" Set the pattern for trailing horizontal whitespace to match and remove. The
" `[:space:]` character class suffices for almost everything in practice, but
" at the time of writing it still only includes ASCII characters. I'm writing
" this because I had a document with lines with a trailing NO-BREAK SPACE
" (U+00A0) which `[:space:]` doesn't catch, so we'll round out the collection
" by adding all the Unicode space characters, since that's easy to do.
"
" <https://jkorpela.fi/chars/spaces.html>
" archived: <http://web.archive.org/web/20230318105634/https://jkorpela.fi/chars/spaces.html>
"
" * U+0020: SPACE
" * U+00A0: NO-BREAK SPACE
" * U+1680: OGHAM SPACE MARK
" * U+180E: MONGOLIAN VOWEL SEPARATOR
" * U+2000: EN QUAD
" * U+2001: EM QUAD
" * U+2002: EN SPACE
" * U+2003: EM SPACE
" * U+2004: THREE-PER-EM SPACE
" * U+2005: FOUR-PER-EM SPACE
" * U+2006: SIX-PER-EM SPACE
" * U+2007: FIGURE SPACE
" * U+2008: PUNCTUATION SPACE
" * U+2009: THIN SPACE
" * U+200A: HAIR SPACE
" * U+200B: ZERO WIDTH SPACE
" * U+202F: NARROW NO-BREAK SPACE
" * U+205F: MEDIUM MATHEMATICAL SPACE
" * U+3000: IDEOGRAPHIC SPACE
" * U+FEFF: ZERO WIDTH NO-BREAK SPACE
"
let s:pattern
\ = '['
\ . '[:space:]'
\ . '\u0020'
\ . '\u00A0'
\ . '\u1680'
\ . '\u180E'
\ . '\u2000-\u200B'
\ . '\u202F'
\ . '\u205F'
\ . '\u3000'
\ . '\uFEFF'
\ . ']\+$'
" Wrapper function to strip both horizontal and vertical trailing whitespace,
" return the cursor to its previous position, and report changes
function! strip_trailing_whitespace#(start, end) abort
" Save cursor position
let pos = getpos('.')
" Whether we made changes
let changed = 0
" If we're going to the end, strip vertical space; we do this first so we
" don't end up reporting having trimmed lines that we deleted
if a:end == line('$')
let vertical = s:StripVertical()
let changed = changed || vertical > 0
endif
" Strip horizontal space
let horizontal = s:StripHorizontal(a:start, a:end)
let changed = changed || horizontal > 0
" Return the cursor
call setpos('.', pos)
" Report what changed
let msg = horizontal.' trimmed'
if exists('vertical')
let msg = msg.', '.vertical.' deleted'
endif
echomsg msg
" Return whether anything changed
return changed
endfunction
" Strip horizontal trailing whitespace, return the number of lines changed
function! s:StripHorizontal(start, end) abort
" Start a count of lines trimmed
let stripped = 0
" Iterate through buffer
let num = a:start
while num <= line('$') && num <= a:end
" If the line has trailing whitespace, strip it off and bump the count
let line = getline(num)
if line =~# s:pattern
call setline(num, substitute(line, s:pattern, '', ''))
let stripped = stripped + 1
endif
" Bump for next iteration
let num = num + 1
endwhile
" Return the number of lines trimmed
return stripped
endfunction
" Strip trailing vertical whitespace, return the number of lines changed
function! s:StripVertical() abort
" Store the number of the last line we found with non-whitespace characters
" on it; start at 1 because even if it's empty it's never trailing
let eof = 1
" Iterate through buffer
let num = 1
while num <= line('$')
" If the line has any non-whitespace characters in it, update our pointer
" to the end of the file text
let line = getline(num)
if line =~# '\S'
let eof = num
endif
" Bump for next iteration
let num = num + 1
endwhile
" Get the number of lines to delete; if there are any, build a range and
" remove them with :delete, suppressing its normal output (we'll do it)
let stripped = line('$') - eof
if stripped
let range = (eof + 1).',$'
silent execute range.'delete'
endif
" Return the number of lines deleted
return stripped
endfunction
