[INTEGRATION] JupyterLab integration due to upstream dependency constraint on Rich

[RobFirth]

Describe the bug
Tracebacks in jupyter notebooks in project virtual environment with argilla are broken as a result of upstream bug in old version of rich. Version constraint on rich = "<=13.0.1", this behaviour was fixed in rich==v13.3.1.

import argilla as rg

def f(x):
    return x/0

f(42)

returns:

╭─────────────────────────────── Traceback (most recent call last) ────────────────────────────────╮
│ /tmp/ipykernel_1885/2875885747.py:1 in <module>                                                  │
│                                                                                                  │
│ [Errno 2] No such file or directory: '/tmp/ipykernel_1885/2875885747.py'                         │
│                                                                                                  │
│ /tmp/ipykernel_1885/3374322498.py:2 in f                                                         │
│                                                                                                  │
│ [Errno 2] No such file or directory: '/tmp/ipykernel_1885/3374322498.py'                         │
╰──────────────────────────────────────────────────────────────────────────────────────────────────╯
ZeroDivisionError: division by zero

Expected behaviour
the traceback gets displayed correctly.

Environment:

• Argilla Version [e.g. 1.12.1]: 1.12.1
• rich: 13.0.1

Additional context

This bug is described in the following issues:

• [BUG] Rich traceback not working proprely within notebook Textualize/rich#2271
• Fix tracebacks in Jupyter Textualize/rich#2754
• Rich Tracebacks fail on Jupyterlab jupyterlab/jupyterlab#13759
• Rich Tracebacks fail on Jupyterlab ipython/ipykernel#1075

This doesn't affect my Argilla workflow other than making it a total pain to debug issues in notebooks! Hopefully a version bump isn't too onerous - it would be a massive QOL improvement.

suggested fix
Version bump rich e.g. rich<=13.3.1

[tomaarsen]

Hello!

When we integrated rich in #2350, I specifically set rich = "<=13.0.1" because rich==13.1.0 was failing for me with this error. At that time, 13.1.0 was the most recent version.

Now I've installed the most recent version (13.4.2), and the aforementioned issue does not occur anymore. Because of this, I think we can lift the "<=13.0.1" restriction without causing crashes for users.

I'll make a PR for this.

• Tom Aarsen

[RobFirth]

Hello!

When we integrated rich in #2350, I specifically set rich = "<=13.0.1" because rich==13.1.0 was failing for me with this error. At that time, 13.1.0 was the most recent version.

Now I've installed the most recent version (13.4.2), and the aforementioned issue does not occur anymore. Because of this, I think we can lift the "<=13.0.1" restriction without causing crashes for users.

I'll make a PR for this.

• Tom Aarsen

Thanks for this @tomaarsen - I was hoping that this was the reason for pinning the version and it wouldn't impact on anything elsewhere. Cheers for the speed 🚀