Token lists from \cs_split_function are not splittable by \seq_set_split
latex3 add tag
joulev 
MWE:

```
\documentclass{article}
\usepackage{expl3}
\begin{document}
\ExplSyntaxOn
\tl_set:Nx \l_tmpa_tl
  {
    \exp_last_unbraced:Nf \use_i:nnn \cs_split_function:N \one_two_three_four
  }
\tl_log:N \l_tmpa_tl
\seq_set_split:NnV \l_tmpa_seq {_} \l_tmpa_tl
\seq_log:N \l_tmpa_seq
\ExplSyntaxOff
\end{document}
```

Log file:

```none
> \l_tmpa_tl=one_two_three_four.
The sequence \l_tmpa_seq contains the items (without outer braces):
>  {one_two_three_four}.
```

The sequence is not split. It should contains four items `{one}{two}{three}{four}`, but it doesn't.

At first I think this is due to a surrounding pair of braces, but it is apparently not the case (as you can see in the output of `\tl_log:N`).

Interestingly, if the delimiter is set to nothing, `\seq_set_split` works well.

```none
\seq_set_split:NnV \l_tmpa_seq {} \l_tmpa_tl

> \l_tmpa_tl=one_two_three_four.
The sequence \l_tmpa_seq contains the items (without outer braces):
>  {o}
>  {n}
>  {e}
>  {_}
>  {t}
>  {w}
>  {o}
>  {_}
>  {t}
>  {h}
>  {r}
>  {e}
>  {e}
>  {_}
>  {f}
>  {o}
>  {u}
>  {r}.
```

Obviously, if the token list is input directly as `one_two_three_four`, `\seq_set_split` has no problem with the delimiter.

```none
\tl_set:Nx \l_tmpa_tl
  {
    one_two_three_four
  }

> \l_tmpa_tl=one_two_three_four.
The sequence \l_tmpa_seq contains the items (without outer braces):
>  {one}
>  {two}
>  {three}
>  {four}.
```
Top Answer
Phelype Oleinik 
That's because internally `\cs_split_function:N` will hit your control sequence with `\string`, so all tokens (except spaces) will be catcode 12, but the `_` in `expl3` code is catcode 11 (and outside it's normally catcode 7), so `\seq_set_split:Nnn` won't find the “the letter \_” token. You need to turn `_` into a “string”, by using `\token_to_str:N _` or `\tl_to_str:n { _ }`.

Also note that `\exp_last_unbraced:Nf` means “`f`-expand the argument, and leave it unbraced”, rather than “the arg is unbraced”, so you _need_ braces around `\cs_split_function:N \one_two_three_four` (otherwise it just works by accident ;-​)

```latex
\documentclass{article}
\usepackage{expl3}
\begin{document}
\ExplSyntaxOn
\tl_set:Nx \l_tmpa_tl
  {
    \exp_last_unbraced:Nf \use_i:nnn
      { \cs_split_function:N \one_two_three_four }
  }
\tl_log:N \l_tmpa_tl
\cs_generate_variant:Nn \seq_set_split:Nnn { NxV }
\seq_set_split:NxV \l_tmpa_seq { \token_to_str:N _ } \l_tmpa_tl
\seq_show:N \l_tmpa_seq
\ExplSyntaxOff
\end{document}
```

The log will show:
```none
The sequence \l_tmpa_seq contains the items (without outer braces):
>  {one}
>  {two}
>  {three}
>  {four}
```
comments
transcript

Enter question or answer id or url (and optionally further answer ids/urls from the same question) from

Separate each id/url with a space. No need to list your own answers; they will be imported automatically.