Since you filter by jobIds, then you don't have to filter in script,
{
"query": {
"function_score": {
"query": {
"match_all": {}
},
"functions": [
{
"filter": {
"terms": {
"_id": [
6,
22,
98,
16
]
}
},
"script_score": {
"script": {
"source": "9999 - Integer.parseInt(doc['_id'].value)"
}
}
},
{
"filter": {
"match_all": {}
},
"script_score": {
"script": {
"source": "return 0"
}
}
}
],
"score_mode": "sum"
}
}
}
If I understand your question correctly, I believe you are looking for something like this.
It looks a bit wacky but it roughly translates to:
"Find me (Vanaraj, 27 years old, in India) or (Ranjit, who is 26 in US)"
# Setup a sample index with data
PUT /stackoverflowtest/_doc/1
{ "Name" : "Vanaraj", "Age" : "27", "country" : "India" }
PUT /stackoverflowtest/_doc/2
{ "Name" : "Ranjit", "Age" : "26", "country" : "US" }
PUT /stackoverflowtest/_doc/3
{ "Name" : "Joe", "Age" : "38", "country" : "US" }
# Query that returns Vanaraj / 27 / India or Ranjit / 26 / US
POST /stackoverflowtest/_search
{
"query": {
"bool": {
"should": [
{
"bool": {
"must": [
{
"match": {
"Name": "Vanaraj"
}
},
{
"match": {
"Age": "27"
}
},
{
"match": {
"country": "India"
}
}
]
}
},
{
"bool": {
"must": [
{
"match": {
"Name": "Ranjit"
}
},
{
"match": {
"Age": "26"
}
},
{
"match": {
"country": "US"
}
}
]
}
}
]
}
}
}
