Implement support for multi-index columns in Tabulator

[philippjfr]

Fixes #6794

Adds support for rendering arbitrarily deep column multi-indexes by using the column grouping feature of tabulator:

import pandas as pd
import numpy as np
import panel as pn
pn.extension('tabulator')

# Define the three levels of column index
level_1 = ['A', 'A', 'A', 'B', 'B', 'B']
level_2 = ['one', 'one', 'two', 'two', 'three', 'three']
level_3 = ['X', 'Y', 'X', 'Y', 'X', 'Y']

# Combine these into a MultiIndex
multi_index = pd.MultiIndex.from_arrays([level_1, level_2, level_3], names=['Level 1', 'Level 2', 'Level 3'])

# Create a DataFrame with this MultiIndex as columns
df = pd.DataFrame(np.random.randn(4, 6), columns=multi_index)

pn.Column(df, pn.widgets.Tabulator(df))

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 82.23%. Comparing base (07f9be3) to head (be65266).

Additional details and impacted files

@@           Coverage Diff           @@
##             main    #7108   +/-   ##
=======================================
  Coverage   82.22%   82.23%           
=======================================
  Files         327      327           
  Lines       48725    48750   +25     
=======================================
+ Hits        40066    40089   +23     
- Misses       8659     8661    +2     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.