Bokeh output in IPySpark is not in correct format

I'm using the latest Zeppelin version compile from the source code (Version 0.11.0-SNAPSHOT)

I'm facing the issue explained in https://issues.apache.org/jira/browse/ZEPPELIN-4771

It means that when running a simple Bokeh code with %ipyspark (or even with %python.ipython)interpreter, the text output is not correctly formatted.

You can try for example with this example where I did a on purpose typo:

%ipyspark

import numpy as np
from bokeh.plotting import figure, show, output_notebook

output_notebook()

x = np.linspace(-6, 6, 500)
y = 8*np.sin*np.sinc

p = figure(width=800, height=300, title="", tools="",
           toolbar_location=None, match_aspect=True)

p.line(x, ey, color="navy", alpha=0.4, line_width=4)

show(p)

then the output is scrabbled like this:
Loading BokehJS ...
 
--------------------------------------------------------------------------- NameError Traceback (most recent call last) Cell In[40], line 12  7 y = 8*np.sin*np.sinc  9 p = figure(width=800, height=300, title="", tools="",  10 toolbar_location=None, match_aspect=True) ---> 12 p.line(x, ey, color="navy", alpha=0.4, line_width=4)  14 show(p) NameError: name 'ey' is not defined