Nested Solr Structure

Question

Is it possible to create a nested data structure in Solr using the content search index configurations? I'm curious if I can create a document like the following, then query against and/or facet the fields in the nested structure.

{   
    "_group":"fe58033195a142f7b33331d85e6120a9"
    "school_name_s": "School 1"
    "programs": [
        {
            "program_name_s":"Program 1",
            "program_type_s":"Program Type 1"
        },
        {
            "program_name_s":"Program 2",
            "program_type_s":"Program Type 1"
        },
        {
            "program_name_s":"Program 3",
            "program_type_s":"Program Type 2"
        }
    ]                
}

I'm currently using Sitecore 10.1.0 and Solr 8.4.0

Answer 1

Short Answer

No. You cannot create nested documents within Solr using the default Sitecore indexing functionality.

Since there's not a lot of background on the goals that you are trying to achieve, I can share some options that are available; each with their own benefits and drawbacks. It'll be up to you to decide whether to implement any of these suggestions.

Option 1:

Use GraphQL (comes with Headless Services 16+) to query your parent and child documents together.

• https://doc.sitecore.com/xp/en/developers/hd/190/sitecore-headless-development/sitecore-graphql-api.html
• https://graphql.org/learn/

Option 2:

Create an index re-crawler that builds a new index using the nested document functionality. There's documentation on how to do this in Solr here:

https://solr.apache.org/guide/8_4/indexing-nested-documents.html

This option does give you the ability to use BlockJoin faceting to search your documents and child documents.

https://solr.apache.org/guide/8_4/blockjoin-faceting.html

I've done this in the past, but some thorough consideration needs to be given to an update strategy on the new index.

WARNING: With the exception of in-place updates, the whole block must be updated or deleted together, not separately. For some applications this may result in tons of extra indexing and thus may be a deal-breaker.

Option 3:

Use a computed field to store your child documents as json within a string field. This is quick and dirty and will get your content into the index, but you can't do any filtering, faceting or block joins on this field.