Speed up Pane.clone with 50%

[MarcSkovMadsen]

Pane.clone did the same look up three times which was costly.

This changes to one lookup and also makes the code more readable.

Measurement 1

Running pytest on below code shows ~3% speed up consistently.

import panel as pn
item = pn.pane.Markdown("Hello World")

def test_html_clone():
    for i in range(20000):
        item.clone()

The reason why I found this is that I can see in apps .clone loops over self.param.values().items(). And loops over self.param.values().items() takes up +5% of the time of loading apps in my tests. My hypothesis is that that one could be improved significantly.

[codecov]

Codecov Report

Attention: 5 lines in your changes are missing coverage. Please review.

Comparison is base (4bfe371) 84.02% compared to head (dac33af) 84.03%.
Report is 3 commits behind head on main.

Files	Patch %	Lines
panel/util/parameters.py	92.53%	5 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #5848      +/-   ##
==========================================
+ Coverage   84.02%   84.03%   +0.01%     
==========================================
  Files         290      291       +1     
  Lines       42420    42479      +59     
==========================================
+ Hits        35643    35698      +55     
- Misses       6777     6781       +4     

Flag	Coverage Δ
ui-tests	40.87% <45.34%> (-0.02%)	⬇️
unitexamples-tests	72.43% <93.02%> (+0.04%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

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

[ahuang11]

Wow super excited for all the speed gains you're discovering!

[philippjfr]

I've done this a few times now so most low hanging fruit are likely gone but after param 2.0 some performance characteristics have likely changed. We should really add a basic asv based performance profiling suite, even if it only covers a small number of test cases initially.