ByteBuffer should have hex bytes initialiser / write method

[weissi]

This is pretty much a companion of #2447 / #2475. The aforementioned bugs/patches introduce the ability to get the hex dump from a ByteBuffer which is awesome.

Next, I'd like to construct/write into a ByteBuffer from a hex dump:

• ByteBuffer(hexEncodedBytes: "68 65 6c 6c 6f 20 77 6f 72 6c 64 0a")
• buffer.writeHexEncodedBytes("68656c6c6f20776f726c640a")

I'd say it just needs to support the .plain format (i.e. 2 hex-digit values with whitespace ignored)

---

Super basic test:

var buffer = ByteBuffer(hexEncodedBytes: "68 65 6c 6c 6f 20 77 6f 72 6c 64 0a")
buffer.writeHexEncodedBytes("68656c6c6f20776f726c64")
buffer.writeHexEncodedBytes("")
buffer.writeHexEncodedBytes("0a")
XCTAssertEqual(ByteBuffer(string: "hello world\nhello world\n"), buffer)

another one:

var allBytes = ByteBuffer()
for x in UInt8.min ... UInt8.max {
    buffer.writeInteger(x)
}

let allBytesHex = allBytes.hexDump(format: .plain)
let allBytesDecoded = ByteBuffer(hexEncodedBytes: allBytesHex)

XCTAssertEqual(allBytes, allBytesDecoded)

[ali-ahsan-ali]

I have taken a look and would like to pick this up after #2734.

[weissi]

Of course this should also parse the .compat format suggested in #2825

[weissi]

@ali-ahsan-ali are you still interested in this?

[ali-ahsan-ali]

Happy if someone else is eager to give it a go. I have started looking at it but no real progress. It feels overly ambitious to try and start working on a second PR when my first has taken quite a while.

[weissi]

Happy if someone else is eager to give it a go. I have started looking at it but no real progress. It feels overly ambitious to try and start working on a second PR when my first has taken quite a while.

I did re-request review on your other PR. Remember to click the ♻️ (re-request) review once you addressed the changed suggested/requested. I'm sure @FranzBusch & @Lukasa just missed that you pushed updates to your existing PR.

Separately, if you're up for it, you can totally have multiple PRs open at the same time, shouldn't slow things down.

[ali-ahsan-ali]

The "good for newcomers" in this repo are the hardest thing ever 😅 (which I VERY much enjoy).

Experimenting with the below and seeing if writing that to the buffer is the right outcome for writing a hex encoded string to buffer.

extension String {
    var hexData: some Sequence<UInt8> {
        sequence(state: self, next: { remainder in
            guard remainder.count >= 2 else { return nil }
            let nextTwo = remainder.prefix(2)
            remainder.removeFirst(2)
            return UInt8(nextTwo, radix: 16) // compatible
        })
    }
}

TLDR: Thanks for the motivation/push. Working on it!

[weissi]

The "good for newcomers" in this repo are the hardest thing every 😅 (which I VERY much enjoy).

Experimenting with the below and seeing if writing that to the buffer is the right outcome for writing a hex encoded string to buffer.

extension String {
    var hexData: some Sequence<UInt8> {
        sequence(state: self, next: { remainder in
            guard remainder.count >= 2 else { return nil }
            let nextTwo = remainder.prefix(2)
            remainder.removeFirst(2)
            return UInt8(nextTwo, radix: 16) // compatible
        })
    }
}

TLDR: Thanks for the push. Working on it!

Awesome, FWIW, you can use

extension String {
   var hexData: some Sequence<UInt8> {
       return ByteBuffer(string: .self).hexDump(format: .plain).utf8
   }
}

to get a sequence of UInt8s. String's .utf8 gives you a UTF8View which is a Sequence<UInt8>

[ali-ahsan-ali]

@weissi please take a look at the PR when you find the time

[ali-ahsan-ali]

I think this can be closed now