I am looking for a g8 equivalent that prints the hex values of the character under the cursor according to the file encoding of the file loaded in a buffer rather than assuming it's utf-8.

Is there such a thing?

  • This doesn't seem possible as near as I can figure out... Even piping characters to external programs (xxd, hexdump) seems difficult, since many functions don't seem to be multibyte-aware enough (at least not with BIG-5)... Commented Mar 3, 2015 at 16:48

2 Answers 2


There's nothing built-in, but the following mapping seems to do the trick.

" g9                Print the hex values of the bytes used in the
"                   character under the cursor / selected text, using the
"                   actual 'fileencoding' of the buffer.
function! s:ShowHexFileEncodingCharacter( isSelection )
    let l:text = (a:isSelection ? ingo#selection#Get() : ingo#text#GetChar(getpos('.')[1:2]))
    if empty(l:text)
        echomsg 'NUL'

    let l:encoding = (empty(&l:fileencoding) ? &encoding : &l:fileencoding)
    let l:output = ingo#buffer#temp#Call(function('GetHexFileEncodingCharacters'), [l:encoding, l:text])
    let l:hex = substitute(l:output, '0a\%(00\)*$\|\%(00\)\+0a$', '', '') " Remove the encoding of the trailing LF.
    echomsg substitute(l:hex, '..\zs', ' ', 'g')
function! GetHexFileEncodingCharacters( fileencoding, text )
    let [l:save_fileencoding, l:save_fileformat] = [&l:fileencoding, &l:fileformat]
    let [&l:fileencoding, &l:fileformat] = [a:fileencoding, 'unix'] " Always use Unix fileformat to make the removal of the newline easier.
        silent call ingo#lines#PutWrapper(1, 'put', a:text)
        silent 1delete _
        %!xxd -ps -c 999
        let [&l:fileencoding, &l:fileformat] = [l:save_fileencoding, l:save_fileformat]
nnoremap <silent> g9 :<C-u>call <SID>ShowHexFileEncodingCharacter(0)<CR>
xnoremap <silent> g9 :<C-u>call <SID>ShowHexFileEncodingCharacter(1)<CR>

It uses the :! command to filter the character through the external xxd command (which ships with Vim and performs a hex dump). This uses some functions of my ingo-library plugin.


For completeness, note that ga uses the encoding (not fileencoding) when it is a non-unicode encoding. You should be aware of the other consequences of setting encoding to a value other than utf-8, though.

The encoding setting is actually the internal encoding used in vim (almost always utf-8 these days - setting it to a different unicode encoding is interpreted as utf-8 as well) - it's used for buffer contents, variables, options, etc. Having it as something other than UTF-8 limits what characters you can have any of those places, and changing it while vim is running will change the meaning of any values (including buffer contents) that are already loaded with non-ASCII text. This also explains why ga and g8 normally always return unicode and UTF-8 values - because the buffer is in encoding (= utf-8) regardless of what encoding the file was in.

  • What would those consequences be? Dire?
    – muru
    Commented Mar 5, 2015 at 19:58
  • encoding is the internal encoding used for text strings everywhere in vim, so you cannot have any variables set that include characters not in the character set, load any files with characters it doesn't support, etc. Changing it mid-stream changes the meaning of any existing non-ascii values (including the contents of file buffers).
    – Random832
    Commented Mar 5, 2015 at 20:00
  • Interesting! Could you add that to the answer as well?
    – muru
    Commented Mar 5, 2015 at 20:01

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.