[Design and Development] Data Fields as Primary Key/Identifier in-place of _id

[TanmoySG]

Currently, _id is generated and used as primary identifier for a data-row(?). For example, now for data inserted , the Primary ID

{
  "name" : "T",
  "student_id" : 0123,
  "age": 21
}

the Identifier is random generated identifier _id,

{
  "axo0193nc4101383" : {"name" : "T", "student_id" : 0123, "age": 21},
}

---

For advanced use-cases, user might want to use any of the field values as identifier instead of _id. Eg instead of _id we want to use student_id as primary key.

{
  "0123" : {"name" : "T", "student_id" : 0123, "age": 21},
}

---

Possible solutions

• Adding a field IdentifierField (string) in the Datum struct

  wunderDB/model/models.go

  Lines 42 to 46 in e7b8cdd

  	type Datum struct {
  	Data interface{} `json:"data"`
  	Metadata Metadata `json:"metadata"`
  	Identifier Identifier `json:"id"`
  	}

• When creating a collection, in addition to schema pass the identifier

{
    "name": "test-metadata1",
    "schema": {"type": "object", "properties": {"name": {"type": "string"}},"required": ["name"]},
    "identifier" : "name"
}

• In code, check if identifier value is in the required array of the schema - request valid only if it is there or if not there can add through code.
• While inserting data, check if identifier field is _id or any other custom field,
  • check if the identifier field is in the data, if not reject
  • if there, check if value already exists data[identifier.value] if not insert the data