Help me think: Is there something smart I can do to fix this? <https://github.com/Ashampoo/kim/blob...

# getting-started

[Stefan Oltmann]

Help me think: Is there something smart I can do to fix this?

https://github.com/Ashampoo/kim/blob/95aba62b09c628d25780fd3818c3d065d628df3d/src/[…]in/kotlin/com/ashampoo/kim/format/isobmff/ISOBMFFImageParser.kt

At the moment I'm thinking about something along the lines if there is some kind of ByteArray that's just a "view" on the real ByteArray - so that the bytes don't have to be doubled. Is there such a construct?

Not happy with the fix to make a copy of every byte read just in case I might need it later (and often won't). That was just my first idea how to solve this to get the code working, but it's certainly not a good solution.

What would you suggest?

[hho]

A Java ByteBuffer is kind of a view on a backing byte array, and it supports `mark()`/`reset()`. Not sure if it solves your problem though.

[Stefan Oltmann]

I'm wondering if there is something lightweight.

Something like copyOfRange() that would not copy the bytes in heap, but point to the same bytes.

[Stefan Oltmann]

I'm working all multiplatform. ByteBuffer is not available to me, but from reading the docs I'm not sure if it would help me.

A reset() function would mean I need to read the bytes twice. This is also not ideal.

[Youssef Shoaib [MOD]]

Sounds like you need either okio or kotlinx-io, both provide the same Buffer APIs.

[Stefan Oltmann]

How would I utilize these to help in this situation?

[Youssef Shoaib [MOD]]

https://square.github.io/okio/3.x/okio/okio/okio/-buffer/copy.html
This "copy" shares all the underlying byte arrays, it only creates a wrapper basically. There might be something more clever that can be done, but this at least should create a performance difference immediately

[Stefan Oltmann]

Ok, that's the right direction I think. Thanks.