nql-example.md

Nutritionix Querying Language (NXQL)

NXQL allows developers to write more advanced queries against our data pool. Examples of what you can do with an advanced query:

• Filter items by their type (show ONLY USDA items or ONLY CPG items)
• Filter nutrient ranges (>100 calories AND <500 calories)

All requests are POST and require a valid JSON object be sent to https://api.nutritionix.com/v1_1/search You can paste any of the below JSON examples into this curl request and they will retrieve results assuming you have added your appKey, and appId.

Request Requirements

• Must provide valid appId and appKey
• Must provide filters, queries, or query properties in your JSON object
• Must set Content-Type to application/json

curl -XPOST https://api.nutritionix.com/v1_1/search -H 'Content-Type: application/json' -d'
{
 "appId":"YOUR_API_ID",
 "appKey":"YOUR_API_KEY",
 "query":"Cookies `n Cream"
}
'

Filtering Item Type and sorting by score

This example will yield only CPG items Item Types(item_type)

• 1 = Restaurant
• 2 = CPG (packaged foods with barcodes)
• 3 = USDA

/**
  *
  * Ensure you pass your API credentials
  * Select the fields you want to return
  * Set the sorting mechanism
  * Apply some filters to the data
  *
  **/

{
  "appId":"YOUR_APP_ID",
  "appKey":"YOU_APP_KEY",  
  "fields":["item_name","brand_name","upc"],
  "sort":{
    "field":"_score",
    "order":"desc"
  },
  "filters":{
    "item_type":2
  }
}

or you want all the item_types except restaurant items

{
  "appId":"YOUR_APP_ID",
  "appKey":"YOU_APP_KEY",
  "filters":{
    "not":{
     "item_type":2
    }
  }
}

Range Filters

You can filter numeric_fields using several operators

• gt
• gte
• lt
• lte

Or

• from
• to

{
 "appId":"YOUR_APP_ID",
 "appKey":"YOU_APP_KEY",
 "filters"{
  "nf_calories":{
   "from":100,
   "to":50
  },
  "nf_sodium":{
   "lte":200
  }
 }
}

Nutritionix Search

You can use the default search where we apply boosting, and multi search factors to yield the most relevant results to your users.

{
  "appId":"YOUR_APP_ID",
  "appKey":"YOU_APP_KEY",
  "query":"Starbucks Frappuccino"
}

Wild Card Searches

This query will yield starbucks frappuccino results

{
  "appId":"YOUR_APP_ID",
  "appKey":"YOU_APP_KEY",
  "query":"Star* Frap*"
}

Different Boolean Search Operators

This query will yield results that have starbucks or frappuccino Available Operators

• OR
• AND (Default)

{
  "appId":"YOUR_APP_ID",
  "appKey":"YOU_APP_KEY",
  "query":"Starbucks OR Frappuccino*"
}

Custom Multi Field Search

If you want to search specific words across multiple fields

{
  "appId":"YOUR_APP_ID",
  "appKey":"YOU_APP_KEY",
  "queries":{
    "item_name":"Kids Fries",
    "brand_name":"McDonalds"
  }
}

Filter to a specific brand, and paginate the results

On the API /:version/search GET endpoint we have the results default sorting by _score. Document scores can very sometimes, so to keep your search results consistent you can sort on a field that doesn't change.

{
  "appId":"YOUR_APP_ID",
  "appKey":"YOUR_APP_KEY",
  "fields":["item_name","brand_name","nf_calories","nf_sodium","item_type"],
  "offset":250,
  "limit":50,
  "sort":{
    "field":"item_name.sortable_na",
    "order":"asc"
  },
  "filters":{
   "brand_id":"513fbc1283aa2dc80c00001a"
  }
}

Sorting by item_name or brand_name

{
  "appId": "YOUR_APP_ID",
  "appKey": "YOUR_APP_KEY",
  "query": "kraft",
  "fields": [
    "item_name",
    "brand_name"
  ],
  "sort": {
    "field": "item_name.sortable_na",
    "order": "asc"
  }
}

NOTE .sortable_na This field is the original name, untouched by our texted analyzers. Thus sortable_na (not analyzed)

• item_name.sortable_na
• brand_name.sortable_na

Advanced Query Full Example

Lets get really fancy with it Here is the breakdown for this search

• appId
• appKey
• fields (an array of item attributes you want to return)
• offset (The starting point in the result set for paging)
• limit (The max number of results to be returned in the page)
• sort (the field who's numerical value you want to sort by in descending order)
• min_score (the minimum score that you want any returned document to have)
• query (A phrase with special operators OR ~ *)
• filters.not.item_type:2 (ensuring that we don't pay any attention to CPG data)
• filters.nf_calories (a range filter to and from a certain numerical range)
• filters.nf_sodium (a boolean filter for less than)

{
  "appId": "YOUR_APP_ID",
  "appKey": "YOUR_APP_KEY",
  "fields": [
    "item_name",
    "brand_name",
    "nf_calories",
    "nf_sodium",
    "item_type"
  ],
  "offset": 0,
  "limit": 50,
  "sort": {
    "field": "nf_calories",
    "order": "desc"
  },
  "min_score": 0.5,
  "query": "starbucks AND frap*",
  "filters": {
    "not": {
      "item_type": 2
    },
    "nf_calories": {
      "from": 0,
      "to": 400
    }
  }
}

This search yield these results.

{
    "total": 3,
    "max_score": null,
    "hits": [
        {
            "_index": "4qnk4vry5gujjxptu5rg",
            "_type": "item",
            "_id": "513fc9c6673c4fbc26001dca",
            "_score": null,
            "fields": {
                "nf_sodium": 100,
                "item_name": "Starbucks Coffee Frappuccino (Yield: 1 bottle)",
                "brand_name": "Gandolfo's New York Delicatessen",
                "nf_calories": 200,
                "item_type": 1
            },
            "sort": [
                200
            ]
        },
        {
            "_index": "4qnk4vry5gujjxptu5rg",
            "_type": "item",
            "_id": "513fc9c6673c4fbc26001dcd",
            "_score": null,
            "fields": {
                "nf_sodium": 100,
                "item_name": "Starbucks Vanilla Frappuccino (Yield: 1 bottle)",
                "brand_name": "Gandolfo's New York Delicatessen",
                "nf_calories": 200,
                "item_type": 1
            },
            "sort": [
                200
            ]
        },
        {
            "_index": "4qnk4vry5gujjxptu5rg",
            "_type": "item",
            "_id": "513fc9c6673c4fbc26001dcc",
            "_score": null,
            "fields": {
                "nf_sodium": 95,
                "item_name": "Starbucks Mocha Frappuccino (Yield: 1 bottle)",
                "brand_name": "Gandolfo's New York Delicatessen",
                "nf_calories": 180,
                "item_type": 1
            },
            "sort": [
                180
            ]
        }
    ]
}

Multi-Field Sorting

Uses the same syntax for sort, but puts them in an array. This will not work properly if each object does not contain field and order properties.

{
  "fields":["_score","item_name","brand_name","item_type"],
  "sort":[
    {
      "field":"item_type",
      "order":"desc"
    },
    {
      "field" : "_score",
      "order" : "asc"
    }
  ],
  "query":"turkey"
}