Query only those documents where image field is not empty - elasticsearch

Question

I have the following mapping **(dynamic strict on the type)**
"created": {
"type": "date"
},
"images": {
"properties": {
"checksum": {
"type": "text",
"index": false
},
"path": {
"type": "text",
"index": false
},
"url": {
"type": "text",
"index": false
}
}
},
I want to query documents where there is a image present
I tried couple of combinations but no luck so far.
This is the last i tried
POST catalog/_search
{
"query": {
"script": {
"script": "doc['images'].values.length > 0"
}
}
}
POST catalog/_search
{
"query": {
"script": {
"script": "doc['images.url'].values.length > 0"
}
}
}
But here it says that field data is not true for text fields. Is there anyway I can do this without changing my mapping.
Ideally this should give me all the records where there is no images. But this is returning all records
POST catalog/_search
{
"query": {
"bool": {
"must_not": [
{
"exists": {
"field": "images"
}
}
]
}
}
}
Here is the example document in which there is a image.
{
"_index": "catalog-2018-03-03",
"_type": "product",
"_id": "151755703145e27e4983a0bd1b70be44",
"_score": 1,
"_source": {
"merchant": {
"link": "http://shophive.com/",
"name": "shophive"
},
"images": [],
"updated": "2018-03-18T13:06:33.583480",
"name": "Plantronics Savi Talk",
"created": "2018-03-18T13:06:33.583459",
"url": "http://www.shophive.com/plantronics-savi-talk",
"price": {
"new": 24999,
"old": 24999,
"discount_percent": 0
},
"category": {
"level_1": {
"url": "computers/tablets/networking",
"name": "Computers/Tablets & Networking "
},
"level_2": {
"url": "tablets/ebook-readers",
"name": "Tablets & eBook Readers"
}
}
}
}
Updated
With the below query I am expecting that elasticsearch would return the documents in which image is missing
POST catalog/product/_search
{
"query": {
"bool": {
"must_not": [
{
"exists": {
"field": "images"
}
}
]
}
}
}
But the result i receive is all the documents in my index and apparently every document has one image. Here is the example document i get with above query
{
"took": 1,
"timed_out": false,
"_shards": {
"total": 5,
"successful": 5,
"failed": 0
},
"hits": {
"total": 20967,
"max_score": 1,
"hits": [
{
"_index": "catalog-2018-03-03",
"_type": "product",
"_id": "151755703145e27e4983a0bd1b70be44",
"_score": 1,
"_source": {
"merchant": {
"link": "http://shophive.com/",
"name": "shophive"
},
"images": [
{
"url": "http://www.shophive.com/media/catalog/product/cache/1/small_image/165x/9df78eab33525d08d6e5fb8d27136e95/p/l/plantronics_savi_talk.jpg",
"path": "full/8e3587bd2b6107f0beafa9b1ba05f476539be0a8.jpg",
"checksum": "fa74ade23c8e80e9590d48d4e59b6b64"
}
],
"updated": "2018-03-18T13:06:33.583480",
"name": "Plantronics Savi Talk",
"created": "2018-03-18T13:06:33.583459",
"url": "http://www.shophive.com/plantronics-savi-talk",
"price": {
"new": 24999,
"old": 24999,
"discount_percent": 0
},
"category": {
"level_1": {
"url": "computers/tablets/networking",
"name": "Computers/Tablets & Networking "
},
"level_2": {
"url": "tablets/ebook-readers",
"name": "Tablets & eBook Readers"
}
}
}
}
}
}

Answer 1

You should leave out the the square brackets in the query as you only have one clause
POST /catalog/_search
{
"query": {
"bool": {
"must_not": {
"exists": {
"field": "images"
}
}
}
}
}
This returns the docs with out images for me and if you need only those that have images
POST /catalog/_search
{
"query": {
"exists": {
"field": "images"
}
}
}