fix: prevent putDimensions from storing duplicate dimensions

[aaronlna]

Issue

#20 (previous, related issue)

Description of changes

It appears a prior change addressing this issue may be allowing both variably-sized dimensions sets, and dimension sets with different keys, to be added, even if they are true duplicates of existing sets in the collection. Examples:

[
  ['A', 'B'],
  ['A', 'B', C'],
  ['A', 'B'], // this would be considered acceptable
]

// Output: 3 dimension sets (should be 2)

[
  ['A', 'B'],
  ['B', 'C'],
  ['A', 'B'], // this would be considered acceptable
]

// Output: 3 dimension sets (should be 2)

This change fixes that aforementioned behavior and adds an additional behavior:

1. duplicate dimension sets replace (and are sorted after) existing dimension sets such
   that newer/latest dimension values are used by the root EMF node

Example:

putDimensions({ "keyA": "valueA1" })
putDimensions({ "keyA": "valueA2" })
putDimensions({ "keyA": "valueA3", "keyB": "valueB1" })
putDimensions({ "keyB": "valueB2", "keyA": "valueA4" })

// expected MetricsDirective dimensions:
[
  { "keyA": "valueA2" },
  { "keyB": "valueB2", "keyA": "valueA4" },
]

// expected EMF target members:
{
  "keyA": "valueA4",
  "keyB": "valueB2",
}

Description of how you validated changes

Unit tests were updated to check for these two edge-cases.

---

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

[Stephen-Bao]

Looks good.