Package capture harvest

[lrafeei]

This PR aims to alleviate some of the time consumed by package data collection in the connect cycle. Instead of collecting all modules used in the application at startup, the agent uses the fast harvest cycle to incrementally collect this data and sends it though the update_loaded_modules endpoint instead of the connect endpoint (through the "environment" key).

Note: despite the term "Jars" not being used by Python to refer to packages/modules, this is the exact string the endpoint needs in order to process the data.

[github-actions]

🦙 MegaLinter status: ❌ ERROR

Descriptor	Linter	Files	Fixed	Errors	Elapsed time
❌ PYTHON	bandit	3		1	5.3s
✅ PYTHON	black	5	1	0	2.18s
✅ PYTHON	flake8	5		0	0.69s
✅ PYTHON	isort	5	1	0	0.29s
✅ PYTHON	pylint	5		0	4.7s

See detailed report in MegaLinter reports
Set VALIDATE_ALL_CODEBASE: true in mega-linter.yml to validate all sources, not only the diff

MegaLinter is graciously provided by

[codecov-commenter]

Codecov Report

Attention: Patch coverage is 96.15385% with 1 line in your changes missing coverage. Please review.

Project coverage is 71.43%. Comparing base (3a001b0) to head (a5faab6).

Files with missing lines	Patch %	Lines
newrelic/core/application.py	94.73%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #1228      +/-   ##
==========================================
- Coverage   81.23%   71.43%   -9.80%     
==========================================
  Files         197      197              
  Lines       21643    21662      +19     
  Branches     3782     3784       +2     
==========================================
- Hits        17581    15474    -2107     
- Misses       2919     5032    +2113     
- Partials     1143     1156      +13     

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