Helm: do not override default exclusions from 'discovery'

[marevers]

Closes #1454

This changes how the ConfigMap contents are generated. I'm not 100% happy with how it works, but it is the best solution that I could come up with without introducing breaking changes.

• Moves the default configuration from config.data to a new template in templates/_config.tpl
• The beyla.defaultConfig template in that file define the original default configuration, retaining those preset values
• A subsequent beyla.config template merges config.data with the output of beyla.defaultConfig, giving precedence to config.data

Caveats are:

• As the single source of truth for the default configuration is now in templates/_config.tpl, while I could technically leave the same default configurations in config.data as well, but that could be confusing
• With mergeOverwrite, empty values are ignored so passing something like discovery: {} would not actually clear the discovery configuration as the default configuration is set. The only way to remove it in this case is to set it to null.

I think the logical problem here stems mostly from the preset configuration (network or application) intersecting with just taking all values from config.data. It might be advisable to change/remove that, but that will of course be a breaking change.

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 81.26%. Comparing base (5770689) to head (a8548f1).
Report is 2 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #1473      +/-   ##
==========================================
+ Coverage   81.11%   81.26%   +0.15%     
==========================================
  Files         149      149              
  Lines       15342    15253      -89     
==========================================
- Hits        12444    12396      -48     
+ Misses       2281     2249      -32     
+ Partials      617      608       -9     

Flag	Coverage Δ
integration-test	59.74% <ø> (+0.04%)	⬆️
k8s-integration-test	60.64% <ø> (+0.01%)	⬆️
oats-test	34.02% <ø> (+0.19%)	⬆️
unittests	52.11% <ø> (+0.34%)	⬆️

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.