refactor(indent): refactor computing of a string's indent size (#27252)

The `get_indent_str_vtab()` function currently calls `tabstop_padding()`
every time a tab is encountered (unless tabstops aren't used).
`tabstop_padding()` either does a division by 'tabstop' If 'vartabstop'
is not set, or iterates through the 'vartabstop' list to find current
tab width. 

Since the virtual column only increases, we can keep track of where the
next tabstop would be, and update this information once it was reached.

`get_indent_str_vtab()` also depends on 'listchars' "tab" value from the
current window, even though it may be called for a line from the same
buffer in a different window. In most cases, it is called with tabstops
enabled (last argument was `false`), so I split the function into one
that uses tabstops and the other that doesn't.

I removed `get_indent_str()` since I couldn't find any calls to it.

src/nvim/change.c

@@ -1149,9 +1149,7 @@ bool open_line(int dir, int flags, int second_line_indent, bool *did_do_comment)
   // indent to use for the new line.
   if (curbuf->b_p_ai || do_si) {
     // count white space on current line
-    newindent = get_indent_str_vtab(saved_line,
-                                    curbuf->b_p_ts,
-                                    curbuf->b_p_vts_array, false);
+    newindent = indent_size_ts(saved_line, curbuf->b_p_ts, curbuf->b_p_vts_array);
     if (newindent == 0 && !(flags & OPENLINE_COM_LIST)) {
       newindent = second_line_indent;  // for ^^D command in insert mode
     }
@@ -1593,9 +1591,7 @@ bool open_line(int dir, int flags, int second_line_indent, bool *did_do_comment)
 
         // Recompute the indent, it may have changed.
         if (curbuf->b_p_ai || do_si) {
-          newindent = get_indent_str_vtab(leader,
-                                          curbuf->b_p_ts,
-                                          curbuf->b_p_vts_array, false);
+          newindent = indent_size_ts(leader, curbuf->b_p_ts, curbuf->b_p_vts_array);
         }
 
         // Add the indent offset

src/nvim/indent.c

@@ -118,6 +118,7 @@ bool tabstop_set(char *var, colnr_T **array)
 /// If "vts" is set then the tab widths are taken from that array,
 /// otherwise the value of ts is used.
 int tabstop_padding(colnr_T col, OptInt ts_arg, const colnr_T *vts)
+  FUNC_ATTR_PURE
 {
   OptInt ts = ts_arg == 0 ? 8 : ts_arg;
   colnr_T tabcol = 0;
@@ -349,83 +350,96 @@ int get_sts_value(void)
   return result;
 }
 
-// Count the size (in window cells) of the indent in the current line.
+/// Count the size (in window cells) of the indent in the current line.
 int get_indent(void)
 {
-  return get_indent_str_vtab(get_cursor_line_ptr(),
-                             curbuf->b_p_ts,
-                             curbuf->b_p_vts_array,
-                             false);
+  return indent_size_ts(get_cursor_line_ptr(), curbuf->b_p_ts, curbuf->b_p_vts_array);
 }
 
-// Count the size (in window cells) of the indent in line "lnum".
+/// Count the size (in window cells) of the indent in line "lnum".
 int get_indent_lnum(linenr_T lnum)
 {
-  return get_indent_str_vtab(ml_get(lnum),
-                             curbuf->b_p_ts,
-                             curbuf->b_p_vts_array,
-                             false);
+  return indent_size_ts(ml_get(lnum), curbuf->b_p_ts, curbuf->b_p_vts_array);
 }
 
-// Count the size (in window cells) of the indent in line "lnum" of buffer
-// "buf".
+/// Count the size (in window cells) of the indent in line "lnum" of buffer "buf".
 int get_indent_buf(buf_T *buf, linenr_T lnum)
 {
-  return get_indent_str_vtab(ml_get_buf(buf, lnum), buf->b_p_ts, buf->b_p_vts_array, false);
+  return indent_size_ts(ml_get_buf(buf, lnum), buf->b_p_ts, buf->b_p_vts_array);
 }
 
-/// Count the size (in window cells) of the indent in line "ptr", with
-/// 'tabstop' at "ts".
-/// If @param list is true, count only screen size for tabs.
-int get_indent_str(const char *ptr, int ts, bool list)
-  FUNC_ATTR_NONNULL_ALL
+/// Compute the size of the indent (in window cells) in line "ptr",
+/// without tabstops (count tab as ^I or <09>).
+int indent_size_no_ts(char const *ptr)
+  FUNC_ATTR_NONNULL_ALL FUNC_ATTR_PURE
 {
-  int count = 0;
+  int tab_size = byte2cells(TAB);
 
-  for (; *ptr; ptr++) {
-    // Count a tab for what it is worth.
-    if (*ptr == TAB) {
-      if (!list || curwin->w_p_lcs_chars.tab1) {
-        // count a tab for what it is worth
-        count += ts - (count % ts);
-      } else {
-        // In list mode, when tab is not set, count screen char width
-        // for Tab, displays: ^I
-        count += ptr2cells(ptr);
-      }
-    } else if (*ptr == ' ') {
-      // Count a space for one.
-      count++;
+  int vcol = 0;
+  while (true) {
+    char const c = *ptr++;
+    if (c == ' ') {
+      vcol++;
+    } else if (c == TAB) {
+      vcol += tab_size;
     } else {
-      break;
+      return vcol;
     }
   }
-  return count;
 }
 
-/// Count the size (in window cells) of the indent in line "ptr", using
-/// variable tabstops.
-/// if "list" is true, count only screen size for tabs.
-int get_indent_str_vtab(const char *ptr, OptInt ts, colnr_T *vts, bool list)
+/// Compute the size of the indent (in window cells) in line "ptr",
+/// using tabstops
+int indent_size_ts(char const *ptr, OptInt ts, colnr_T *vts)
+  FUNC_ATTR_NONNULL_ARG(1) FUNC_ATTR_PURE
 {
-  int count = 0;
+  assert(char2cells(' ') == 1);
 
-  for (; *ptr; ptr++) {
-    if (*ptr == TAB) {  // count a tab for what it is worth
-      if (!list || curwin->w_p_lcs_chars.tab1) {
-        count += tabstop_padding(count, ts, vts);
-      } else {
-        // In list mode, when tab is not set, count screen char width
-        // for Tab, displays: ^I
-        count += ptr2cells(ptr);
-      }
-    } else if (*ptr == ' ') {
-      count++;  // count a space for one
+  int vcol = 0;
+  int tabstop_width, next_tab_vcol;
+
+  if (vts == NULL || vts[0] < 1) {  // tab has fixed width
+    // can ts be 0 ? This is from tabstop_padding().
+    tabstop_width = (int)(ts == 0 ? 8 : ts);
+    next_tab_vcol = tabstop_width;
+  } else {  // tab has variable width
+    colnr_T *cur_tabstop = vts + 1;
+    colnr_T *const last_tabstop = vts + vts[0];
+
+    while (cur_tabstop != last_tabstop) {
+      int cur_vcol = vcol;
+      vcol += *cur_tabstop++;
+      assert(cur_vcol < vcol);
+
+      do {
+        char const c = *ptr++;
+        if (c == ' ') {
+          cur_vcol++;
+        } else if (c == TAB) {
+          break;
+        } else {
+          return cur_vcol;
+        }
+      } while (cur_vcol != vcol);
+    }
+
+    tabstop_width = *last_tabstop;
+    next_tab_vcol = vcol + tabstop_width;
+  }
+
+  assert(tabstop_width != 0);
+  while (true) {
+    char const c = *ptr++;
+    if (c == ' ') {
+      vcol++;
+      next_tab_vcol += (vcol == next_tab_vcol) ? tabstop_width : 0;
+    } else if (c == TAB) {
+      vcol = next_tab_vcol;
+      next_tab_vcol += tabstop_width;
     } else {
-      break;
+      return vcol;
     }
   }
-  return count;
 }
 
 /// Set the indent of the current line.
@@ -828,10 +842,12 @@ int get_breakindent_win(win_T *wp, char *line)
     prev_tick = buf_get_changedtick(wp->w_buffer);
     prev_vts = wp->w_buffer->b_p_vts_array;
     if (wp->w_briopt_vcol == 0) {
-      prev_indent = get_indent_str_vtab(line,
-                                        wp->w_buffer->b_p_ts,
-                                        wp->w_buffer->b_p_vts_array,
-                                        wp->w_p_list);
+      if (wp->w_p_list && !wp->w_p_lcs_chars.tab1) {
+        prev_indent = indent_size_no_ts(line);
+      } else {
+        prev_indent = indent_size_ts(line, wp->w_buffer->b_p_ts,
+                                     wp->w_buffer->b_p_vts_array);
+      }
     }
     prev_listopt = wp->w_briopt_list;
     prev_list = 0;

test/unit/indent_spec.lua

@@ -1,6 +1,8 @@
 local helpers = require('test.unit.helpers')(after_each)
 local itp = helpers.gen_itp(it)
 
+local to_cstr = helpers.to_cstr
+local ffi = helpers.ffi
 local eq = helpers.eq
 
 local indent = helpers.cimport('./src/nvim/indent.h')
@@ -28,3 +30,31 @@ describe('get_sts_value', function()
     eq(tabstop, indent.get_sts_value())
   end)
 end)
+
+describe('indent_size_ts()', function()
+  itp('works for spaces', function()
+    local line = to_cstr((' '):rep(7) .. 'a ')
+    eq(7, indent.indent_size_ts(line, 100, nil))
+  end)
+
+  itp('works for tabs and spaces', function()
+    local line = to_cstr('   \t  \t \t\t   a ')
+    eq(19, indent.indent_size_ts(line, 4, nil))
+  end)
+
+  itp('works for tabs and spaces with empty vts', function()
+    local vts = ffi.new('int[1]') -- zero initialized => first element (size) == 0
+    local line = to_cstr('   \t  \t \t\t       a ')
+    eq(23, indent.indent_size_ts(line, 4, vts))
+  end)
+
+  itp('works for tabs and spaces with vts', function()
+    local vts = ffi.new('int[3]')
+    vts[0] = 2 -- zero indexed
+    vts[1] = 7
+    vts[2] = 2
+
+    local line = to_cstr('      \t  \t \t\t   a ')
+    eq(18, indent.indent_size_ts(line, 4, vts))
+  end)
+end)