fix: Fix ast-grep runner when there are special characters

.changeset/pretty-monkeys-watch.md

@@ -0,0 +1,5 @@
+---
+"codemod": patch
+---
+
+Make ast-grep runner unicode-aware

apps/cli/package.json

@@ -92,7 +92,6 @@
     "@ast-grep/cli": "catalog:",
     "@ast-grep/napi": "catalog:",
     "@codemod.com/workflow": "workspace:*",
-    "@nodejs-loaders/alias": "^1.1.1",
     "@octokit/rest": "catalog:",
     "blessed": "catalog:",
     "esbuild": "^0.23.0",

packages/runner/src/engines/ast-grep.ts

@@ -128,25 +128,45 @@ export const runAstGrepCodemod = async (
       ? `powershell -Command "${astCommandBase}"`
       : astCommandBase;
 
-  const { stdout } = await execPromise(astCommand);
+  const { stdout, stderr } = await execPromise(astCommand);
   const matches = JSON.parse(stdout.trim()) as AstGrepCompactOutput[];
   // Sort in reverse order to not mess up replacement offsets
   matches.sort((a, b) => b.range.byteOffset.start - a.range.byteOffset.start);
 
-  let newData = oldData;
+  const encoder = new TextEncoder();
+  const decoder = new TextDecoder();
+  const oldDataBytes = encoder.encode(oldData);
+
+  let newDataBytes = oldDataBytes;
+
   for (const result of matches) {
     const { replacementOffsets, replacement } = result;
     if (!replacementOffsets) {
       continue;
     }
+    const replacementBytes = encoder.encode(replacement || "");
+
+    const startOffset = replacementOffsets.start;
+    const endOffset = replacementOffsets.end;
+    const newSize =
+      newDataBytes.length - (endOffset - startOffset) + replacementBytes.length;
+    const updatedBytes = new Uint8Array(newSize);
+
+    updatedBytes.set(newDataBytes.subarray(0, startOffset));
 
-    newData =
-      newData.slice(0, replacementOffsets.start) +
-      replacement +
-      newData.slice(replacementOffsets.end);
+    updatedBytes.set(replacementBytes, startOffset);
+
+    updatedBytes.set(
+      newDataBytes.subarray(endOffset),
+      startOffset + replacementBytes.length,
+    );
+
+    newDataBytes = updatedBytes;
   }
 
-  if (typeof newData !== "string" || oldData === newData) {
+  const newData = decoder.decode(newDataBytes);
+
+  if (oldData === newData) {
     return commands;
   }
 

packages/workflow/src/astGrep/astGrep.ts

@@ -38,7 +38,6 @@ const fileExtensionToLang: Record<string, Lang> = {
   cpp: Lang.Cpp,
   hpp: Lang.Cpp,
   cs: Lang.CSharp,
-  dart: Lang.Dart,
   ex: Lang.Elixir,
   exs: Lang.Elixir,
   go: Lang.Go,
@@ -153,6 +152,7 @@ const runExternalAstGrepRule = async (rule: any, filepath: string) => {
     path.dirname(require.resolve("@ast-grep/cli/package.json")),
     binaries.sg,
   );
+  console.log(binaryPath);
   await execFile(binaryPath, [
     "scan",
     "--update-all",