I want to write Elasticsearch query where I am able to look into a nested array, and check a key for all those elements inside the array, if all of them have that value then the root document should be selected.
this is a sample data :
[
{
"_index": "allproperties",
"_type": "_doc",
"_id": "5241050",
"_score": 0.0,
"_source": {
"id": 5241050,
"type": "HOUSE",
"state": "NSW",
"agency": {
"id": 31,
"code": "AU_LNT",
"name": "Luthor Properties",
"config": "{}"
},
"branch": {
"id": 89,
"name": "Luthor Sales Unit Trust",
"tradingName": "Luthor Properties",
"internalCode": "AU_LNT_G"
},
"images": null,
"leases": [
{
"id": 26439,
"bond": 2000,
"leased": true,
"source": "PORTAL",
"status": "SIGNED",
"endDate": "2022-11-06T00:00:00",
"tenants": [
{
"id": 11106,
"role": "TENANT",
"user": {
"id": 38817,
"dob": "1900-01-01",
"name": "Liam",
"email": "[email protected]",
"phone": "+61400000000",
"surname": "Tempo",
"salutation": "Mr.",
"invitations": null,
"lastInAppActivity": null
},
"address": "45675 Bruce Hwy, Coolbie QLD 4850, Australia",
"extendedData": "{\"rentSplitAmount\":22.62}",
"roleAdjective": "PRIMARY",
"addressComponents": {
"state": "QLD",
"suburb": "Coolbie",
"country": "Australia",
"postcode": "4850",
"streetName": "Bruce Hwy",
"unitNumber": null,
"poboxNumber": null,
"streetNumber": "45675",
"addressComponentsUnavailable": null
}
}
],
"signDate": "2022-10-05T04:58:08.887",
"startDate": "2022-10-06T00:00:00",
"moveInDate": "2022-10-06T00:00:00",
"appointments": [
{
"id": 3506,
"type": "REMOTE",
"date_time": "2022-10-12T04:56:00+00:00",
"createdBy": "Lex Luthor",
"createdDate": "2022-10-05T04:56:52.936",
"lastModifiedBy": "Lex Luthor",
"lastModifiedDate": "2022-10-05T04:56:53.51"
}
],
"createdDate": "2022-10-05T04:55:42.247",
"rentalAmount": 500,
"dateAvailable": null,
"extendedData": {
"last_rent_adjustment_date": 1665014400000
},
"leaseDuration": 32,
"rentalFrequency": "2",
"signedManually": false
}
],
"refId": "b66326eb-a6b2-42b6-b058-b46847e13399",
"source": "PT",
"status": null,
"suburb": "Hurstville",
"address": "456 Forest Road Hurstville NSW AUSTRALIA 2220",
"country": "Australia",
"bedrooms": 3,
"category": null,
"pmsCode": "84c34d15-a0ab-4791-b9e3-1bdac215b99c",
"postcode": "2220",
"agencyId": 31,
"bathrooms": 3,
"branchId": 89,
"carspaces": 3,
"createdBy": "system",
"streetname": "Forest Road",
"unitnumber": null,
"description": null,
"createdDate": "2022-10-05T04:51:12.619",
"entitlements": [
{
"id": 3453799,
"role": "LANDLORD",
"user": {
"id": 22855,
"dob": null,
"name": "please enter ownership name",
"email": "[email protected]",
"phone": "0400000000",
"surname": "",
"salutation": null,
"lastInAppActivity": null
},
"company": {
"id": 137,
"abn": "",
"acn": "",
"phone": "0400000000",
"address": "1234 Park Avenue New York NY USA 10037-1702",
"companyName": "please enter ownership name",
"displayName": "please enter ownership name",
"email": "[email protected]"
},
"roleAdjective": "GROUP"
},
{
"id": 3453800,
"role": "AGENT",
"user": {
"id": 20054,
"dob": null,
"name": "Paul",
"email": "[email protected]",
"phone": "+61403084232",
"surname": "Botti",
"salutation": null,
"lastInAppActivity": null
},
"company": null,
"roleAdjective": "MANAGING"
},
{
"id": 3453801,
"role": "AGENT",
"user": {
"id": 20054,
"dob": null,
"name": "Paul",
"email": "[email protected]",
"phone": "+61403084232",
"surname": "Botti",
"salutation": null,
"lastInAppActivity": null
},
"company": null,
"roleAdjective": "LEASING"
}
],
"streetnumber": "456",
"advertised": false,
"commission_type": null,
"lastModifiedBy": "system",
"lastModifiedDate": "2022-10-05T04:58:09.043",
"_meta": {
"branches": {
"id": [
89
]
},
"agencies": {
"id": [
31
]
},
"user_property_entitlement": {
"id": [
3453799,
3453800,
3453801
]
},
"users": {
"id": [
20054,
22855,
38817
]
},
"companies": {
"id": [
137
]
},
"leases": {
"id": [
26439
]
},
"user_lease_entitlement": {
"id": [
11106
]
},
"appointments": {
"id": [
3506
]
}
}
},
"sort": [
1664945472619,
0.0
]
}
]
In this particular case, leases is a nested document, I want to select the documents where all the leases are in cancelled state or leases is null, i.e; either all of the objects inside leases should have status as CANCELLED or the leases key should be null. Already went through this question, but did'nt quite get it as its an old answer from 2015 and methods used in this got deprecated.
Try this query:
GET idx_test/_search?filter_path=hits.hits
{
"query": {
"bool": {
"must_not": [
{
"nested": {
"path": "leases",
"query": {
"bool": {
"must_not": [
{
"bool": {
"should": [
{
"term": {
"leases.status.keyword": "CANCELLED"
}
},
{
"bool": {
"must_not": [
{
"exists": {
"field": "leases"
}
}
]
}
}
]
}
}
]
}
}
}
}
]
}
}
}