Elasticsearch - Search and filter by price - elasticsearch

Question

I have some array with promocodes (comes from request):
$promocodes = ['K1H5E1F1', 'M4C8A5K6', 'A3B9A45KL'];
And I have products data in Elasticsearch (as example, I will give data of one product):
// First product (2 promocodes matched, take a lower price 265.5 and filter this product at this price)
"price": 199,
"promocodes" : [
{
"code" : "K1H5E1F1",
"price" : 265.5
},
{
"code" : "LKDS3534K",
"price" : 357
},
{
"code" : "A3B9A45KL",
"price" : 327.5
}
]
// Second product (1 promocode matched, take a price 700 and filter this product at this price)
"price": 800,
"promocodes" : [
{
"code" : "AJ543HJB",
"price" : 500
},
{
"code" : "M4C8A5K6",
"price" : 700
}
]
// Third product (0 promocode matched, take a base price 900 and filter this product at this price)
"price": 900,
"promocodes" : [
{
"code" : "AJ87HJ90",
"price" : 750
}
]
I need to filter products data by price based on promocodes. If you set a range for the price and have promocodes, then you need to filter the products. If the product has the same promocode, then you need to take the price for this promotional code, not the main price. If 2 promocodes match for one product, then you need to take a lower price. In my example, the same product has 2 promotional codes for one product, I need to take the lower price out of 2 prices for the promocode and filter for that particular price.
This request does not filter prices as I need:
GET dev_products/_search
{
"query": {
"bool": {
"must": [
{
"range": {
"price": {
"gte": 100,
"lte": 350
}
}
},
{
"nested": {
"path": "promocodes",
"query": {
"terms": {
"promocodes.code": [
'K1H5E1F1',
'M4C8A5K6',
'A3B9A45KL'
]
}
}
}
}
]
}
}
}
I don't know how to make a request correctly, I ask you for help.

Answer 1

You need to use inner hits.
{
"query": {
"bool": {
"must": [
{
"range": {
"price": {
"gte": 100,
"lte": 350
}
}
},
{
"nested": {
"path": "promocodes",
"query": {
"terms": {
"promocodes.code": [
"K1H5E1F1",
"A3B9A45KL"
]
}
},
"inner_hits": {
"sort": {"promocodes.price": "asc"},----> sort nested document by price
"size": 1 ---> return top 1 document
}
}
}
]
}
}
}
Result:
"hits" : [
{
"_index" : "index4",
"_type" : "_doc",
"_id" : "NTBFgm0BFLPFo7KPt70j",
"_score" : 2.0,
"_source" : {
"price" : 199,
"promocodes" : [
{
"code" : "K1H5E1F1",
"price" : 265.5
},
{
"code" : "LKDS3534K",
"price" : 357
},
{
"code" : "A3B9A45KL",
"price" : 327.5
}
]
},
"inner_hits" : { -----> inner hits contains nested data
"promocodes" : {
"hits" : {
"total" : {
"value" : 2,
"relation" : "eq"
},
"max_score" : null,
"hits" : [ -----> returns one matched field
{
"_index" : "index4",
"_type" : "_doc",
"_id" : "NTBFgm0BFLPFo7KPt70j",
"_nested" : {
"field" : "promocodes",
"offset" : 0
},
"_score" : null,
"_source" : {
"code" : "K1H5E1F1",
"price" : 265.5
},
"sort" : [
265.5
]
}
]
}
}
}
}
]
EDIT:
Below logic checks if promocode has match then return document with promocode value in innerhits. If promocode has no match and parent price is in range(gte and lte value) then return that document.
GET dev_products/_search
{
"_source": "price",
"query": {
"bool": {
"should": [
{
"bool": {
"must": [
{
"range": {
"price": {
"gte": 100,
"lte": 350
}
}
}
],
"must_not": [
{
"nested": {
"path": "promocodes",
"query": {
"bool": {
"must": [
{
"terms": {
"promocodes.code.keyword": [
"K1H5E1F1",
"A3B9A45KL"
]
}
}
]
}
},
"inner_hits": {
"sort": {
"promocodes.price": "asc"
},
"size": 1
}
}
}
]
}
},
{
"nested": {
"path": "promocodes",
"query": {
"bool": {
"must": [
{
"terms": {
"promocodes.code.keyword": [
"K1H5E1F1",
"A3B9A45KL"
]
}
},
{
"range": {
"promocodes.price": {
"gte": 100,
"lte": 350
}
}
}
]
}
},
"inner_hits": {
"sort": {
"promocodes.price": "asc"
},
"size": 1
}
}
}
]
}
}
}
EDIT-2
Query
GET dev_products/_search
{
"_source": "price",
"query": {
"bool": {
"should": [
{
"bool": {
"must": [
{
"range": {
"price": {
"gte": 100,
"lte": 350
}
}
}
],
"must_not": [
{
"nested": {
"path": "promocodes",
"query": {
"bool": {
"must": [
{
"terms": {
"promocodes.code.keyword": [
"K1H5E1F1",
"A3B9A45KL"
]
}
}
]
}
}
}
}
]
}
},
{
"bool": {
"must": [
{
"nested": {
"path": "promocodes",
"query": {
"bool": {
"must": [
{
"terms": {
"promocodes.code.keyword": [
"K1H5E1F1",
"A3B9A45KL"
]
}
},
{
"range": {
"promocodes.price": {
"lte": 350,
"gte": 100
}
}
}
]
}
},
"inner_hits": {
"sort": {
"promocodes.price": "asc"
},
"size": 1
}
}
}
],
----> don't include document if any matched promcode has value less than given range
"must_not": [
{
"nested": {
"path": "promocodes",
"query": {
"bool": {
"must": [
{
"terms": {
"promocodes.code.keyword": [
"K1H5E1F1",
"A3B9A45KL"
]
}
},
{
"range": {
"promocodes.price": {
"lt": 100
}
}
}
]
}
}
}
}
]
}
}
]
}
}
}

Answer 2

If the range for price and promocodes.price is equal to "gte":270,"lte":271 and in terms of promocodes.code is equal to ["promo1","promo2","promo4"] the request does not work - in fact, he should not choose this product, since the price according to the lowest promotional code is 265.5 and does not fall into the diapason of range, but he still selects this product and does not add the desired Promocode for inner_hits (for some reason he chooses "promo2" for inner_hits with price 270).
"price": 275,
"promocodes" : [
{
"code" : "promo1",
"price" : 265.5
},
{
"code" : "promo2",
"price" : 270
},
{
"code" : "promo3",
"price" : 250
}
]