Re: [PATCH v3 5/6] iio: pressure: Add timestamp and scan_masks for BMP280 driver

[Andy Shevchenko <andriy.shevchenko@linux.intel.com>]

On Wed, Mar 20, 2024 at 10:31:39PM +0100, Vasileios Amoiridis wrote:
> On Wed, Mar 20, 2024 at 10:38:03PM +0200, Andy Shevchenko wrote:
> > On Wed, Mar 20, 2024 at 07:45:16PM +0100, Vasileios Amoiridis wrote:
> > > On Wed, Mar 20, 2024 at 01:07:07PM +0200, Andy Shevchenko wrote:
> > > > On Tue, Mar 19, 2024 at 01:29:24AM +0100, Vasileios Amoiridis wrote:

...

> > > > > +enum bmp280_scan {
> > > > > +	BMP280_TEMP,
> > > > > +	BMP280_PRESS,
> > > > > +	BME280_HUMID
> > > > 
> > > > The last is not a terminator, please leave trailing comma.
> > > > 
> > > > > +};
> > > 
> > > What do you mean it is not a terminator? In general with the enum
> > > variables I would write:
> > > 
> > > 	enum var { a, b, c };
> > 
> > This example is different to what you used. I.o.w. _this_ example is okay.
> > 
> > > Why in this case there is a comma needed after the BME280_HUMID element?
> > 
> > It's pure style issue that helps to avoid the unneeded churn in the future in
> > case the list is getting expanded. You can easily imagine what I mean.
> > 
> 
> Ok, that definitely makes sense, thank you! In general, should this be applied
> to structs as well?

Yes, to structs and/or arrays initializers when the list has a potential
expanding. We don't have trailing comma when:
1) it's a terminator entry (nothing must be after);
2) it's on the one line (as in your above example).

-- 
With Best Regards,
Andy Shevchenko