feature: hydrate AsyncPG span attributes at creation (#3643)

[forafox]

Description

This change enables filtering of AsyncPG spans by moving attribute hydration from inside the span context to before span creation. Previously, spans were created with no attributes and attributes were added after span creation using span.set_attribute(), which prevented samplers from filtering based on request attributes.

The fix moves the _hydrate_span_from_args() call outside the span context and passes the hydrated attributes directly to start_as_current_span() using the attributes parameter. This ensures that database attributes (system, name, user, statement, network info) are available during span creation for sampling decisions.

Fixes #3643

Type of change

• Bug fix (non-breaking change which fixes an issue)
• New feature (non-breaking change which adds functionality)
• Breaking change (fix or feature that would cause existing functionality to not work as expected)
• This change requires a documentation update

How Has This Been Tested?

The changes were tested by:

• Running existing AsyncPG instrumentation tests to ensure no regressions
• Verifying that attributes are now available during span creation by examining the modified code structure
• Confirming that the attributes parameter is properly passed to start_as_current_span() in both _do_execute and _do_cursor_execute methods

The implementation maintains backward compatibility and does not change the external API.

Does This PR Require a Core Repo Change?

• Yes. - Link to PR:
• No.

Checklist:

See contributing.md for styleguide, changelog guidelines, and more.

• Followed the style guidelines of this project
• Changelogs have been updated
• Unit tests have been added
• Documentation has been updated

[linux-foundation-easycla]

The committers listed above are authorized under a signed CLA.

• ✅ login: forafox / name: Andrey Karabanov (a6670d7, 32136cf, e99879c)

[xrmx]

Which attributes do you need for sampling decisions? If the tests don't fail it looks like we don't have tests for not recording :)
And we need a changelog entry.

[forafox]

Which attributes do you need for sampling decisions? If the tests don't fail it looks like we don't have tests for not recording :) And we need a changelog entry.

The sampler now receives DB system/name/user, DB statement (and parameters when capture_parameters is on), and peer host/port/transport because we hydrate attributes before start_as_current_span. Added tests covering execute and cursor paths with non-recording spans to ensure those attributes are present for sampling decisions, and added a changelog entry

[tammy-baylis-swi]

Thanks for contributing! Please also have a look at the docker-tests, and do a tox -e precommit on local to address any formatting issues.

[forafox]

Thanks for contributing! Please also have a look at the docker-tests, and do a tox -e precommit on local to address any formatting issues.

@tammy-baylis-swi

Hi! Thanks for the feedback — I’ve addressed the formatting issues and updated the docker tests as requested.
Everything should be fixed now