*Naming Conventions*

[Neil Cerutti]

Neil Cerutti a écrit :

I'm not sure this is really an "advantage" here.

Why not?

 

[Marius Gedminas]

Certainly i and j are just as generic, but they have the
advantage over 'item' of being more terse.

Python programmers usually prefer readability over terseness. Finding
a variable named 'i' that is not an integer index would be surprising
to me.

 

[Neil Cerutti]

Python programmers usually prefer readability over terseness.
Finding a variable named 'i' that is not an integer index would
be surprising to me.

Terseness and readability can go together in some circumstances,
though. But every case should be handled with taste and the
application experience, obviously. In a situation where 'i' might
be misleading I wouldn't use it.

 

[bruno.desthuilliers]

(snip)


Why not?

Because
1/ too much terseness (like too much verbosity) goes against
readability,
2/ 'i' and 'j' are canonical names for C-style for loops indices - and
Python's for(each) loops are very different...

A few special cases aside, one-letter identifiers are bad when it
comes to readability. As far as I'm concern, I stopped using 'j' even
in C code a long time ago - as soon as I have too distinct loop
indices, I can never remember what are 'i' and 'j' supposed to
represent, so I prefer to name them more explicitly (ie : 'row' and
'col').

Just make the test:


for (i=0; i < mi; i++)
for (j-0; j < mj; j++)
f(a[j]);


vs

for (row=0; row < nb_rows; row++) {
for (col=0; col < nb_cols; col++) {
init_item_at(grid[row][col]);
}
}