fix(NODE-5578): improve objectId deserialisation by skipping buffer allocation

[billouboq]

Description

Improve ObjectId deserialisation by around 20% by skipping one buffer allocation.
Also add a benchmark that deserialise lot's of ObjectIds

Double check the following

• Ran npm run check:lint script
• Self-review completed using the steps outlined here
• PR title follows the correct format: type(NODE-xxxx)[!]: description
  • Example: feat(NODE-1234)!: rewriting everything in coffeescript
• Changes are covered by tests
• New TODOs have a related JIRA ticket

[W-A-James]

Hi @billouboq, thanks so much for this PR. Taking a look in the debugger, with this change, the ObjectId will not own the Uint8Array that is passed in. Because of this the buffer could be changed later on and mutate the value of the ObjectId, potentially leading to data corruption. So while this does improve performance, it breaks an underlying assumption we try to maintain with our BSON types.

[billouboq]

Yeah no problem @W-A-James , I can understand. Thanks for the heads up 👍