test(runner): Test the runner and bench with a comment.

packages/big-endian/bench/decode.bench.js

@@ -1,33 +1,35 @@
-import { bench, describe } from "vitest"
-import decode from "../src/index.js"
+import { bench, describe } from "vitest";
+import decode from "../src/index.js";
 
-const SIZE_512x512 = 512 * 512
+const SIZE_512x512 = 512 * 512;
+
+// REMOVE THIS COMMENT. IT IS JUST BEING ADDED SO THE BENCH CAN RUN.
 
 function makeBuffer(byteLen) {
-  const data = new Uint8Array(byteLen)
-  for (let i = 0; i < byteLen; i++) data[i] = (i * 37) & 0xff
-  return data
+  const data = new Uint8Array(byteLen);
+  for (let i = 0; i < byteLen; i++) data[i] = (i * 37) & 0xff;
+  return data;
 }
 
 describe("big-endian decode (byte-swap)", () => {
-  const data16 = makeBuffer(SIZE_512x512 * 2)
+  const data16 = makeBuffer(SIZE_512x512 * 2);
 
   bench("16-bit unsigned + swap, 512x512", () => {
-    decode({ bitsAllocated: 16, pixelRepresentation: 0 }, data16)
-  })
+    decode({ bitsAllocated: 16, pixelRepresentation: 0 }, data16);
+  });
 
   bench("16-bit signed + swap, 512x512", () => {
-    decode({ bitsAllocated: 16, pixelRepresentation: 1 }, data16)
-  })
+    decode({ bitsAllocated: 16, pixelRepresentation: 1 }, data16);
+  });
 
   // The 16-bit benches above do real per-pixel swap work and are left
   // unbatched. The 8-bit path is a bare property assignment, so a single
   // call is swamped by fixed harness overhead and cache-model variation
   // across runner CPUs; batch x100 so the body is measurable.
-  const data8 = makeBuffer(SIZE_512x512)
+  const data8 = makeBuffer(SIZE_512x512);
   bench("8-bit passthrough, 512x512 x100", () => {
     for (let i = 0; i < 100; i++) {
-      decode({ bitsAllocated: 8 }, data8)
+      decode({ bitsAllocated: 8 }, data8);
     }
-  })
-})
+  });
+});