vincentaxhe has asked for the wisdom of the Perl Monks concerning the following question:

I use Term::ReadLine, but It handle wide character(chinese) poorly, though the input is recognized correctly, is there a better module be able to deal wide character input, if just like type in a normal console

    state $term = Term::ReadLine->new('note');
    my $prompt = "LINE:";
    chomp($note = $term->readline($prompt));
    $note = decode 'UTF-8', $note;
    $term->addhistory($note) if $note =~ /\S/;

[download]

Replies are listed 'Best First'.
Re: Term::ReadLine handle wide character poorly
by The_Dj (Scribe) on May 14, 2024 at 03:33 UTC
    Hi

    I tested your code above without the decode and it worked perfectly* as I expected.

    By default it was using Term::ReadLine::Gnu but I also tried using Term::ReadLine::Perl5 and it also worked exactly the same

    * I don't have a chinese keyboard, but I was copying 弛 and 地 from wikipedia

    What exactly is it doing wrong for you?
[reply]
[d/l]
[select]

      It swallow wide character to squeeze in a non-wide character length, it start to do that after input eight or nine wide characters, it keeps doing that, then wrap line upward, very clumsy line input environment

[reply]
      Can you paste a string here I could test with?
[reply]

        Its behavior varies depending you paste repeatly and type in, paste will show '[27m', type in will squeeze character, neither is desirable. try paste '一生'.

        I use fcitx in arch, you will need 'fcitx5 fcitx5-chinese-addons fcitx5-configtool fcitx5-gtk-git fcitx5-pinyin-zhwiki fcitx5-qt', relogin and some environment settings. here is my hyprland env settings related to fcitx. make it work takes some efforts.

        env = XMODIFIERS, @im=fcitx
# env = GTK_IM_MODULE, fcitx
env = QT_IM_MODULE, fcitx
env = SDL_IM_MODULE, fcitx
env = ECORE_IMF_MODULE,xim

        [download]
[reply]
[d/l]
        sorry, its alacritty I use, kitty and foot works fine.
[reply]
Re: Term::ReadLine handle wide character poorly
by The_Dj (Scribe) on May 15, 2024 at 03:38 UTC
    I see an issue in your sample code;
    You don't need $term->addhistory($note) if $note =~ /\S/; - The readline method already adds to the history.
    But using addhistory you are doubling everything in your history.

[reply]
[d/l]
[select]
      indeed, I preload some history by     $term->addhistory($_) foreach split "\n", $tasks[$selectedtask]{note} // ""; I assume its all manually, and perldoc also give example like this. thx
[reply]
[d/l]
Re: Term::ReadLine handle wide character poorly
by ForgotPasswordAgain (Vicar) on May 14, 2024 at 16:24 UTC
    I haven't tried it, but from what I read you might be able to do: binmode($term->IN, ':utf8');
[reply]
[d/l]
      sorry, its alacritty I use, kitty and foot works fine.
[reply]

        That does rather point the finger at alacritty then. Are you on the latest version? If so, maybe raise the issue with them?

        🦛

[reply]

Back to Seekers of Perl Wisdom