fix(utils): normalized ptree.bytes() return type to Uint8Array

Bun's `Response(stream).bytes()` returns the raw `ArrayBuffer` once the
body arrives in more than one chunk, which happens for subprocess stdout
past ~128 KB. The public contract of `ptree.ChildProcess.bytes()` is
`Promise<Uint8Array>`, and callers — most visibly the `ssh://` read
path's `decodeUtf8Text` — rely on `Uint8Array` methods such as `.indexOf`
and `.subarray`. On larger remote text files this surfaced as:

  TypeError: bytes.indexOf is not a function

Normalize the result at the boundary: when `Response.bytes()` hands back
an `ArrayBuffer`, wrap it in a zero-copy `Uint8Array` view before
returning. Adds a regression test that drives a 256 KB stdout payload
through `ptree.spawn(...).bytes()` and asserts the contract.

Fixes #3712
This commit is contained in:
roboomp
2026-06-28 10:04:53 +00:00
parent 4373dde48a
commit 7d6a5c6de0
3 changed files with 37 additions and 1 deletions
+4
View File
@@ -2,6 +2,10 @@
## [Unreleased]
### Fixed
- Normalized `ptree.ChildProcess.bytes()` to always return a `Uint8Array`. Bun's `Response(stream).bytes()` returns the raw `ArrayBuffer` once the body arrives in more than one chunk (subprocess stdout past ~128 KB), which broke callers relying on `Uint8Array` methods like `.indexOf` and `.subarray` — including the `ssh://` read path's `decodeUtf8Text`, which crashed with `indexOf is not a function` on larger remote text files. ([#3712](https://github.com/can1357/oh-my-pi/issues/3712))
## [16.2.0] - 2026-06-27
### Added
+6 -1
View File
@@ -247,7 +247,12 @@ export class ChildProcess<In extends InMask = InMask> {
}
async bytes(): Promise<Uint8Array> {
return new Response(this.stdout).bytes();
// Bun's `Response(stream).bytes()` returns the raw `ArrayBuffer` once the
// stream emits more than one chunk (subprocess stdout chunks past ~128 KB).
// Normalize at the contract boundary so every caller — SSH read,
// `decodeUtf8Text`, callers slicing with `.subarray` — sees a `Uint8Array`.
const body = (await new Response(this.stdout).bytes()) as Uint8Array | ArrayBuffer;
return body instanceof Uint8Array ? body : new Uint8Array(body);
}
// ── Wait ─────────────────────────────────────────────────────────────
+27
View File
@@ -0,0 +1,27 @@
import { describe, expect, it } from "bun:test";
import { spawn } from "@oh-my-pi/pi-utils/ptree";
describe("ptree.ChildProcess.bytes()", () => {
// Regression for https://github.com/can1357/oh-my-pi/issues/3712:
// `Response(stream).bytes()` returns the raw `ArrayBuffer` once the body
// arrives in more than one chunk (which happens for subprocess stdout past
// ~128 KB). Downstream code — e.g. the SSH read path's `decodeUtf8Text` —
// relied on `Uint8Array` methods (`.indexOf`, `.subarray`) and crashed.
it("returns a Uint8Array regardless of stdout size", async () => {
// 256 KB is comfortably past the multi-chunk boundary observed on Bun
// 1.3.x; the test then asserts only on the contract, not on the exact
// chunk threshold, so it stays robust to future Bun runtime changes.
const size = 256 * 1024;
const child = spawn(["bun", "-e", `process.stdout.write("a".repeat(${size}))`]);
const bytes = await child.bytes();
await child.exitedCleanly;
expect(bytes).toBeInstanceOf(Uint8Array);
expect(bytes.length).toBe(size);
// The two methods the SSH read path depends on.
expect(typeof bytes.indexOf).toBe("function");
expect(typeof bytes.subarray).toBe("function");
expect(bytes.indexOf(0)).toBe(-1);
expect(bytes.subarray(0, 4)).toEqual(new Uint8Array([0x61, 0x61, 0x61, 0x61]));
});
});