[Monitor] Generate exporter with TypeSpec

[pvaneck]

Generation from TypeSpec from the branch in this PR: Azure/azure-rest-api-specs#38795

Noteable changes

• The time field in TelemetryItem seems to have always been typed as datetime.datetime, but was being set as a ISO8601 date string. I have adjusted the tests and the _create_telemetry_item method to set the time to a datetime object and have adjusted the tests accordingly. The datetime objects will be serialized to the ISO8601 formatted strings (e.g. 2025-12-15T13:45:30.0000000Z when sent in a request anyway, so this should be fine.
• In TypeSpec, defaults in models are treated as server-side defaults. The version property in the Domain model has a default of 2, however, the generated code does not include this as a default. The code was updated to explicitly set the version argument when constructing subclasses of Domain.

Notes:

• Since two different TypeSpec projects are contained in this project (Exporter and LiveMetrics), I am using the batch generation feature of tsp-client where the Exporter project will be emitted into azure/monitor/opentelemetry/exporter/_generated. LiveMetrics will be emitted to azure/monitor/opentelemetry/exporter/_quickpulse/_generated in the future.
• One test, tests/trace/test_trace.py::TestAzureTraceExporter::test_get_otel_resource_envelope, was failing after the regeneration. It is comparing OTel Resource attributes with MetricsData.properties, but the properties field is type and serialized as Dict[str, str]. Thus, all values end up as strings when checking the values of properties.

Asks for team

1. Validate whether using datetime objects instead of strings when creating TelemetryItems is acceptable.
2. Validate whether properties being strings inside MetricsData is acceptable.

[rads-1996]

@pvaneck I can create a separate PR that runs the black format checks on the existing exporter code. I will share it once it is ready for review.

[pvaneck]

@rads-1996 Thanks, but I believe it was already formatted with #44462. I just need to rebase off of main.

[rads-1996]

why are these function calls changed?