DocumentSearchServiceImpl and Repetitions

Kernel
,
1

Hi. I just learned about the DocumentSearchServiceImpl from https://elearning.mgm-tp.com/learner_module/show/34223?lesson_id=131358&section_id=447313

I have several questions on this:

  • I wonder why DocumentSearchServiceImpl is used.
    – It obviously simplifies the search for documents compared to the seen complexity in DocumentQueryService
    – However, it’s from an internal package (com.mgmtp.a12.kernel.md.document.internal.service) and is this intended to be used by projects?
    – Is there another “simpler” solution possible with a different public API?
  • I wonder what the “repetition” Array means.
    – There is no Javadoc on this, so I can only guess.
    – What is the default value?
    – How is it expected to be filled?
    – Is there an easier solution to the repetition array?

Thanks.

2

Hi @stefan-grey-canyon ,

I have changed the tag to Kernel because the API you are mentioning is a kernel API. This is a search service that is meant to be used for searching a fields in the document. It is internal because it is hidden behind the Public API interface and I believe it should be used via DocumentServiceFactory.

DocumentQueryService is used to search documents by the field values and this is Data Services API. This operates the set of multiple documents.

3

Just a note - I stumbled upon that, as well. IMO any experienced Java developer would guess that IDocumentSearchService searches for and returns IDocuments.

4

@stefan-grey-canyon The repetition array is meant to add indices to document element paths so as to uniquely identify element instances in a document, as I understood it. E.g. if you have a document model such as (pseudocode)

{
	Group: {
		name: "root",
		repeatability: 1,
		elements: [
			{
				Group: {
					name: "repeatableGroup",
					repeatability: 10,
					elements: [
						{
							Field: {
								name: "Field"
							}
						}
					]
				}
			}
		]
	}
}

and a document

{
	root:
	{
		repeatableGroup: [
			{Field: "abc"},
			{Field: "xyz"}
		]

	}
}

querying for the value at document path root/repeatableGroup/Field would give you “abc” with repetitions array [1, 1, 1] and “xyz” with repetitions array [1, 2, 1] (non-repeatable groups/fields always have 1 and indices are one-based)

I agree that this is a bit confusing and could benefit from improved documentation.