NIAD-3416: Performance improvements

[ORybak5]

What

Caching unmarshaller in XmlUnmarshallUtil class

Why

Creating unmarshallers is a very expensive operation and it takes (unnecessarily) a lot of resources. Once we create an unmarshaller it should be cached and reused for better efficiency.

Type of change

Please delete options that are not relevant.

• Bug fix (non-breaking change which fixes an issue)
• New feature (non-breaking change which adds functionality)
• Internal change (non-breaking change with no effect on the functionality affecting end users)
• Breaking change (fix or feature that would cause existing functionality to not work as expected)

Checklist:

• I have performed a self-review of my code
• I have made corresponding changes to the documentation
• My changes generate no new warnings
• I have added tests that prove my fix is effective or that my feature works
• New and existing unit tests pass locally with my changes
• I have updated the Changelog with details of my change in the UNRELEASED section if this change will affect end users
• A corresponding change has been made to the Mapping Documentation repository

[github-actions]

Images built and published to ECR using a Build Id of PR-925-fd807bc

[github-actions]

• Surviving mutants in this change: 2
• Killed mutants in this change: 16

class	surviving	killed
🧟uk.nhs.adaptors.pss.translator.util.XmlUnmarshallUtil	2	1
💯uk.nhs.adaptors.pss.translator.util.ResourceReferenceUtil	0	15

See https://pitest.org

• Arcmutate licence valid until 2026-07-16 (yyyy/mm/dd) (non-commercial use only)
• Project uses Spring, but the Arcmutate Spring plugin is not present.

[MartinWheelerMT]

Can we improve the description / why / what of this PR please.

The what should describe what has been done - i.e:

• Refactored SomeMethod in SomeFile.
• Removed ThisFile as no longer needed.
• Added XML files for new test cases etc.

The why should be the reasoning for this changes - i.e.:

• SomeMethod was being used to select the last item from a list by parsing all values, it is more efficient to simply read the last values directly.
• Removed files which are no longer in use due to test refactoring.
• Mutation testing revealed two areas of concern, which are now covered by additional tests.

[MartinWheelerMT]

This file still has places where we have assertions such as: assertThat(references.size()).isOne() which is less efficient than assertThat(references).hasSize(1)