I am currently using ImportJSON to import Sendgrid Email with data Keenio Extraction Query API URL by calling the ImportJSON function in a Google Spreadsheet cell of Sheet DATA.
=ImportJSON("https://api.keen.io/3.0/projects/"& PROJECT_KEY & "/queries/extraction?api_key=" & API_KEY & "&event_collection=" & EVT_COL & "&timezone=" & TIMEZONE & "&latest=" & LATEST & "&property_names..........", PTDATA!$AB$1)
In Sheet PTDATA, in the last column cell i am setting a random number for ImportJSON to recalculate. The function runs on Spreadsheet open event. I have also added a custom menu to call the ReCalcCell custom function.
function onOpen() {
var ui = SpreadsheetApp.getUi();
// Or DocumentApp or FormApp.
ui.createMenu('IMPORT DATA')
.addItem('KEENIO DATA', 'ReCalcCell')
.addToUi();
}
function ReCalcCell(){
var min = Math.ceil(0);
var max = Math.floor(9999);
var randomNum = Math.floor(Math.random() * (max - min + 1)) + min
var ss = SpreadsheetApp.getActiveSpreadsheet();
var sh = ss.getSheetByName("PTDATA");
sh.getRange("$AB$1").setValue(randomNum);
}
PTDATA sheet has specific column header names for which i want to pull the data from DATA sheet. Towards the right of these columns, i have other calculation columns which work on these specific columns.
Since the columns in DATA sheet always appear in a random / shuffled order, i had to write a small custom function GCL which takes in a header name and returns its datarange address from DATA sheet as a string.
function GCL(header,dummy) {
var ss = SpreadsheetApp.getActiveSpreadsheet();
var sheet = ss.getSheetByName("DATA");
var headings = sheet.getRange(1, 1, 1, sheet.getLastColumn()); // get the range representing the whole sheet
var width = headings.getWidth();
var lrow = sheet.getLastRow();
// search every cell in row 1 from A1 till the last column
for (var i = 1; i <= width; i++) {
var data = headings.getCell(1,i).getValue();
if (data == header) {
return ((sheet.getSheetName() + "!" + columnToLetter(i)+"2:" + columnToLetter(i) + lrow).toString()); // return the column range if we find it
break; // exit when found
}
}
return(-1); // return -1 if it doesn't exist
}
function columnToLetter(column)
{
var temp, letter = '';
while (column > 0)
{
temp = (column - 1) % 26;
letter = String.fromCharCode(temp + 65) + letter;
column = (column - temp - 1) / 26;
}
return letter;
}
Then i use the custom function GCL in each specific column to get it's datarange. Once data is populated, the PDATA sheet is used to create different Pivots for reporting purposes.
=ARRAYFORMULA(INDIRECT(GCL(A1,$AB$1)))
The problems i am facing is that though the ImportJSON data populates the DATA sheet:
- The columns appear shuffled everytime, so my calculation columns cannot calculate as the references go away. This renders the pivots useless! To counter this issue, i had to create the PDATA sheet to pull in specific columns using the custom function GCL.
- The custom function GCL does not always refresh and most of the time shows #Ref error.
BTW, my JSON output from Keenio looks like this:
{
"result":
[
{
"sg_event_id": "92-OndRfTs6fZjNdHWzLBw",
"timestamp": 1529618395,
"url": "https://noname.com?utm_campaign=website&utm_source=sendgrid.com&utm_medium=email",
"ip": "192.168.1.1",
"event": "click",
"keen": {
"timestamp": "2018-06-21T21:59:55.000Z",
"created_at": "2018-06-21T22:00:28.532Z",
"id": "555c1f7c5asdf7000167d87b"
},
"url_offset": {
"index": 38,
"type": "text"
},
"sg_message_id": "F5mwV1rESdyKFA_2bn1IEQ.filter0042p3las1-15933-5B2A68E8-36.0",
"useragent": "Mozilla/4.0 (compatible; MSIE 8.0; Windows NT 5.1; Trident/4.0)",
"email": "[email protected]"
}, {
"sg_event_id": "bjMlfsSfRyuXEVy8LndsYA",
"timestamp": 1529618349,
"url": "https://noname.com?utm_campaign=website&utm_source=sendgrid.com&utm_medium=email",
"ip": "192.168.1.1",
"event": "click",
"keen": {
"timestamp": "2018-06-21T21:59:09.000Z",
"created_at": "2018-06-21T21:59:39.491Z",
"id": "555c1f7c5asdf7000167d87b"
},
"url_offset": {
"index": 36,
"type": "text"
},
"sg_message_id": "F5mwV1rESdyKFA_2bn1IEQ.filter0042p3las1-15933-5B2A68E8-36.0",
"useragent": "Mozilla/4.0 (compatible; MSIE 8.0; Windows NT 5.1; Trident/4.0)",
"email": "[email protected]"
}, {
"sg_event_id": "fru_s2s1RtueuqBMNoIoTg",
"timestamp": 1529618255,
"url": "https://noname.com?utm_campaign=website&utm_source=sendgrid.com&utm_medium=email",
"ip": "192.168.1.1",
"event": "click",
"keen": {
"timestamp": "2018-06-21T21:57:35.000Z",
"created_at": "2018-06-21T21:58:20.374Z",
"id": "555c1f7c5asdf7000167d87b"
},
"url_offset": {
"index": 29,
"type": "text"
},
"sg_message_id": "F5mwV1rESdyKFA_2bn1IEQ.filter0042p3las1-15933-5B2A68E8-36.0",
"useragent": "Mozilla/4.0 (compatible; MSIE 8.0; Windows NT 5.1; Trident/4.0)",
"email": "[email protected]"
}
]
}
My questions are:
- Is there a way to parse the JSON result without use of ImportJSON, which has to be entered as a custom function in a cell that also depends on recalculation? ImportJSON sometimes doesn't work properly.
- How can this code be refactored or optimized so that it can always return data to PDATA sheet columns?
- Is there a better way of accomplishing what i want without resorting to custom functions like GCL in the PDATA Sheet or ImportJSON in DATA sheet?


JSON.parse(). Aboutit can always return data to PDATA sheet columns, what does "always" mean? Do you want to update the data using a time-driven trigger? Can you provide a sample spreadsheet? I think that it will help users think of your solution. Of course, please remove your private information. - Tanaike