "version": "0.4.4",
+  "version": "0.4.5",
   "description": "Miru (見る) — hybrid code search for agents. Bun, TypeScript, Takara embeddings.",
   "type": "module",
   "license": "MIT",

package/src/cache.ts

@@ -1,6 +1,7 @@
 import { createHash } from "node:crypto";
 import { rm } from "node:fs/promises";
-import { join, resolve } from "node:path";
+import { join } from "node:path";
+import { computeSourceCacheKey } from "./utils.ts";
 import { resolveEmbeddingDimensions, resolveEmbeddingModel } from "./embeddings/openai.ts";
 import {
   loadBm25,
@@ -30,8 +31,8 @@ export function resolveCacheFolder(): string {
   return base;
 }
 
-export function findIndexCachePath(path: string): string {
-  const normalized = resolve(path);
+export function findIndexCachePath(path: string, ref?: string | null): string {
+  const normalized = computeSourceCacheKey(path, ref);
   const subdir = createHash("sha256").update(normalized, "utf-8").digest("hex");
   return join(resolveCacheFolder(), subdir, "index");
 }

package/src/chunking/structural.ts

@@ -5,7 +5,16 @@ import {
   splitLinesKeepEnds,
 } from "./lines.ts";
 
-const SUPPORTED_STRUCTURAL_LANGUAGES = new Set(["python", "go", "typescript", "javascript"]);
+const SUPPORTED_STRUCTURAL_LANGUAGES = new Set([
+  "python",
+  "go",
+  "typescript",
+  "javascript",
+  "cpp",
+  "c",
+]);
+
+const BRACE_FAMILY_LANGUAGES = new Set(["cpp", "c"]);
 
 export function chunkStructural(
   source: string,
@@ -164,9 +173,27 @@ function declPattern(language: string): RegExp {
   if (language === "go") {
     return /^\s*(func\b|type\b.*\b(struct|interface)\b)/;
   }
+  if (language === "cpp") {
+    return /^\s*(?:(?:template\s*<[^>]*>\s*)?(?:class|struct|namespace|enum(?:\s+class)?)\b|(?!(?:if|for|while|switch|catch|return|sizeof|static_assert|else|do)\b)(?:[\w:~<>,*&]+\s+)+\w+\s*\([^;{}]*\)\s*(?:const)?\s*(?:override)?(?:\s*=\s*0)?)/;
+  }
+  if (language === "c") {
+    return /^\s*(?:struct|enum|union|typedef)\b|(?!(?:if|for|while|switch|return|sizeof|else|do)\b)(?:[\w:*\s]+\s+)+\w+\s*\([^;{}]*\)/;
+  }
   return /^\s*(export\s+)?(async\s+function\b|function\b|class\b|interface\b|type\b|(?:const|let|var)\s+[A-Za-z_$][\w$]*\s*=\s*(?:async\s*)?(?:\([^)]*\)|[A-Za-z_$][\w$]*)(?:\s*:\s*[^=]+)?\s*=>)/;
 }
 
+function isBraceFamilyTypeDecl(text: string, language: string): boolean {
+  if (language === "cpp") {
+    return /^\s*(?:(?:template\s*<[^>]*>\s*)?(?:class|struct|namespace|enum(?:\s+class)?)\b)/.test(
+      text,
+    );
+  }
+  if (language === "c") {
+    return /^\s*(?:struct|enum|union)\b/.test(text);
+  }
+  return false;
+}
+
 function braceUnits(lines: LineGroup[], language: string): ChunkBoundary[] {
   const units: ChunkBoundary[] = [];
   const full = lines.map((l) => l.text).join("");
@@ -178,6 +205,13 @@ function braceUnits(lines: LineGroup[], language: string): ChunkBoundary[] {
     if (!re.test(text)) {
       continue;
     }
+    if (
+      BRACE_FAMILY_LANGUAGES.has(language) &&
+      lineIndent(text) > 0 &&
+      !isBraceFamilyTypeDecl(text, language)
+    ) {
+      continue;
+    }
     if (firstDeclStart === null) {
       firstDeclStart = lines[i]?.start;
     }

package/src/cli.ts

@@ -1,5 +1,4 @@
 #!/usr/bin/env bun
-import { resolve } from "node:path";
 import { type AgentId, writeAgentFile } from "./agents.ts";
 import { clearCache } from "./cache.ts";
 import {
@@ -27,7 +26,7 @@ import { MiruIndex } from "./miru-index.ts";
 import { ensureCredentials, runClearCredentials, runSetup } from "./setup.ts";
 import { withSpinner } from "./spinner.ts";
 import type { ContentType, SearchResult } from "./types.ts";
-import { formatResults, resolveChunk, resolveContent } from "./utils.ts";
+import { formatResults, resolveChunk, resolveContent, resolveSearchPath } from "./utils.ts";
 
 loadEnvFiles();
 normalizeTakaraApiKeyEnv();
@@ -283,7 +282,7 @@ async function runCli(argv: string[]): Promise<void> {
   }
 
   if (command === "clear") {
-    const path = resolve(rest[0] ?? process.cwd());
+    const path = resolveSearchPath(rest[0] ?? process.cwd());
     await runClear(path);
     return;
   }
@@ -298,7 +297,7 @@ async function runCli(argv: string[]): Promise<void> {
       printCommandHelp("search");
       process.exit(1);
     }
-    const path = resolve(sizedRest[1] ?? process.cwd());
+    const path = resolveSearchPath(sizedRest[1] ?? process.cwd());
     await runSearch(path, query, topK, content, jsonFlag);
     return;
   }
@@ -311,7 +310,7 @@ async function runCli(argv: string[]): Promise<void> {
       process.exit(1);
     }
     const line = Number(lineRaw);
-    const path = resolve(sizedRest[2] ?? process.cwd());
+    const path = resolveSearchPath(sizedRest[2] ?? process.cwd());
     try {
       await runFindRelated(path, filePath, line, topK, content, jsonFlag);
     } catch (err) {

package/src/index/files.ts

@@ -21,11 +21,14 @@ const EXTENSION_TO_LANGUAGE: Record<string, string> = {
   ".kt": "kotlin",
   ".kts": "kotlin",
   ".c": "c",
-  ".h": "c",
+  ".h": "cpp",
   ".cpp": "cpp",
   ".cc": "cpp",
   ".cxx": "cpp",
   ".hpp": "cpp",
+  ".hh": "cpp",
+  ".hxx": "cpp",
+  ".h++": "cpp",
   ".cs": "csharp",
   ".rb": "ruby",
   ".php": "php",

package/src/utils.ts

@@ -62,3 +62,8 @@ export function computeSourceCacheKey(source: string, ref?: string | null): stri
   }
   return resolve(source);
 }
+
+/** Resolve a local search path; leave git URLs unchanged for remote indexing. */
+export function resolveSearchPath(path: string): string {
+  return isGitUrl(path) ? path : resolve(path);
+}