vimdiff: Jump to next difference inside line?

Question

vimdiff is very handy for comparing files. However, I often use it on files with long lines and relatively few differences inside the lines.

vimdiff will correctly highlight differences inside a line (whole line pink, differing characters red). In these cases, it would be nice to be able to jump to the next difference inside the line.

You can jump to the "next difference" (]c), but this will jump to the next line with a difference.

Is there a way to go to the next different character inside the current line?

Answer 1

I see two solutions:

1. you would have to test the current syntax highlighting to jump to the red part in the line.
2. you would have to extract the current line in both buffers and find the first character that differs to position correctly the cursor

Both solutions need to be executed after the ]c, and require vim scripting.

EDIT: Here is a first draft that seems to work:

nnoremap <expr> <silent> <F3>   (&diff ? "]c:call \<sid>NextDiff()\<cr>" : ":cn\<cr>")

function! s:GotoWinline(w_l)
  normal! H
  while winline() < a:w_l
    normal! j
  endwhile
  " todo: beware of cases where the window is too little
endfunction

" Better ]c, [c jump
function! s:NextDiff()
  if ! &diffopt =~ 'filler' | return | endif

  let ignore_blanks = &diffopt =~ 'iwhite'

  " Assert: called just after a ]c or a [c
  " Forces the cursos to be synchronized in all synced windows
  " let diff_l = line()
  try 
    let foldenable = &foldenable
    set nofoldenable

    let w_l = winline() " problematic with enabled lines (from diff...)
    " echomsg w_l.'|'.line('.').'|'.getline('.')

    let lines = {}
    windo if &diff | call <sid>GotoWinline(w_l) | let lines[winnr()]={'text':getline('.'), 'number':line('.')} | endif
  finally
    let &foldenable = foldenable
  endtry

  " echomsg string(lines)
  if len(lines) < 2 | return | endif

  let indices = repeat([0], len(lines))
  let tLines = values(lines)
  let found = 0
  " infinite loop on two empty texts...
  while ! found
    let c = ''
    let next_idx = []
    let i = 0
    while i != len(indices)
      let crt_line = tLines[i].text
      let n = indices[i]
      if len(crt_line) == n
    let found = 1
    break
      endif

      let c2 = (len(crt_line) == n) ? 'EOL' : crt_line[n]
      if empty(c) 
    let c = c2
      endif

      " checks match
      let n += 1
      if c =~ '\s'
    if (c2 != c) && (ignore_blanks && c2 !~ '\s')
      let found = 1
      break
    else " advance
      while ignore_blanks && (n == len(crt_line) || crt_line[n] =~ '\s')
        let n += 1
      endwhile
    endif
      else
    if c2 != c
      let found = 1
      break
    endif
      endif
      let next_idx += [n]

      let i += 1
    endwhile
    if found | break | endif

    let indices = next_idx
  endwhile

  " now goto the right column
  let windows = keys(lines)
  " Assert len(windows) == len(indices)
  let w = 0
  while w != len(windows)
    " echomsg 'W#'.windows[w].' -> :'(tLines[w].number).'normal! '.(indices[w]+1).'|'
    exe windows[w].'wincmd w'
    silent! exe (tLines[w].number).'normal! 0'.(indices[w]).'l'
    let w += 1
  endwhile
  " echomsg string(indices)
endfunction

Answer 2

This is an easy workaround:

You can use set wrap.

This will create problems if the difference causes text to wrap in unequal number of lines.

Answer 3

I can't figure out how to do this with vimdiff either, but you might check out wdiff instead. It shows differences between two files one word at a time.

I had to compile it from source:

curl http://ftp.gnu.org/gnu/wdiff/wdiff-1.2.1.tar.gz > wdiff-1.2.1.tar.gz
tar -xzvf wdiff-1.2.1.tar.gz
cd wdiff-1.2.1
./configure
make
make install

Answer 4

Judging by the documentation, it can't be done.