Re: split: Allow splitting by line count (instead of byte size)

[Pádraig Brady]

On 12/01/2021 18:16, John wrote:
> * On Tuesday, January 12, 2021 07:31, "Pádraig Brady" said:
> > The disadvantage is that we'd be pulling some wc logic into split,
> > but it wouldn't be providing any efficiency advantages.
> > Achieving this in shell is also simple enough and portable
> >
> >     lines=$(($(wc -l < "file") / 2))
> >     split -l $lines file
>
>    You're not wrong that it's simple enough and portable (that's what I did 
> myself when I had to do the next data set in the project)...but it's also 
> equally applicable to bytes, so why have split count the number of bytes in the 
> input when another tool could do that?
>
> bytes=$(($(wc -c < "file") / 2))
> split -b $bytes file
>
>    I know this example is a little pedantic and/or facetious, and I'm not trying to be 
> snarky here (well, maybe a little snarky?  :)); the goal here is just to illustrate that 
> while I agree that "do one thing and do it well" is a very noble goal, 
> sometimes there's such a thing as being _too_ modular.

Well doing this for bytes is arguably more useful.
Also doing this for bytes is computationally inexpensive to implement.
I.E. --number=N doesn't hide much processing,
while --number=L/N would.
It's generally best to be explicit with expensive operations.

>    On the other hand, its not being worth the effort _is_ a valid argument in 
> my book; it's up to the people actually doing the work whether they think their 
> effort is justified.  I know _I_ would like it, but I assume I'm an edge case 
> since split has existed this long without this feature and I can only speak for 
> myself.

cheers,
Pádraig