Added state_hash property to devices - fixes #86

[jath03]

No description provided.

[felipecrs]

Hm... thank you!

But I think this is not more optimized than what I am currently doing.

However, you made me realize that I could simplify the implementation by hashing the entire data attribute.

Something like:

def _compute_state_hash(self) -> int:
        """Compute a hash of device states."""
        state_data = ""
        for device in self.coordinator.client.devices:
            state_data += str(device.data)
        return hash(state_data)

Also, it's completely fine if you disagree this being a good idea to be added to openrgb-python.

[felipecrs]

Oh, worth noting that we may get something like this from OpenRGB itself:

https://gitlab.com/CalcProgrammer1/OpenRGB/-/issues/5178#note_2870690945

[jath03]

Maybe eventually, but sounds like the recommendation is to do this ourselves for now. I wouldn't mind having this in the library.

I wasn't really expecting to get any performance differences, and updating devices is already a relatively slow operation, so I doubt this would make any significant impact.

The main benefit to having this in the library is that it gets updated/computed only when devices updates are received from openrgb. So you're not computing hashes all the time whenever you want to check the state.

[jath03]

One thing to note is that python's hash is not stable across processes, if that matters to you.

[jath03]

Eh, I'm reconsidering the decision to hash on every update. Adding extra cycles when most people will not need it is unnecessary. Sounds like you have a solution that will work for you, so I'm closing this.