PR: Set SPY_PARENT_PID environment variable in kernelspec (IPython console)

[ccordoba12]

Description of Changes

• That's necessary so that the kernel poller checks the Spyder pid for crashes instead of the conda run one. That way, if Spyder dies unexpectedly, the kernel will die as well.
• Depends on PR: Use our own poller to check if the kernel parent is alive on Unix spyder-kernels#551.

Issue(s) Resolved

Fixes #22414.

Affirmation

By submitting this Pull Request or typing my (user)name below,
I affirm the Developer Certificate of Origin
with respect to all commits and content included in this PR,
and understand I am releasing the same under Spyder's MIT (Expat) license.

I certify the above statement is true and correct: @ccordoba12

[ccordoba12]

@dalthviz, please check if this solution works for you on Windows.

[dalthviz]

Gave check to this by creating a big dataframe and then trying to view it with the DataFrame editor. Created the dataframe using the following code:

import numpy as np
import pandas as pd

np.random.seed(42)

num_columns = 1000
num_rows = 1000000

data = np.random.rand(num_rows, num_columns)
df = pd.DataFrame(data, columns=[f'Column{i+1}' for i in range(num_columns)])

So, while the Variable Explorer is still loading I tried to close Spyder repeatedly which then cause it to crash. From what I can see after 5-6 seconds my available RAM is restored (which I think could be a sign that the kernel process is killed).

As a side note, searched for Python process after the crash and seems like there are still 4 Python process running even after my RAM is restored. Not totally sure if those are expected but letting you know