bug#71025: set-fontset-font doesn't show some scripts as glyphless characters

[Rodrigo Morales <me@rodrigomorales.site>]

Eli Zaretskii <eliz@gnu.org> writes:

> Do you have site-init files in your Emacs installation?  Or some
> early-init file?  Maybe something other than the init file is
> affecting this.  Otherwise, I have no idea why this doesn't work for
> you, as it does for me.

load-history doesn't show other files in the ~/.config directory besides
~/.config/emacs/init.el. Here's how I found it out.

I used the same configuration file.

#+HEADER: :tangle ~/.config/emacs/init.el
#+BEGIN_SRC elisp
(set-fontset-font t 'brahmi nil)
(set-fontset-font t 'egyptian nil)
(set-fontset-font t 'ethiopic nil)
(set-fontset-font t 'tagbanwa nil)
(set-fontset-font t 'han nil)
#+END_SRC

When Emacs GUI opened, I visited the file =/tmp/a.txt=. Brahmi and Egyptian Hieroglyphs were shown as hexadecimal codes. Amharic and Tagbanwa and Chinese were shown with other fonts. I used =describe-char= to find out which font those scripts were using.

Amharic: ftcrhb:-GOOG-Noto Sans Ethiopic-regular-normal-normal-*-12-*-*-*-*-0-iso10646-1 (#x2B)
Tagbanwa: ftcrhb:-GOOG-Noto Sans Tagbanwa-regular-normal-normal-*-12-*-*-*-*-0-iso10646-1 (#x12)
Chinese: ftcrhb:-GOOG-Noto Sans CJK KR-regular-normal-normal-*-12-*-*-*-*-0-iso10646-1 (#x2703)

#+HEADER: :tangle /tmp/a.txt
#+BEGIN_SRC text
The following lines were retrieved from the HELLO buffer which is
opened by calling view-hello-file in GNU Emacs 29.3.
Brahmi (𑀩𑁆𑀭𑀸𑀳𑁆𑀫𑀻)       𑀦𑀫𑀲𑁆𑀢𑁂
Egyptian Hieroglyphs (𓂋𓏤𓈖𓆎𓅓‌𓏏𓊖) 𓅓𓊵𓏏𓊪, 𓇍𓇋𓂻𓍘𓇋
Amharic (አማርኛ)  ሠላም
Tagbanwa (ᝦᝪᝯ)  ᝫᝩᝬᝥ ᝣᝮᝧᝯ
Chinese (中文,普通话,汉语)     你好
#+END_SRC

I launched =emacs= and I pressed =C-h v load-history RET=. The content of the =*Help*= buffer can be found here: http://web.archive.org/web/20240520161841/http://0x0.st/XPww.txt . I searched =.config= inside the =*Help*= buffer and I only found one match: =/home/rodrigo/.config/emacs/init.el=, so I don't think there are other configuration files that are influencing this behavior.

> Or maybe your default font supports those scripts, in which case Emacs
> will use it regardless of the fontset setup?  What happens if you
> change the default font to something that doesn't support these scripts?

I changed the default font to =Cozette=.

#+HEADER: :tangle ~/.config/emacs/init.el
#+BEGIN_SRC elisp
(add-to-list 'default-frame-alist '(font . "Cozette"))
(set-fontset-font t 'brahmi nil)
(set-fontset-font t 'egyptian nil)
(set-fontset-font t 'ethiopic nil)
(set-fontset-font t 'tagbanwa nil)
(set-fontset-font t 'han nil)
#+END_SRC

When Emacs GUI opened, I visited the file =/tmp/a.txt=. Brahmi and Egyptian Hieroglyphs were shown as hexadecimal codes. Amharic and Tagbanwa and Chinese were shown with other fonts. I used =describe-char= to find out which font those scripts were using.

Amharic: ftcrhb:-GNU -Unifont Sample-regular-normal-normal-*-13-*-*-*-d-0-iso10646-1 (#x1221)
Tagbanwa: ftcrhb:-GNU -Unifont Sample-regular-normal-normal-*-13-*-*-*-d-0-iso10646-1 (#x176C)
Chinese: ftcrhb:-GNU -Unifont Sample-regular-normal-normal-*-13-*-*-*-d-0-iso10646-1 (#x4F5F)

Because Unifont was being used for those scripts, I decided to delete
that font from my system just to see what happens. I removed
=unifont-15.1.05.bdf= from the directory =~/.fonts= and executed
=fc-cache -f=. I launched Emacs again and visited =/tmp/a.txt=. I used
=describe-char= to find out the font that were using for those
scripts.

Amharic: ftcrhb:-GOOG-Noto Sans Ethiopic-regular-normal-normal-*-13-*-*-*-*-0-iso10646-1 (#x2B)
Tagbanwa: ftcrhb:-GOOG-Noto Sans Tagbanwa-regular-normal-normal-*-13-*-*-*-*-0-iso10646-1 (#x12)
Chinese: ftcrhb:-GOOG-Noto Sans CJK KR-regular-normal-normal-*-13-*-*-*-*-0-iso10646-1 (#x2703)

Note that these fonts were the same used as when I didn't change the
default font through =(add-to-list 'default-frame-alist '(font . "Cozette"))=.