Any tool to convert tabs with 2 spaces to 4 spaces?

Question

Is there any tool I can use to convert tabs with 2 spaces to 4 spaces?

It'll be great if it also can change code formatting like positions of { } and strip unwanted spaces (eg. > 2 empty lines)

Answer 1

There's a devious sequence of operations in vim and relatives that will convert from '2 space' indenting to '4 space' indenting.

First, set tabstops to 2, shiftwidth to 2, and ensure expandtab is off:

:set ts=2 sw=2 noet

Now, from the top of the file, shift the file one unit right, then one unit left. This converts the blanks into tabs:

:1
>G<G

Verify that you have tabs where you expect them.

Now set tabstops to 4, shiftwidth to 4, and expand tab back on:

:set ts=4 sw=4 et

And repeat the shifting:

:1
>G<G

This avoids reformatting anything except the leading space on each line. You could encode it all in a single (complex) mapped instruction:

:map v :set ts=2 sw=2 noet^V^M:1^V^M>G<G:set ts=4 sw=4 et^V^M>G<G

If you are formatting C code, or C++, you may be better of with one of the specialized formatting programs - the classic one is (GNU) indent, but there are many others, including astyle. These can be controlled to deal with the indentation too - so you wouldn't have to go through the machinations.

Trailing space is trivial:

:g/[ ^I][ ^I]*$/s///

However, I do it so often (other people's code) that I have a script or program called stb (strip trailing blanks) to do the job. It works well for me; it is not strictly necessary, but is much less typing than the (trivial?) regex above. The script version is a sed script:

sed 's/[ ^I][ ^I]*$//' "$@"

Answer 2

I believe what you are looking for is astyle. I'm fairly sure that others exist, but that is the one I have seen used most for what you are trying to do. It also integrates well with a number of IDEs, in the off chance that your IDE does not support such things in and of itself.

Answer 3

If you want to convert 2 spaces to 4 spaces, you can use a sed trick like this. The pattern matches all the leading white space, and then doubles it.

$ sed 's/^\(\s*\)/\1\1/g' test.txt