Use param in context copy instead of added function

[markkuhn]

Issue #, if available: N/A

Description of changes:
This is a small change related to this comment on a PR in the node lib: awslabs/aws-embedded-metrics-node#136 (comment)

To keep all emf libraries consistent I have added that change to this library.

By submitting this pull request, I confirm that you can use, modify, copy, and redistribute this contribution, under the terms of your choice.

[hussam789]

PR Code Suggestions ✨

Category	Suggestion	Impact
Possible issue	Use deepcopy for dimensions The current implementation directly assigns self.dimensions to new_dimensions when preserve_dimensions is True, which creates a reference instead of a deep copy. This can lead to unexpected behavior if the original dimensions are modified after copying. aws_embedded_metrics/logger/metrics_context.py [164] -new_dimensions: List[Dict] = [] if not preserve_dimensions else self.dimensions +new_dimensions: List[Dict] = [] if not preserve_dimensions else copy.deepcopy(self.dimensions) Apply this suggestion Suggestion importance[1-10]: 8 __ Why: The suggestion correctly identifies a potential bug related to aliasing by referencing self.dimensions directly when preserve_dimensions is True, and the use of copy.deepcopy would ensure that subsequent modifications do not affect the copied context. However, it assumes that importing the copy module is handled elsewhere.	Medium
More