perf: avoid cloning args on every pre/post

[vkarpov15]

Cloning args on every pre/post was a big drain on performance, especially for pre() where we don't have to add callbacks anymore.

Before:

{
  "results": [
    {
      "scenario": "1 hook, 1 arg",
      "iterations": 200000,
      "totalMs": 49.221,
      "avgUsPerExec": 0.246
    },
    {
      "scenario": "5 hooks, 1 arg",
      "iterations": 200000,
      "totalMs": 136.745,
      "avgUsPerExec": 0.684
    },
    {
      "scenario": "10 hooks, 1 arg",
      "iterations": 200000,
      "totalMs": 243.199,
      "avgUsPerExec": 1.216
    },
    {
      "scenario": "5 hooks, 3 args",
      "iterations": 200000,
      "totalMs": 137.098,
      "avgUsPerExec": 0.685
    },
    {
      "scenario": "5 hooks, 3 args + callback",
      "iterations": 200000,
      "totalMs": 135.913,
      "avgUsPerExec": 0.68
    },
    {
      "scenario": "10 hooks, 3 args + callback",
      "iterations": 200000,
      "totalMs": 244.986,
      "avgUsPerExec": 1.225
    }
  ]
}

After:

{
  "results": [
    {
      "scenario": "1 hook, 1 arg",
      "iterations": 200000,
      "totalMs": 34.19,
      "avgUsPerExec": 0.171
    },
    {
      "scenario": "5 hooks, 1 arg",
      "iterations": 200000,
      "totalMs": 45.294,
      "avgUsPerExec": 0.226
    },
    {
      "scenario": "10 hooks, 1 arg",
      "iterations": 200000,
      "totalMs": 64.934,
      "avgUsPerExec": 0.325
    },
    {
      "scenario": "5 hooks, 3 args",
      "iterations": 200000,
      "totalMs": 41.757,
      "avgUsPerExec": 0.209
    },
    {
      "scenario": "5 hooks, 3 args + callback",
      "iterations": 200000,
      "totalMs": 42.56,
      "avgUsPerExec": 0.213
    },
    {
      "scenario": "10 hooks, 3 args + callback",
      "iterations": 200000,
      "totalMs": 61.592,
      "avgUsPerExec": 0.308
    }
  ]
}

With post, the difference is less pronounced because we still have lingering callbacks. But still non-trival, before:

{
  "results": [
    {
      "scenario": "1 hook, 1 arg",
      "iterations": 200000,
      "totalMs": 38.381,
      "avgUsPerExec": 0.192
    },
    {
      "scenario": "5 hooks, 1 arg",
      "iterations": 200000,
      "totalMs": 98.818,
      "avgUsPerExec": 0.494
    },
    {
      "scenario": "10 hooks, 1 arg",
      "iterations": 200000,
      "totalMs": 156.182,
      "avgUsPerExec": 0.781
    },
    {
      "scenario": "5 hooks, 3 args",
      "iterations": 200000,
      "totalMs": 88.307,
      "avgUsPerExec": 0.442
    },
    {
      "scenario": "5 hooks, 3 args + callback",
      "iterations": 200000,
      "totalMs": 98.874,
      "avgUsPerExec": 0.494
    },
    {
      "scenario": "10 hooks, 3 args + callback",
      "iterations": 200000,
      "totalMs": 158.878,
      "avgUsPerExec": 0.794
    }
  ]
}

After:

{
  "results": [
    {
      "scenario": "1 hook, 1 arg",
      "iterations": 200000,
      "totalMs": 45.687,
      "avgUsPerExec": 0.228
    },
    {
      "scenario": "5 hooks, 1 arg",
      "iterations": 200000,
      "totalMs": 81.51,
      "avgUsPerExec": 0.408
    },
    {
      "scenario": "10 hooks, 1 arg",
      "iterations": 200000,
      "totalMs": 128.164,
      "avgUsPerExec": 0.641
    },
    {
      "scenario": "5 hooks, 3 args",
      "iterations": 200000,
      "totalMs": 75.607,
      "avgUsPerExec": 0.378
    },
    {
      "scenario": "5 hooks, 3 args + callback",
      "iterations": 200000,
      "totalMs": 77.089,
      "avgUsPerExec": 0.385
    },
    {
      "scenario": "10 hooks, 3 args + callback",
      "iterations": 200000,
      "totalMs": 121.383,
      "avgUsPerExec": 0.607
    }
  ]
}

This difference is significant enough to show up in Mongoose's saveSimple() benchmark, which we have been striving to optimize.

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 4407e9e035

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[chatgpt-codex-connector]

Rebuild errorArgs after overwrite in error-handler flow

When an error-handling post middleware throws/callbacks Kareem.overwriteResult, this branch updates args and newArgs but then continues without refreshing errorArgs. That means the next error handler is called with stale pre-overwrite arguments, whereas the previous implementation recomputed args per middleware iteration. This can corrupt chained error-handling behavior whenever one handler overwrites the result and a later handler depends on the updated args.

Useful? React with 👍 / 👎.

[Copilot]

Pull request overview

Optimizes Kareem hook execution performance by reducing per-hook argument cloning in execPre() / execPost(), and adds benchmark scripts to measure the impact.

Changes:

• Removes per-pre-hook argument cloning in execPre() and calls each pre hook with the current args array.
• Refactors execPost() to reuse a single newArgs array and callback wrapper across post hooks, rebuilding only when needed (e.g. overwriteResult).
• Adds benchmarks/execPre.js and benchmarks/execPost.js to measure pre/post hook execution cost.

Reviewed changes

Copilot reviewed 3 out of 3 changed files in this pull request and generated 4 comments.

File	Description
index.js	Removes per-hook arg cloning and refactors post-hook arg/callback handling for performance.
benchmarks/execPre.js	Adds a benchmark runner for execPre() scenarios.
benchmarks/execPost.js	Adds a benchmark runner for execPost() scenarios.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[vkarpov15]

@hasezoey @AbdelrahmanHafez can you please take a look?

[hasezoey]

Looks good to me, though consider resolving copilot's suggestions..