don't change colours on blank lines

[landfillbaby]

builds on @obfusk's changes. fixes #24

[obfusk]

The variable names are a bit inconsistent now: g_blankLine vs g_changeEmpty.
I originally only skipped "empty" lines. But now the code also handles "blank" (i.e. empty or only whitespace) lines.
I suggest renaming variables and command line options to use "blank".

[landfillbaby]

how's this?

[obfusk]

LGTM. I think it would be better to switch Author/Co-authored-by though.

You could maybe add \f and \v as blank characters.
Full-width spaces (　) might be a bit difficult to detect with the current code.
Not sure if there are other blank characters...

(<shameless plug>my python version skips all of these 😅</shameless plug>)

[landfillbaby]

done

[obfusk]

Should

		if (c == '\n' || c == '\f' || c == '\v') {
[...]
		(c != ' ' && c != '\t' && c != '\r'))) {

not be

		if (c == '\n') {
[...]
		(c != ' ' && c != '\t' && c != '\r' && c != '\f' && c != '\v'))) {

instead?

[obfusk]

How about this?

#include <cctype>
[...]
else if (g_blankLine && (g_changeBlank || !isspace(c))) {

[landfillbaby]

i was thinking that since \f and \v both do effectively cause a new line, (at least they do on Android Termux, my PC is out of commission,) they should be treated as such, but if you think they should be ignored like the horizontal whitespace we can change it
and yes maybe isspace would be good to use

[obfusk]

I think \f and \v should normally not be found in the input anyway (I don't recall ever using them), and -- since different terminals don't always handle them identically -- only treating \n as newline is best. And isspace also makes the code a bit cleaner :) But I think there is probably no "correct answer" for all situations.

[obfusk]

Throw in some \bs and all bets are off :)

[landfillbaby]

ok yes. \n only, and !isspace(c) done

[landfillbaby]

closing in favour of #33