I am learning C programming.
I met a problem about vim configuration.
The following is my vim setting written in .vimrc:
set nu
set shiftwidth=4
set tabstop=4
set softtabstop=8
set autoindent
set cindent
set smartindent
syntax on
In order to see where the tabs are, I have used :set list:
#include <stdio.h>$
int main(void)$
{$
^Iprintf("ab\n");$
^Ireturn 0;$
}$
Now in line 4 under Insert mode, I want to see how many tabs I can get after I strike Tab key in different locations.
Here's the result:
When I strike Tab after n, I get ^Iprin^I^Itf("ab\n");$.
When I strike Tab after \n, I get ^Iprintf("ab\n^I^I");$.
The above two situations are quite understandable to me.
But when I strike Tab after a, I get ^Iprintf("a^Ib\n");$.
This is beyond my understanding.
Can anybody explain why there is only one tab replaced?
I thought there should be two because softtabstop takes up 8 spaces which is two times tabstop (4).
This is the result of the interaction between your softtabstop and (regular) tabstop.
Your softtabstop says when hitting the Tab key, the results should be aligned with multiples of 8 columns. When you strike Tab after a, you're already at column 13, so just one 4-width tab character is enough to get there. In the other cases you were at 16 and 8, respectively, so it took two 4-width tab characters to align to your 8-width soft tab.
PS: if you want weird behaviours (not recommended of course), set tabstop=3 softtabstop=8. Then you get combinations of spaces and tab characters when you hit the Tab key.
PPS: All this is part of why I don't like literal Tab characters. They're unstable, their display depending on each developer's tab-stop settings. In my organization, I push hard for everyone to use :set expandtab (or the equivalent in each developer's editor) and use space characters for stable and reproducible display. The only place I tolerate tab characters in my work is in makefiles, where I don't have a choice.