feat: methods for reading temp as a u16 value

[antonio-hickey]

This library has an emphasis on platform agnostic, but I ran into a LOT of problems trying to work with the f32 values returned from the methods in this library when running on bare metal.

I know using u16 is a little less accurate than f32, but it's nice to be able to call a method that just works out of the box when running on bare metal.

This PR introduces 3 new methods for Mlx90614 and 2 new methods for Mlx90615 which are all just clones of existing temperature methods but persevering their raw u16 value type.

I'm not 100% sold on the method names, and definitely open to suggestions.

Also willing to write additional docs, tests, and examples for these :) , if you think this feature will be merged.

Fixes #3

[antonio-hickey]

@eldruin

hey was hoping you could give me a hint here, my tests are failing do to ChecksumMismatch thrown from Self::check_pec:

    pub(crate) fn read_u16(&mut self, register: u8) -> Result<u16, Error<E>> {
        let mut data = [0; 3];
        self.i2c
            .write_read(self.address, &[register], &mut data)
            .map_err(Error::I2C)?;
        let pec = data[2];
        Self::check_pec(
            &[
                self.address << 1,
                register,
                (self.address << 1) + 1,
                data[0],
                data[1],
            ],
            pec,
        )?;
        Ok(u16::from(data[0]) | (u16::from(data[1]) << 8))
    }

I noticed this works when I actually use the method in code, but fails test.

[eldruin]

You need to calculate the PEC for your transfer and put that in the test as well. In this case it is the third byte. See for example here

[antonio-hickey]

Hey @eldruin got test's passing, never heard of PEC before so that was fun to learn about.

Let me know if there's any changes or more doc's / tests you want to see for this feature. :)

[eldruin]

I have published this in version 0.2.1.