charclass() sometimes returns 2 for punctuation

[kennypete]

Steps to reproduce

1. In a new buffer, source:

for c in range(32, 48, 8)
  echo nr2char(c) .. ' charclass=' .. nr2char(c)->charclass()
endfor

Echos (correctly):

  charclass=0
( charclass=1
0 charclass=2  

2. In a help buffer, set ma to make it modifiable, then put the same for loop and source it:

Echos:

  charclass=0
( charclass=2
0 charclass=2  

Screenshots to illustrate:

Expected behaviour

charclass() should always return 1 for the left parenthesis (and any punctuation character):

builtin.txt help:

charclass({string})					charclass()
		Return the character class of the first character in {string}.
		The character class is one of:
			0	blank
			1	punctuation
			2	word character
			3	emoji
			other	specific Unicode class

The same thing happens for all punctuation characters, returning 2 instead of 1.

Version of Vim

9.1.90

Environment

Same result in Win11 gvim, Debian WSL Huge GTK3 GUI.

Also happens in Neovim 0.10.2, incidentally, just to see whether the same thing happens in it, and it does (not that I care particularly).

Logs and stack traces

No response

[chrisbra]

That is most likely because Vims helpfiles set custom 'iskeyword' settings.

[kennypete]

That is most likely because Vims helpfiles set custom 'iskeyword' settings.

Okay, thanks. Since it's an internal Vim thing (and not, e.g., returning a value of a Unicode field), the only thing needed is to cover that distinction in the help? I'm in the process of updating a couple of help files, so will add that caveat.

The 'bug' label should be removed and 'enhancement' added? (i.e., enhancing the documentation)

It's kind of ironic, if you provide a working example of using charclass() in help file itself, that some/common punctuation characters return 2, though, aye 😉:

vim9script
var pop: list<string>
var eg: list<string> = [' ', '՛', '!', 'a', '♥', '가']
for i in range(0, len(eg) - 1)
  add(pop, eg[i] .. ': ' .. eg[i]->charclass())
endfor
popup_notification(pop, {title: "charclass()", time: 4000})