fix(ALL): improve vim.str_utfindex and vim.str_byteindex usage

lua/mini/align.lua

@@ -2031,9 +2031,12 @@ if vim.fn.has('nvim-0.10') == 0 then
   end
 end
 
+H.str_utfindex = function(s, i) return vim.str_utfindex(s, 'utf-32', i) end
+if vim.fn.has('nvim-0.11') == 0 then H.str_utfindex = function(s, i) return (vim.str_utfindex(s, i)) end end
+
 H.str_utf_end = function(s, n) return n >= s:len() and 0 or vim.str_utf_end(s, n) end
 if vim.fn.has('nvim-0.10') == 0 then
-  H.str_utf_end = function(s, n) return n >= s:len() and 0 or (vim.str_byteindex(s, vim.str_utfindex(s, n)) - n) end
+  H.str_utf_end = function(s, n) return n >= s:len() and 0 or (vim.str_byteindex(s, H.str_utfindex(s, n)) - n) end
 end
 
 H.is_any_point_inside_any_span = function(points, spans)

lua/mini/completion.lua

@@ -1917,6 +1917,9 @@ H.fit_to_width = function(text, width)
   return t_width <= width and text or ('…' .. vim.fn.strcharpart(text, t_width - width + 1, width - 1))
 end
 
+H.str_byteindex = function(s, i) return vim.str_byteindex(s, 'utf-32', i) end
+if vim.fn.has('nvim-0.11') == 0 then H.str_byteindex = function(s, i) return vim.str_byteindex(s, i) end end
+
 -- Simulate splitting single line `l` like how it would look inside window with
 -- `wrap` and `linebreak` set to `true`
 H.wrap_line = function(l, width)
@@ -1928,7 +1931,7 @@ H.wrap_line = function(l, width)
     -- Simulate wrap by looking at breaking character from end of current break
     -- Use `pcall()` to handle complicated multibyte characters (like Chinese)
     -- for which even `strdisplaywidth()` seems to return incorrect values.
-    success, width_id = pcall(vim.str_byteindex, l, width)
+    success, width_id = pcall(H.str_byteindex, l, width)
 
     if success then
       local break_match = vim.fn.match(l:sub(1, width_id):reverse(), '[- \t.,;:!?]')

lua/mini/diff.lua

@@ -1488,12 +1488,15 @@ H.compute_worddiff_changed_parts = function(ref_line, buf_line)
   return ref_ranges, buf_ranges
 end
 
+H.str_utfindex = function(s, i) return vim.str_utfindex(s, 'utf-32', i) end
+if vim.fn.has('nvim-0.11') == 0 then H.str_utfindex = function(s, i) return (vim.str_utfindex(s, i)) end end
+
 H.slice_line = function(line)
   -- Intertwine every proper character with '\n'
   local line_len = line:len()
   local sliced, starts, ends
   -- Make short route for a very common case of no multibyte characters
-  if vim.str_utfindex(line) == line_len then
+  if H.str_utfindex(line) == line_len then
     sliced, starts, ends = line:gsub('(.)', '%1\n'), {}, {}
     for i = 1, string.len(line) do
       starts[i], ends[i] = i, i

lua/mini/icons.lua

@@ -2132,10 +2132,13 @@ H.get_impl = {
   os = function(name) return H.os_icons[name] end,
 }
 
+H.str_byteindex = function(s, i) return vim.str_byteindex(s, 'utf-32', i) end
+if vim.fn.has('nvim-0.11') == 0 then H.str_byteindex = function(s, i) return vim.str_byteindex(s, i) end end
+
 H.style_icon = function(glyph, name)
   if MiniIcons.config.style ~= 'ascii' then return glyph end
   -- Use `vim.str_byteindex()` and `vim.fn.toupper()` for multibyte characters
-  return vim.fn.toupper(name:sub(1, vim.str_byteindex(name, 1)))
+  return vim.fn.toupper(name:sub(1, H.str_byteindex(name, 1)))
 end
 
 H.filetype_match = function(filename)

lua/mini/jump2d.lua

@@ -478,8 +478,8 @@ MiniJump2d.gen_spotter.pattern = function(pattern, side)
 
       -- Unify how spot is chosen in case of multibyte characters
       -- Use `+-1` to make sure that result is at start of multibyte character
-      local utf_index = vim.str_utfindex(line, spot) - 1
-      spot = vim.str_byteindex(line, utf_index) + 1
+      local utf_index = H.str_utfindex(line, spot) - 1
+      spot = H.str_byteindex(line, utf_index) + 1
 
       -- Add spot only if it referces new actually visible column
       if spot ~= res[#res] then table.insert(res, spot) end
@@ -1226,4 +1226,10 @@ H.merge_unique = function(tbl_1, tbl_2)
   return res
 end
 
+H.str_utfindex = function(s, i) return vim.str_utfindex(s, 'utf-32', i) end
+if vim.fn.has('nvim-0.11') == 0 then H.str_utfindex = function(s, i) return (vim.str_utfindex(s, i)) end end
+
+H.str_byteindex = function(s, i) return vim.str_byteindex(s, 'utf-32', i) end
+if vim.fn.has('nvim-0.11') == 0 then H.str_byteindex = function(s, i) return vim.str_byteindex(s, i) end end
+
 return MiniJump2d

lua/mini/map.lua

@@ -1159,7 +1159,7 @@ H.mask_from_strings = function(strings, _)
     local mask_row = H.tbl_repeat(true, n_cols)
 
     -- Detect whitespace
-    s_ext:gsub('()%s', function(j) mask_row[vim.str_utfindex(s_ext, j)] = false end)
+    s_ext:gsub('()%s', function(j) mask_row[H.str_utfindex(s_ext, j)] = false end)
     res[i] = mask_row
   end
 
@@ -1679,11 +1679,10 @@ end
 
 H.set_extmark_safely = function(...) pcall(vim.api.nvim_buf_set_extmark, ...) end
 
-H.str_width = function(x)
-  -- Use first returned value (UTF-32 index, and not UTF-16 one)
-  local res = vim.str_utfindex(x)
-  return res
-end
+H.str_utfindex = function(s, i) return vim.str_utfindex(s, 'utf-32', i) end
+if vim.fn.has('nvim-0.11') == 0 then H.str_utfindex = function(s, i) return (vim.str_utfindex(s, i)) end end
+
+H.str_width = function(x) return H.str_utfindex(x) end
 
 H.tbl_repeat = function(x, n)
   local res = {}

lua/mini/operators.lua

@@ -1211,10 +1211,16 @@ H.get_mark = function(mark_name) return vim.api.nvim_buf_get_mark(0, mark_name)
 
 H.set_mark = function(mark_name, mark_data) vim.api.nvim_buf_set_mark(0, mark_name, mark_data[1], mark_data[2], {}) end
 
+H.str_utfindex = function(s, i) return vim.str_utfindex(s, 'utf-32', i) end
+if vim.fn.has('nvim-0.11') == 0 then H.str_utfindex = function(s, i) return (vim.str_utfindex(s, i)) end end
+
+H.str_byteindex = function(s, i) return vim.str_byteindex(s, 'utf-32', i) end
+if vim.fn.has('nvim-0.11') == 0 then H.str_byteindex = function(s, i) return vim.str_byteindex(s, i) end end
+
 H.get_next_char_bytecol = function(markcoords)
   local line = vim.fn.getline(markcoords[1])
-  local utf_index = vim.str_utfindex(line, math.min(line:len(), markcoords[2] + 1))
-  return vim.str_byteindex(line, utf_index)
+  local utf_index = H.str_utfindex(line, math.min(line:len(), markcoords[2] + 1))
+  return H.str_byteindex(line, utf_index)
 end
 
 -- Indent ---------------------------------------------------------------------

lua/mini/pick.lua

@@ -3662,10 +3662,16 @@ H.seq_along = function(arr)
   return res
 end
 
+H.str_utfindex = function(s, i) return vim.str_utfindex(s, 'utf-32', i) end
+if vim.fn.has('nvim-0.11') == 0 then H.str_utfindex = function(s, i) return (vim.str_utfindex(s, i)) end end
+
+H.str_byteindex = function(s, i) return vim.str_byteindex(s, 'utf-32', i) end
+if vim.fn.has('nvim-0.11') == 0 then H.str_byteindex = function(s, i) return vim.str_byteindex(s, i) end end
+
 H.get_next_char_bytecol = function(line_str, col)
   if type(line_str) ~= 'string' then return col end
-  local utf_index = vim.str_utfindex(line_str, math.min(line_str:len(), col))
-  return vim.str_byteindex(line_str, utf_index)
+  local utf_index = H.str_utfindex(line_str, math.min(line_str:len(), col))
+  return H.str_byteindex(line_str, utf_index)
 end
 
 H.is_file_text = function(path)

lua/mini/surround.lua

@@ -1877,6 +1877,12 @@ H.is_point_inside_spans = function(point, spans)
   return false
 end
 
+H.str_utfindex = function(s, i) return vim.str_utfindex(s, 'utf-32', i) end
+if vim.fn.has('nvim-0.11') == 0 then H.str_utfindex = function(s, i) return (vim.str_utfindex(s, i)) end end
+
+H.str_byteindex = function(s, i) return vim.str_byteindex(s, 'utf-32', i) end
+if vim.fn.has('nvim-0.11') == 0 then H.str_byteindex = function(s, i) return vim.str_byteindex(s, i) end end
+
 -- Work with operator marks ---------------------------------------------------
 H.get_marks_pos = function(mode)
   -- Region is inclusive on both ends
@@ -1922,10 +1928,10 @@ H.get_marks_pos = function(mode)
     -- Use `math.min()` because it might lead to 'index out of range' error
     -- when mark is positioned at the end of line (that extra space which is
     -- selected when selecting with `v$`)
-    local utf_index = vim.str_utfindex(line2, math.min(#line2, pos2[2]))
+    local utf_index = H.str_utfindex(line2, math.min(#line2, pos2[2]))
     -- This returns the last byte inside character because `vim.str_byteindex()`
     -- 'rounds upwards to the end of that sequence'.
-    pos2[2] = vim.str_byteindex(line2, utf_index)
+    pos2[2] = H.str_byteindex(line2, utf_index)
   end
 
   return {