Dynamically Parsing JSON with Groovy - json

Question

I have a JSON document pulled back from a support system API. With my code, I want to pull out the pre-configured fields dynamically, presuming that the JSON may have more or fewer of the desired fields when my program calls the API.
I have some code that works, though it seems very convoluted and inefficient.
Here is a snippet of the pieces of JSON that I'm interested in:
{
"rows": [
{
"assignee_id": 1,
"created": "2017-01-25T14:13:19Z",
"custom_fields": [],
"fields": [],
"group_id": 2468,
"priority": "Low",
"requester_id": 2,
"status": "Open",
"subject": "Support request",
"ticket": {
"description": "Ticket descritpion",
"id": 1000,
"last_comment": {
"author_id": 2,
"body": "Arbitrary text",
"created_at": "2017-02-09T14:21:38Z",
"public": false
},
"priority": "low",
"status": "open",
"subject": "Support request",
"type": "incident",
"url": "Arbitrary URL"
},
"updated": "2017-02-09T14:21:38Z",
"updated_by_type": "Agent"
},
{
"assignee_id": 1,
"created": "2017-02-09T14:00:18Z",
"custom_fields": [],
"fields": [],
"group_id": 3579,
"priority": "Normal",
"requester_id": 15,
"status": "Open",
"subject": "Change request",
"ticket": {
"description": "I want to change this...",
"id": 1001,
"last_comment": {
"author_id": 20,
"body": "I want to change the CSS on my website",
"created_at": "2017-02-09T14:12:12Z",
"public": true
},
"priority": "normal",
"status": "open",
"subject": "Change request",
"type": "incident",
"url": "Arbitrary URL"
},
"updated": "2017-02-09T14:12:12Z",
"updated_by_type": "Agent"
}
]
}
I have an ArrayList called wantedFields that I build up from a config to define which information I want to pull out from the JSON:
["id","subject","requester_id","status","priority","updated","url"]
The complexity is that data is replicated in the API, and I only want to pull out data once, with a preference for the data in "rows" where applicable. My method for doing this is below. It feels like I'm repeating code but I can't really see how to make this work more efficiently. The JSON is held as "viewAsJson".
def ArrayList<Map<String,Object>> assignConfiguredFields(viewAsJson, wantedFields) {
//Pull out configured fields from JSON and store as Map to write as CSV later
ArrayList<Map<String,Object>> listOfDataToWrite = new ArrayList<Map<String,Object>>()
ArrayList<String> rowKeyList = new ArrayList<String>()
def validationRow = viewAsJson.rows.get(0)
//Compare one row object to config first
validationRow.each { k, v ->
if (wantedFields.contains(k)) {
wantedFields.remove(k)
rowKeyList.add(k)
}
}
ArrayList<String> ticketKeyList = new ArrayList<String>()
def validationTicket = viewAsJson.rows.ticket.get(0)
//Compare one ticket object to config first
validationTicket.each { k, v ->
if (wantedFields.contains(k)) {
wantedFields.remove(k)
ticketKeyList.add(k)
}
}
def rows = viewAsJson.rows
def tickets = viewAsJson.rows.ticket
//Pull matching ticket objects from JSON and store in Map
ArrayList<Map<String,Object>> tickList= new ArrayList<>()
ArrayList<Map<String,Object>> rowList= new ArrayList<>()
rows.each { row ->
Map<String,Object> rowMap = new HashMap<>()
row.each { k, v ->
if(rowKeyList.contains(k))
rowMap.put(k,v)
}
rowList.add(rowMap)
}
tickets.each { ticket ->
Map<String,Object> ticketMap = new HashMap<>()
ticket.each { k, v ->
if(ticketKeyList.contains(k))
ticketMap.put(k, v)
}
tickList.add(ticketMap)
}
for (int i = 0; i < rowList.size(); i++) {
HashMap<String,Object> dataMap = new HashMap<>()
dataMap.putAll(rowList.get(i))
dataMap.putAll(tickList.get(i))
listOfDataToWrite.add(dataMap)
}
println listOfDataToWrite
return listOfDataToWrite
}
I know there should be some validation for if the wantedFields ArrayList is still populated. I've iterated on this code so many times I just forgot to re-add that this time.

Answer 1

I don't know if you still need this code but why not try something like this.
Have a translation map and run each row through it.
Object tranverseMapForValue(Map source, String keysToTranverse, Integer location = 0){
List keysToTranverseList = keysToTranverse.split(/\./)
tranverseMapForValue(source, keysToTranverseList, location)
}
Object tranverseMapForValue(Map source, List keysToTranverse, Integer location = 0){
if(source.isEmpty() || keysToTranverse.isEmpty()){
return null
}
String key = keysToTranverse[location]
if(source[key] instanceof Map){
return tranverseMapForValue(source[key], keysToTranverse, location + 1)
}
else{
return source[key]
}
}
Map translation = [
"ticket.id": "id",
"ticket.subject": "subject",
"requester_id": "requester_id",
"ticket.status": "status",
"priority": "priority",
"updated": "updated",
"ticket.url": "url"
]
List rows = []
json.rows.each{ row ->
Map mapForRow = [:]
translation.each{ sourceKey, newKey ->
mapForRow << [(newKey): tranverseMapForValue(row, sourceKey)]
}
rows.add(mapForRow)
}