Use LRU cache for statement parser

[JonasKunz]

What does this PR do?

Replaces the caching for our Statement parsing with a simple LRU cache.
Previously we, used a bounded Map and made some exclusions to caching:

• Non-prepared statements were not cached
• Too long or too short statements were not cached

This could lead to edge cases where the caching is not effective, which we encountered with a user.
This PR switches to the same caching approach also used by the OTel-agent: Simply cache everything, but use a bounded LRU cache.

Checklist

• This is something else
  • I have updated CHANGELOG.asciidoc

[v1v]

run docs-build

[jackshirazi]

We don't need any unit/integration tests beyond what is already there, but we should report a manual test that shows the difference. Not urgent though