feat(config): sampler plugin loading via entry points for declarative config

[MikeGoldsmith]

Depends on #5093 (shared load_entry_point utility). Best reviewed after that merges.

Description

Extends sampler support in declarative file configuration to handle custom (plugin) samplers, matching the spec's PluginComponentProvider mechanism.

Background

Python's typed dataclass models (generated from JSON Schema) drop unknown keys. A YAML key like my_custom_sampler: {} would be silently lost before reaching the factory. Other SDKs avoid this by working with non-typed structs/maps, which preserve the unknown key as the plugin name.

Solution

Sampler and ParentBasedSampler are changed from @dataclass to TypeAlias = dict[str, Any] in models.py. This reflects how the loader already stores nested fields (raw dicts via OpenTelemetryConfiguration(**data) with no recursive conversion), and means the plugin name is preserved as a dict key through the pipeline.

_create_sampler() now has a single code path: extract the one key as the sampler name, look it up in _SAMPLER_REGISTRY (known names: always_on, always_off, trace_id_ratio_based, parent_based), then fall back to load_entry_point("opentelemetry_sampler", name) for unknown names.

Custom sampler example

tracer_provider:
  sampler:
    my_custom_sampler: {}

[project.entry-points."opentelemetry_sampler"]
my_custom_sampler = "my_package:MySamplerClass"

Closes #5071