Parse Complex JSON -- Map - json

Question

I need to parse the complex JSON (below) IN SCALA to get the values of "expression" and "value" in "measure" key i.e I need List (COUNT, COUNT_DISTINCT ...) and List (1,tbl1.USER_ID ...).
I tried multiple options, but it is not working. Any help is appreciated
{
"uuid": "uuidddd",
"last_modified": 1559080222953,
"version": "2.6.1.0",
"name": "FULL_DAY_2_mand_date",
"is_draft": false,
"model_name": "FULL_DAY_1_may05",
"description": "",
"null_string": null,
"dimensions": [
{
"name": "PLATFORM",
"table": "tbl1",
"column": "PLATFORM",
"derived": null
},
{
"name": "OS_VERSION",
"table": "tbl1",
"column": "OS_VERSION",
"derived": null
}
],
"measures": [
{
"name": "_COUNT_",
"function": {
"expression": "COUNT",
"parameter": {
"type": "constant",
"value": "1"
},
"returntype": "bigint"
}
},
{
"name": "UU",
"function": {
"expression": "COUNT_DISTINCT",
"parameter": {
"type": "column",
"value": "tbl1.USER_ID"
},
"returntype": "hllc(12)"
}
},
{
"name": "CONT_SIZE",
"function": {
"expression": "SUM",
"parameter": {
"type": "column",
"value": "tbl1.SIZE"
},
"returntype": "bigint"
}
},
{
"name": "CONT_COUNT",
"function": {
"expression": "SUM",
"parameter": {
"type": "column",
"value": "tbl1.COUNT"
},
"returntype": "bigint"
}
}
],
"dictionaries": [],
"rowkey": {
"rowkey_columns": [
{
"column": "tbl1.OS_VERSION",
"encoding": "dict",
"encoding_version": 1,
"isShardBy": false
},
{
"column": "tbl1.PLATFORM",
"encoding": "dict",
"encoding_version": 1,
"isShardBy": false
},
{
"column": "tbl1.DEVICE_FAMILY",
"encoding": "dict",
"encoding_version": 1,
"isShardBy": false
}
]
},
"hbase_mapping": {
"column_family": [
{
"name": "F1",
"columns": [
{
"qualifier": "M",
"measure_refs": [
"_COUNT_",
"CONT_SIZE",
"CONT_COUNT"
]
}
]
},
{
"name": "F2",
"columns": [
{
"qualifier": "M",
"measure_refs": [
"UU"
]
}
]
}
]
},
"aggregation_groups": [
{
"includes": [
"tbl1.PLATFORM",
"tbl1.OS_VERSION"
],
"select_rule": {
"hierarchy_dims": [],
"mandatory_dims": [
"tbl1.DATE_HR"
],
"joint_dims": []
}
}
],
"signature": "ttrrs==",
"notify_list": [],
"status_need_notify": [
"ERROR",
"DISCARDED",
"SUCCEED"
],
"partition_date_start": 0,
"partition_date_end": 3153600000000,
"auto_merge_time_ranges": [
604800000,
2419200000
],
"volatile_range": 0,
"retention_range": 0,
"engine_type": 4,
"storage_type": 2,
"override_kylin_properties": {
"job.queuename": "root.production.P0",
"is-mandatory-only-valid": "true"
},
"cuboid_black_list": [],
"parent_forward": 3,
"mandatory_dimension_set_list": [],
"snapshot_table_desc_list": []
}
This is a snippet of the code I tried, and it is giving a null list
import org.json4s._
import org.json4s.jackson.JsonMethods._
implicit val formats = org.json4s.DefaultFormats
case class Function (
expression: String,
parameter: Parameter,
returntype: String
)
case class Parameter (
`type`: String,
value: String
)
case class Measures (
name: String,
function: Function
)
case class AllMeasuresData(uuid: String, measure: List[Measures])
val data = parse(tmp).extract[AllMeasuresData]
val names = data.measure.map(_.name)
println(names)
case class AllMeasuresData(uuid: String, measure: List[Measures])
val data = parse(tmp).extract[AllMeasuresData]
val names = data.measure.map(_.name)
println(names)

Answer 1

There are couple typos in your ADT:
Here is what you need:
case class Function (
expression: String,
parameter: Parameter,
returntype: String
)
case class Parameter (
`type`: String,
value: String
)
case class Measures (
name: String,
function: Function
)
case class AllMeasuresData(uuid: String, measures: List[Measures])
There is also an extra comma int the json, here is the correct:
{
"uuid":"uuidddd",
"last_modified":1559080222953,
"version":"2.6.1.0",
"name":"FULL_DAY_2_mand_date",
"is_draft":false,
"model_name":"FULL_DAY_1_may05",
"description":"",
"null_string":null,
"dimensions":[
{
"name":"PLATFORM",
"table":"tbl1",
"column":"PLATFORM",
"derived":null
},
{
"name":"OS_VERSION",
"table":"tbl1",
"column":"OS_VERSION",
"derived":null
} // There was an extra trailing comma here
],
"measures":[
{
"name":"_COUNT_",
"function":{
"expression":"COUNT",
"parameter":{
"type":"constant",
"value":"1"
},
"returntype":"bigint"
}
},
{
"name":"UU",
"function":{
"expression":"COUNT_DISTINCT",
"parameter":{
"type":"column",
"value":"tbl1.USER_ID"
},
"returntype":"hllc(12)"
}
},
{
"name":"CONT_SIZE",
"function":{
"expression":"SUM",
"parameter":{
"type":"column",
"value":"tbl1.SIZE"
},
"returntype":"bigint"
}
},
{
"name":"CONT_COUNT",
"function":{
"expression":"SUM",
"parameter":{
"type":"column",
"value":"tbl1.COUNT"
},
"returntype":"bigint"
}
}
],
"dictionaries":[
],
"rowkey":{
"rowkey_columns":[
{
"column":"tbl1.OS_VERSION",
"encoding":"dict",
"encoding_version":1,
"isShardBy":false
},
{
"column":"tbl1.PLATFORM",
"encoding":"dict",
"encoding_version":1,
"isShardBy":false
},
{
"column":"tbl1.DEVICE_FAMILY",
"encoding":"dict",
"encoding_version":1,
"isShardBy":false
}
]
},
"hbase_mapping":{
"column_family":[
{
"name":"F1",
"columns":[
{
"qualifier":"M",
"measure_refs":[
"_COUNT_",
"CONT_SIZE",
"CONT_COUNT"
]
}
]
},
{
"name":"F2",
"columns":[
{
"qualifier":"M",
"measure_refs":[
"UU"
]
}
]
}
]
},
"aggregation_groups":[
{
"includes":[
"tbl1.PLATFORM",
"tbl1.OS_VERSION"
],
"select_rule":{
"hierarchy_dims":[
],
"mandatory_dims":[
"tbl1.DATE_HR"
],
"joint_dims":[
]
}
}
],
"signature":"ttrrs==",
"notify_list":[
],
"status_need_notify":[
"ERROR",
"DISCARDED",
"SUCCEED"
],
"partition_date_start":0,
"partition_date_end":3153600000000,
"auto_merge_time_ranges":[
604800000,
2419200000
],
"volatile_range":0,
"retention_range":0,
"engine_type":4,
"storage_type":2,
"override_kylin_properties":{
"job.queuename":"root.production.P0",
"is-mandatory-only-valid":"true"
},
"cuboid_black_list":[
],
"parent_forward":3,
"mandatory_dimension_set_list":[
],
"snapshot_table_desc_list":[
]
}
Now you can run:
val data = parse(tmp).extract[AllMeasuresData]
val names = data.measures.map(_.name)
println(names)
// Displays
// List(_COUNT_, UU, CONT_SIZE, CONT_COUNT)

Answer 2

Your Parameter class does not match the JSON because you have used type1 rather than type as the field name. Use backticks to use "type" as a field name even though it is a reserved word:
case class Parameter (
`type`: String,
value: String
)
You also need to change the Function class as it has returntype1 rather than returntype:
case class Function (
expression: String,
parameter: Parameter,
returntype: String
)
The names of the fields in Scala must exactly match the names of the fields in the JSON. Extra fields in the JSON are ignored, but all the fields in the Scala must have matching fields in the JSON. If there are optional fields in the JSON then the Scala field type should be Option[...].