llm chat - readline problems still present

[kevinburke]

When I open llm chat, I expect that using the left and right arrow keys will navigate the cursor but instead I get nasty ANSI codes printed to the screen.

$ llm chat
Chatting with gpt-4
Type 'exit' or 'quit' to exit
Type '!multi' to enter multiple lines, then '!end' to finish
> foo bar ^[[D^[[D^[[D^[[D^[[D

I am using the latest version of llm, commit 5cc8efe (the readline support commit), installed by checking out this commit locally, starting a virtualenv and running "python setup.py install", then running e.g. ./venv/bin/llm.

I also verified that my Python version has readline support:

$ python
Python 3.10.13 (main, Nov  7 2023, 20:58:26) [Clang 15.0.0 (clang-1500.0.40.1)] on darwin
Type "help", "copyright", "credits" or "license" for more information.
>>> import readline
>>> print(readline.__doc__)
Importing this module enables command line editing using GNU readline.

Let me know what else I can do to help troubleshoot.

[kugurerdem]

I had the same experience. Running rlwrap llm chat resolved the issue for me when I wanted to chat through llm. This is essentially a workaround that worked for me, but the problem still remains.

It's a bit unfortunate that this great program has issues with readline. I would really like to see this problem being addressed and solved.

[simonw]

Working with ChatGPT: https://chat.openai.com/share/b5e1a34e-e365-4455-8d17-265ef80101c0 I found this magic incantation which seems to do the right thing:

    readline.parse_and_bind("\\e[D: backward-char")
    readline.parse_and_bind("\\e[C: forward-char")

Weirdly obscure!

[simonw]

Demo of that fix: