Hello everyone. I am writing with the following problem I am facing. I have created a field "test" which is an object type. In it I write an object with the following structure:

{1: {id:1, price: 10},2: {id:2, price: 15},3: {id:3, price: 20}}

and when I run a multisearch request:

{
  "searches": [
    {
      "q": "adidas",
      "per_page": 16,
      "page": 1,
      "case_sensitive": false,
      "collection": "products_4878",
      "filter_by": "test.2.id:=2",
      "group_by": "item_id",
      "group_limit": 1,
      "sort_by": "test.2.price:asc",
      "query_by": "name",
      "include_fields": "id,item_id,name"
    }
  ]
}

there is no problem, but if i use "test.10" got this error "Could not find a filter field named

customer_group.10.id

in the schema." for "filter_by", this error "Could not find a facet field named

customer_group.10.price

in the schema." for "facet_by" and this error "Could not find a field named

customer_group.10.id

in the schema." for "group_by". I used version 27.1. How can i fih that problem? Option for dinamicly update of schema is not OK. there is option for many create and delete records for test with different id's

Do you also have another field called

customer_group

in your schema? It will be useful if you can post a fully reproducible example: 1. Your collection schema 2. One sample doc 3. The query which fails

There is schema:

{
  "name": "test_collection",
  "enable_nested_fields": true,
  "fields": [
    {
      "name": "name",
      "type": "string"
    },
    {
      "name": "customer_group",
      "type": "object",
      "facet": true,
      "optional": true
    }
  ]
}

2 documents:

[
{
    "id": "1",
    "name": "Stark Industries",
    "customer_group": {
        "1": {
            "id": 1,
            "price": 10
        },
        "2": {
            "id": 2,
            "price": 15
        },
        "3": {
            "id": 3,
            "price": 20
        }
    }
},
{
    "id": "2",
    "name": "Stark Industries 2",
    "customer_group": {
        "3": {
            "id": 3,
            "price": 10
        },
        "4": {
            "id": 4,
            "price": 15
        },
        "5": {
            "id": 5,
            "price": 20
        }
    }
}
]

And multysearch query with success response:

{
  "searches": [
    {
      "collection": "test_collection",
      "q": "*",
      "filter_by": "customer_group.3.id:=3"
    }
  ]
}

and with error:

{
  "searches": [
    {
      "collection": "test_collection",
      "q": "*",
      "filter_by": "customer_group.6.id:=6"
    }
  ]
}

Got it. So since that field doesn't exist yet we throw an error. I understand that this doesn't quite work for dynamic fields of an object. We are going to fix this.

great. thanks.

This is on my plate for this week. I will post an update when we have a RC build with the fix.

Excellent news. Thank you very much

Not fully done yet. It works for query_by, and facet_by but not yet for group_by or filter_by. I will post an update when all 4 are done.

@Kishore Nallan group_by is most important to me unfortunately. Will there be any development with him?

What do you expect to happen when group by field is not found in any docs?

I have field item_id, and object test with that structure: {1: {id:1, price: 10},2: {id:2, price: 15},3: {id:3, price: 20}}. when group by item_id, test.1 if no records with test.1 then group only by item_id. if i use onli test.1 for group and no records with that match, there is no grouped documents 🙂

Ok will check on this

very thanks 🙂

Don't have a build yet, will share once filter by is done.

ok, thanks

Please check

28.0.rc22

-- we have added a

validate_field_names

boolean parameter which you can use to control the field validation behavior. This is now implemented for group by and filter by as well.