Version: v5.2

Normalization script

Create a normalization script to take raw IoT data and output the data in the data structure for readings on the platform.

Output data structure

Your normalization script must transform the raw IoT data to the following format:

{​
  "readings": [​
    {​
      // IoT reading data goes here​
      "_ts": "",​
      "_tsMetadata": {​
        "_sourceId": ""​
      }​
    }​
  ]​
}

Property	Type	Description	Required
`"readings"`	Array of Objects	Contains an array of reading objects, which also contain a timestamp and metadata.	Required
`"_ts"`	String	A timestamp you create	Required
`"_tsMetadata"`	Object	Create an object that contains a `"_sourceId"` property.	Required
`"_sourceId"`	String	Contains the source id of the sensor that takes the reading.	Required

Input and output data examples

Raw IoT input data

{
  "type": "text",
  "messageId": "ID:activemq-39563-1689584032552-5:4:-1:1:1",
  "correlationId": null,
  "destination": {
    "type": "queue",
    "name": "Consumer.<ANY_STRING>.VirtualTopic.lnt_RfmbfrED"
  },
  "replyTo": null,
  "priority": 4,
  "expiration": 0,
  "timestamp": 1689584797284,
  "redelivered": false,
  "properties": {
    "ActiveMQ.MQTT.QoS": {
      "type": "integer",
      "boolean": null,
      "byte": null,
      "short": null,
      "integer": 0,
      "long": null,
      "float": null,
      "double": null,
      "string": null
    }
  },
  "payloadText": "{\"from\":\"BLE\",\"mac\":\"30:ae:7b:e1:f5:28\",\"to\":\"GATEWAY\",\"time\":1689584797,\"deviceCode\":\"01012354-ad7d-4a0a-91db-a08435b7c9e3\", \"type\":\"reportAttribute\",\"data\":{\"value\":{\"device_list\":[{\"data\":\"0201061AFF5B07050305730073005E27D80085018E04AA023A079B0C000011079ECADC240EE5A9E093F3A3B50100406E09094869626F75414952\",\"modelstr\":\"HibouAIRe\",\"scan_rssi\":-51,\"Board_id\":\"5d9f693f-6f95-4ea5-a610-5fe7bf061f40\",\"ble_addr\":\"E8:E5:ED:BE:5E:D9\",\"addr_type\":1,\"scan_time\":1684043612,\"CO2\":0,\"HibouType\":\"PM\",\"dev_name\":\"HibouAIR\",\"ALS\":115,\"PM1.0\":\"68.200\",\"PM2.5\":\"185.000\",\"Pressure\":\"1007.800\",\"PM10\":\"322.700\",\"Temperature\":\"21.600\",\"Humidity\":\"38.900\",\"VOC\":1166,\"connectable\":1}]},\"attribute\":\"mod.device_list\",\"mac\":\"30:ae:7b:e1:f5:28\"},\"count\":1,\"name\":\"kafka-telemetry-1_sXrwGTdXah_timeseries\",\"id\":\"1-1\",\"csv\":\"execution7.csv\"}"
}

Normalized data output

{
  "readings": [
    {
      "data": "0201061AFF5B07050305730073005E27D80085018E04AA023A079B0C000011079ECADC240EE5A9E093F3A3B50100406E09094869626F75414952",
      "modelstr": "HibouAIRe",
      "scan_rssi": -51,
      "Board_id": "5d9f693f-6f95-4ea5-a610-5fe7bf061f40",
      "ble_addr": "E8:E5:ED:BE:5E:D9",
      "addr_type": 1,
      "scan_time": 1684043612,
      "CO2": 0,
      "HibouType": "PM",
      "dev_name": "HibouAIR",
      "ALS": 115,
      "PM1.0": "68.200",
      "PM2.5": "185.000",
      "Pressure": "1007.800",
      "PM10": "322.700",
      "Temperature": "21.600",
      "Humidity": "38.900",
      "VOC": 1166,
      "connectable": 1,
      "_ts": "2023-05-14T05:53:32.000Z",
      "_tsMetadata": {
        "_sourceId": "5d9f693f-6f95-4ea5-a610-5fe7bf061f40"
      }
    }
  ]
}

Creating a normalization script

The following procedure creates the script that transforms the raw IoT input data to the normalized output in the previous example:

Important: Only use the var keyword to create variables, do not use const or let.

Important: Do not use console logging functions in your script.

Note: You can use arrow functions.

To skip to the final script, see Script example.

Create a functions object to contain your constructReadings and normalize methods.
```
var functions = {

}
```
In the functions object, create a normalize method that takes a string message as its parameter.
```
...
  normalize: function(strMessage) {

  }
```

In the normalize method, parse the string message parameter.

  ...
    var message = JSON.parse(strMessage);​
    var payloadText = message.payloadText;​
    var payload = JSON.parse(payloadText);​
    var data = payload.data.value.device_list;

Create a readings array to push normalized readings to.
```
    ...
    var readings = [];
```

Construct a readings object for each device using a constructReadings method you define later, then push the object to the readings array.

    ...
    data.forEach(d => {
        var readingsObj = {...d, ...this.constructReadings(d.Board_id, d.scan_time)};
        readings.push(readingsObj)
    });

Declare a constructReadings method that takes the reading's source id and timestamp as parameters.
```
...
  constructReadings: (sourceId, ts) => {

  }
```
In the constructReadings method, create a new Date object based on the timestamp parameter multiplied by 1000.
```
  ...
    var date = new Date(ts*1000);
```

In the constructReadings function, construct a readings object with "_ts", "_tsMetadata", and "_tsMetadata._sourceId" properties, then return that object.

  ...
    var readingsObj = {
      "_ts": date.toISOString(),
      "_tsMetadata":{
        "_sourceId":sourceId
      }
    };
    return readingsObj;

Script example

var functions = {
  normalize: function(strMessage){
    var message = JSON.parse(strMessage);
    var payloadText=message.payloadText;
    var payload = JSON.parse(payloadText);
    if (typeof payload === 'string') { 
        payload = JSON.parse(payload);
    }
    var data = payload.data.value.device_list;
    

    var readings = [];
  
    data.forEach(d => {
        var readingsObj = {...d, ...this.constructReadings(d.Board_id, d.scan_time)};
        readings.push(readingsObj)
    });

    var result = {readings};
    return JSON.stringify(result);
  },
  constructReadings: (sourceId, ts) => {
    var date = new Date(ts*1000);
    var readingsObj = {
      "_ts": date.toISOString(),
      "_tsMetadata":{
          "_sourceId":sourceId
      }
    };
    return readingsObj;
  }
}