Google Apps Script: A Beginner’s Guide

What is Google Apps Script?

Google Apps Script is a cloud-based scripting language for extending the functionality of Google Apps and building lightweight cloud-based applications.

It means you write small programs with Apps Script to extend the standard features of Google Workspace Apps. It’s great for filling in the gaps in your workflows.

With Apps Script, you can do cool stuff like automating repeatable tasks, creating documents, emailing people automatically and connecting your Google Sheets to other services you use.

Writing your first Google Script

In this Google Sheets script tutorial, we’re going to write a script that is bound to our Google Sheet. This is called a container-bound script.

(If you’re looking for more advanced examples and tutorials, check out the full list of Apps Script articles on my homepage.)

Hello World in Google Apps Script

Let’s write our first, extremely basic program, the classic “Hello world” program beloved of computer teaching departments the world over.

Begin by creating a new Google Sheet.

Then click the menu: Extensions > Apps Script

This will open a new tab in your browser, which is the Google Apps Script editor window:

Google Apps Script Editor

By default, it’ll open with a single Google Script file (code.gs) and a default code block, myFunction():

function myFunction() {
  
}

In the code window, between the curly braces after the function myFunction() syntax, write the following line of code so you have this in your code window:

function myFunction() {
  Browser.msgBox("Hello World!");
}

Your code window should now look like this:

Hello World Apps Script

Google Apps Script Authorization

Google Scripts have robust security protections to reduce risk from unverified apps, so we go through the authorization workflow when we first authorize our own apps.

When you hit the run button for the first time, you will be prompted to authorize the app to run:

Google Apps Script Authorization

Clicking Review Permissions pops up another window in turn, showing what permissions your app needs to run. In this instance the app wants to view and manage your spreadsheets in Google Drive, so click Allow (otherwise your script won’t be able to interact with your spreadsheet or do anything):

Google Apps Script Access

❗️When your first run your apps script, you may see the “app isn’t verified” screen and warnings about whether you want to continue.

In our case, since we are the creator of the app, we know it’s safe so we do want to continue. Furthermore, the apps script projects in this post are not intended to be published publicly for other users, so we don’t need to submit it to Google for review (although if you want to do that, here’s more information).

Click the “Advanced” button in the bottom left of the review permissions pop-up, and then click the “Go to Starter Script Code (unsafe)” at the bottom of the next screen to continue. Then type in the words “Continue” on the next screen, click Next, and finally review the permissions and click “ALLOW”, as shown in this image (showing a different script in the old editor):

More information can be found in this detailed blog post from Google Developer Expert Martin Hawksey.

Running a function in Apps Script

Once you’ve authorized the Google App script, the function will run (or execute).

If anything goes wrong with your code, this is the stage when you’d see a warning message (instead of the yellow message, you’ll get a red box with an error message in it).

Return to your Google Sheet and you should see the output of your program, a message box popup with the classic “Hello world!” message:

Message Box Google Sheets

Click on Ok to dismiss.

Great job! You’ve now written your first apps script program.

Rename functions in Google Apps Script

We should rename our function to something more meaningful.

At present, it’s called myFunction which is the default, generic name generated by Google. Every time I want to call this function (i.e. run it to do something) I would write myFunction(). This isn’t very descriptive, so let’s rename it to helloWorld(), which gives us some context.

So change your code in line 1 from this:

function myFunction() {
  Browser.msgBox("Hello World!");
}

to this:

function helloWorld() {
  Browser.msgBox("Hello World!");
}

Note, it’s convention in Apps Script to use the CamelCase naming convention, starting with a lowercase letter. Hence, we name our function helloWorld, with a lowercase h at the start of hello and an uppercase W at the start of World.

Adding a custom menu in Google Apps Script

In its current form, our program is pretty useless for many reasons, not least because we can only run it from the script editor window and not from our spreadsheet.

Let’s fix that by adding a custom menu to the menu bar of our spreadsheet so a user can run the script within the spreadsheet without needing to open up the editor window.

This is actually surprisingly easy to do, requiring only a few lines of code. Add the following 6 lines of code into the editor window, above the helloWorld() function we created above, as shown here:

function onOpen() {
  const ui = SpreadsheetApp.getUi();
  ui.createMenu('My Custom Menu')
      .addItem('Say Hello', 'helloWorld')
      .addToUi();
}

function helloWorld() {
  Browser.msgBox("Hello World!");
}

If you look back at your spreadsheet tab in the browser now, nothing will have changed. You won’t have the custom menu there yet. We need to re-open our spreadsheet (refresh it) or run our onOpen() script first, for the menu to show up.

To run onOpen() from the editor window, first select then run the onOpen function as shown in this image:

Google Apps Script Function Menu

Now, when you return to your spreadsheet you’ll see a new menu on the right side of the Help option, called My Custom Menu. Click on it and it’ll open up to show a choice to run your Hello World program:

Custom menu

Run functions from buttons in Google Sheets

An alternative way to run Google Scripts from your Sheets is to bind the function to a button in your Sheet.

For example, here’s an invoice template Sheet with a RESET button to clear out the contents:

Button with apps script in google sheets

For more information on how to do this, have a look at this post: Add A Google Sheets Button To Run Scripts

Google Apps Script Examples

Macros in Google Sheets

Another great way to get started with Google Scripts is by using Macros. Macros are small programs in your Google Sheets that you record so that you can re-use them (for example applying standard formatting to a table). They use Apps Script under the hood so it’s a great way to get started.

Read more: The Complete Guide to Simple Automation using Google Sheets Macros

Custom function using Google Apps Script

Let’s create a custom function with Apps Script, and also demonstrate the use of the Maps Service. We’ll be creating a small custom function that calculates the driving distance between two points, based on Google Maps Service driving estimates.

The goal is to be able to have two place-names in our spreadsheet, and type the new function in a new cell to get the distance, as follows:

GAS custom function for maps

The solution should be:

GAS custom map function output

Copy the following code into the Apps Script editor window and save. First time, you’ll need to run the script once from the editor window and click “Allow” to ensure the script can interact with your spreadsheet.

function distanceBetweenPoints(start_point, end_point) {
  // get the directions
  const directions = Maps.newDirectionFinder()
     .setOrigin(start_point)
     .setDestination(end_point)
     .setMode(Maps.DirectionFinder.Mode.DRIVING)
     .getDirections();
  
  // get the first route and return the distance
  const route = directions.routes[0];
  const distance = route.legs[0].distance.text;
  return distance;
}

Saving data with Google Apps Script

Let’s take a look at another simple use case for this Google Sheets Apps Script tutorial.

Suppose I want to save copy of some data at periodic intervals, like so:

save data in google sheet

In this script, I’ve created a custom menu to run my main function. The main function, saveData(), copies the top row of my spreadsheet (the live data) and pastes it to the next blank line below my current data range with the new timestamp, thereby “saving” a snapshot in time.

The code for this example is:

// custom menu function
function onOpen() {
  const ui = SpreadsheetApp.getUi();
  ui.createMenu('Custom Menu')
      .addItem('Save Data','saveData')
      .addToUi();
}

// function to save data
function saveData() {
  const ss = SpreadsheetApp.getActiveSpreadsheet();
  const sheet = ss.getSheets()[0];
  const url = sheet.getRange('Sheet1!A1').getValue();
  const follower_count = sheet.getRange('Sheet1!B1').getValue();
  const date = sheet.getRange('Sheet1!C1').getValue();
  sheet.appendRow([url,follower_count,date]);
}

See this post: How To Save Data In Google Sheets With Timestamps Using Apps Script, for a step-by-step guide to create and run this script.

Google Apps Script example in Google Docs

Google Apps Script is by no means confined to Sheets only and can be accessed from other Google Workspace tools.

Here’s a quick example in Google Docs, showing a script that inserts a specific symbol wherever your cursor is:

Google Docs Apps Script

We do this using Google App Scripts as follows:

1. Create a new Google Doc

2. Open script editor from the menu: Extensions > Apps Script

3. In the newly opened Script tab, remove all of the boilerplate code (the “myFunction” code block)

4. Copy in the following code:

// code to add the custom menu
function onOpen() {
  const ui = DocumentApp.getUi();
  ui.createMenu('My Custom Menu')
      .addItem('Insert Symbol', 'insertSymbol')
      .addToUi();
}

// code to insert the symbol
function insertSymbol() {  
  // add symbol at the cursor position
  const cursor = DocumentApp.getActiveDocument().getCursor();
  cursor.insertText('§§');
  
}

5. You can change the special character in this line

cursor.insertText('§§');

to whatever you want it to be, e.g.

cursor.insertText('( ͡° ͜ʖ ͡°)');

6. Click Save and give your script project a name (doesn’t affect the running so call it what you want e.g. Insert Symbol)

7. Run the script for the first time by clicking on the menu: Run > onOpen

8. Google will recognize the script is not yet authorized and ask you if you want to continue. Click Continue

9. Since this the first run of the script, Google Docs asks you to authorize the script (I called my script “test” which you can see below):

Docs Apps Script Auth

10. Click Allow

11. Return to your Google Doc now.

12. You’ll have a new menu option, so click on it:

My Custom Menu > Insert Symbol

13. Click on Insert Symbol and you should see the symbol inserted wherever your cursor is.

Google Apps Script Tip: Use the Logger class

Use the Logger class to output text messages to the log files, to help debug code.

The log files are shown automatically after the program has finished running, or by going to the Executions menu in the left sidebar menu options (the fourth symbol, under the clock symbol).

The syntax in its most basic form is Logger.log(something in here). This records the value(s) of variable(s) at different steps of your program.

For example, add this script to a code file your editor window:

function logTimeRightNow() {
  const timestamp = new Date();
  Logger.log(timestamp);
}

Run the script in the editor window and you should see:

Google Apps Script Execution Logs

Real world examples from my own work

I’ve only scratched the surface of what’s possible using G.A.S. to extend the Google Apps experience.

Here are a couple of interesting projects I’ve worked on:

1) A Sheets/web-app consisting of a custom web form that feeds data into a Google Sheet (including uploading images to Drive and showing thumbnails in the spreadsheet), then creates a PDF copy of the data in the spreadsheet and automatically emails it to the users. And with all the data in a master Google Sheet, it’s possible to perform data analysis, build dashboards showing data in real-time and share/collaborate with other users.

2) A dashboard that connects to a Google Analytics account, pulls in social media data, checks the website status and emails a summary screenshot as a PDF at the end of each day.

Marketing dashboard using Google Apps Script

3) A marking template that can send scores/feedback to students via email and Slack, with a single click from within Google Sheets. Read more in this article: Save time with this custom Google Sheets, Slack & Email integration

My own journey into Google Apps Script

My friend Julian, from Measure School, interviewed me in May 2017 about my journey into Apps Script and my thoughts on getting started:

Google Apps Script Resources

For further reading, I’ve created this list of resources for information and inspiration:

Course

Documentation

Official Google Documentation

Google Workspace Developers Blog

Communities

Google Apps Script Group

Stack Overflow GAS questions

Elsewhere On The Internet

A huge big up-to-date list of Apps Script resources hosted on GitHub.

For general Javascript questions, I recommend this JavaScript tutorial page from W3 Schools when you’re starting out.

When you’re more comfortable with Javascript basics, then I recommend the comprehensive JavaScript documentation from Mozilla.

Imagination and patience to learn are the only limits to what you can do and where you can go with GAS. I hope you feel inspired to try extending your Sheets and Docs and automate those boring, repetitive tasks!

Related Articles

Sheet Sizer! Build A Tool To Measure Your Google Sheets Size With Apps Script

In this tutorial, you’ll use Apps Script to build a tool called Sheet Sizer, to measure the size of your Google Sheets!

Google Sheets has a limit of 10 million cells, but it’s hard to know how much of this space you’ve used.

Sheet Sizer will calculate the size of your Sheet and compare it to the 10 million cell limit in Google Sheets.

Sheet Sizer

Sheet Sizer: Build a sidebar to display information

Step 1:
Open a blank sheet and rename it to “Sheet Sizer”

Step 2:
Open the IDE. Go to Tools > Script editor

Step 3:
Rename your script file “Sheet Sizer Code”

Step 4:
In the Code.gs file, delete the existing “myFunction” code and copy in this code:

/**
* Add custom menu to sheet
*/
function onOpen() {

  SpreadsheetApp.getUi()
    .createMenu('Sheet Sizer')
    .addItem('Open Sheet Sizer', 'showSidebar')
    .addToUi();
}

/**
* function to show sidebar
*/
function showSidebar() {
 
  // create sidebar with HTML Service
  const html = HtmlService.createHtmlOutputFromFile('Sidebar').setTitle('Sheet Sizer');
 
// add sidebar to spreadsheet UI
  SpreadsheetApp.getUi().showSidebar(html);
}

There are two functions: onOpen, which will add the custom menu to your Sheet, and showSidebar, which will open a sidebar.

The text between the /* ... */ or lines starting with // are comments.

Step 5:
Click the + next to Files in the left menu, just above the Code.gs filename.

Add an HTML file and call it “Sidebar”. It should look like this:

Apps Script HTML files

Step 6:
In the Sidebar file, on line 7, between the two BODY tags of the existing code, copy in the following code:

<input type="button" value="Close" onclick="google.script.host.close()" />

This code adds a “Close” button to the sidebar.

Your Sidebar file should now look like this:

<!DOCTYPE html>
<html>
  <head>
    <base target="_top">
  </head>
  <body>

    <input type="button" value="Close" onclick="google.script.host.close()" />

  </body>
</html>

Step 7:
Don’t forget to hit Save

Step 8:
Select the Code.gs file, then select the onOpen function in the menu bar (from the drop down next to the word Debug). Then hit Run.

Step 9:
When you run for the first time, you have to accept the script permissions. If you see an “App isn’t verified” screen, click on Advanced, then “Go to…” and follow the prompts. (More info here.)

Step 10:
After authorizing the app in step 8, jump back to your Google Sheet. You should see a new custom menu “Sheet Sizer” in the menu bar, to the right of the Help menu.

Click the menu to open the sidebar.

Step 11:
Close the menu using the button.

Here’s what you’ve built so far:

Google Sheets sidebar with Apps Script

Sheet Sizer: Add a new button and functionality to the sidebar

Step 12:
In the Sidebar file, after the first BODY tag, on line 6, and before the INPUT tag on line 7, add a new line.

Paste in the new button code:

<input type="button" value="Get Sheet Size" onclick="getSheetSize()" />

When clicked, this will run a function called getSheetSize.

Step 13:
Add the getSheetSize function into the Sidebar file with the following code.

Copy and paste this after the two INPUT tags but before the final BODY tag, on line 9.

<script>
function getSheetSize() {
  
  google.script.run.auditSheet();

}
</script>

When the button is clicked to run the getSheetSize function (client side, in the sidebar), it will now run a function called auditSheet in our Apps Script (server side).

Step 14:
Go to the Code.gs file

Step 15:
Copy and paste this new function underneath the rest of your code:

/**
* Get size data for a given sheet url
*/
function auditSheet() {
  // get Sheet
  const ss = SpreadsheetApp.getActiveSpreadsheet();
  const sheet = ss.getActiveSheet();
 
  // get sheet name
  const name = sheet.getName();

  // get current sheet dimensions
  const maxRows = sheet.getMaxRows();
  const maxCols = sheet.getMaxColumns();
  const totalCells = maxRows * maxCols;

  // output
  SpreadsheetApp.getUi().alert(totalCells);
}

This gets the active Sheet of your spreadsheet and calculates the total number of cells as max rows multiplied by max columns.

Finally, the last line displays an alert popup to show the total number.

Step 16:
Back in your Google Sheet, run Sheet Sizer from the custom Sheet Sizer menu.

When you click on the “Get Sheet Size” button, you should see a popup that shows the number of cells in your Sheet:

Sheet Sizer sidebar

Sheet Sizer: Display the Sheet size in the sidebar

Step 17:
Delete this line of code in the auditSheet function:

SpreadsheetApp.getUi().alert(totalCells);

Step 18:
Paste in this new code, to replace the code you deleted in Step 4:

// put variables into object
const sheetSize = 'Sheet: ' + name +
  '<br>Row count: ' + maxRows +
  '<br>Column count: ' + maxCols +
  '<br>Total cells: ' + totalCells +
  '<br><br>You have used ' + ((totalCells / 5000000)*100).toFixed(2) + '% of your 10 million cell limit.';

return sheetSize;

Now, instead of showing the total number of cells in an alert popup, it sends the result back to the sidebar.

Let’s see how to display this result in your sidebar.

Step 19:
Go to the Sidebar file and copy this code after the two INPUT tags but before the first SCRIPT tag:

<div id="results"></div>

This is a DIV tag that we’ll use to display the output.

Step 20:
Staying in the Sidebar file, replace this line of code:

google.script.run.auditSheet();

with this:

google.script.run.withSuccessHandler(displayResults).auditSheet();

This uses the withSuccessHandler callback function, which we’ve called: displayResults

It runs when the Apps Script function auditSheet successfully executes on the server side. The return value of that auditSheet function is passed to a new function called displayResults, which we’ll create now.

Step 21:
Underneath the getSheetSize function, add this function:

function displayResults(results) {
  // display results in sidebar
  document.getElementById("results").innerHTML = results;
}

When this function runs, it adds the results value (our total cells count) to that DIV tag of the sidebar you added in step 7.

Step 22:
Back in your Google Sheet, run Sheet Sizer from the custom Sheet Sizer menu.

When you click on the “Get Sheet Size” button, you should see a popup that shows the number of cells in your Sheet:

Apps Script to measure Sheet Size

Sheet Sizer: Handle multiple sheets

Step 23:
Modify your auditSheet code to this:

/**
* Get size data for a given sheet url
*/
function auditSheet(sheet) {

  // get spreadsheet object
  const ss = SpreadsheetApp.getActiveSpreadsheet();

  // get sheet name
  const name = sheet.getName();

  // get current sheet dimensions
  const maxRows = sheet.getMaxRows();
  const maxCols = sheet.getMaxColumns();
  const totalCells = maxRows * maxCols;

  // put variables into object
  const sheetSize = {
    name: name,
    rows: maxRows,
    cols: maxCols,
    total: totalCells
  }

  // return object to function that called it
  return sheetSize;

}

Step 24:
In your Code.gs file, copy and paste the following code underneath the existing code:

/**
* Audits all Sheets and passes full data back to sidebar
*/
function auditAllSheets() {

  // get spreadsheet object
  const ss = SpreadsheetApp.getActiveSpreadsheet();
  const sheets = ss.getSheets();

  // declare variables
  let output = '';
  let grandTotal = 0;

  // loop over sheets and get data for each
  sheets.forEach(sheet => {

    // get sheet results for the sheet
    const results = auditSheet(sheet);
    
    // create output string from results
    output = output + '<br><hr><br>Sheet: ' + results.name +
      '<br>Row count: ' + results.rows + 
      '<br>Column count: ' + results.cols +
      '<br>Total cells: ' + results.total + '<br>';

    // add results to grand total
    grandTotal = grandTotal + results.total;

  });

  // add grand total calculation to the output string
  output = output + '<br><hr><br>' + 
    'You have used ' + ((grandTotal / 5000000)*100).toFixed(2) + '% of your 10 million cell limit.';

  // pass results back to sidebar
  return output;

}

This adds a new function, auditAllSheets, which loops over all the sheets in your Google Sheet and calls the auditSheet function for each one. The results for each Sheet are joined together into a result string, called output.

Your Code.gs file should now look like this.

Step 25:
Jump back to your Sidebar file and replace this line of code:

google.script.run.withSuccessHandler(displayResults).auditSheet();

with this:

google.script.run.withSuccessHandler(displayResults).auditAllSheets();

The callback function is the general auditAllSheets function, not the specific individual sheet function.

Step 26:
Back in your Google Sheet, add another sheet to your Google Sheet (if you haven’t already) and run Sheet Sizer.

It will now display the results for all the sheets within your Google Sheet!

Sheet Sizer: Add CSS styles

This step is purely cosmetic to make the sidebar more aesthetically pleasing.

Step 27:

Add these CSS lines inside the HEAD tags of the sidebar file:

<!-- Add CSS code to format the sidebar from google stylesheet -->
<link rel="stylesheet" href="https://ssl.gstatic.com/docs/script/css/add-ons1.css">

<style>
  body {
    padding: 20px;
  }
</style>

Step 28:

Add DIV BLOCKS to the elements of the sidebar:

<div class="block">
  <input type="button" class="create" value="Get Sheet Size" onclick="getSheetSize()" /> 
</div>

<div class="block">
  <input type="button" value="Close" onclick="google.script.host.close()" />
</div>

<div class="block">
  <div id="results"></div> 
</div>

Sheet Sizer: Global Variables

Step 29:

One final improvement from me is to move the 10 million cell limit number into a global variable.

Currently, it’s buried in your script so it’s difficult to find and update if and when Google updates the Sheets cell limit.

It’s not good practice to hard-code variables in your code for this reason.

The solution is to move it into a global variable at the top of your Code.gs file, with the following code:

/**
 * Global Variable
 */
const MAX_SHEET_CELLS = 10000000;

And then change the output line by replacing the 5000000 number with the new global variable MAX_SHEET_CELLS:

// add grand total calculation to the output string
output = output + '<br><hr><br>' + 
  'You have used ' + ((grandTotal / MAX_SHEET_CELLS)*100).toFixed(2) + '% of your 10 million cell limit.';

Here is your final Sheet Sizer tool in action:

Sheet Sizer

Click here to see the full code for the Sheet Sizer tool on GitHub.

Next steps: have a go at formatting the numbers shown in the sidebar, by adding thousand separators so that 1000 shows as 1,000 for example.

A Guide To The New Google Apps Script IDE

In December 2020, Google launched a new Google Apps Script Integrated Development Environment or IDE for short.

The Apps Script team knocked it out of the park with this new editor.

New Google Apps Script IDE

If you’re new to Apps Script, check out my beginner guide to Google Apps Script first.

Continue reading A Guide To The New Google Apps Script IDE

How To Merge Cells In Google Sheets And When To Be Careful

In this tutorial, you’ll learn how to merge cells in Google Sheets, when to use merged cells in Google Sheets, the pros and cons of using merged cells, and finally, how to identify them with Apps Script.
Continue reading How To Merge Cells In Google Sheets And When To Be Careful

Sentiment Analysis For Google Tables Using Apps Script

2024 Update: Table’s features and capabilities are now integrated into AppSheet. Read more here.

In this post we’ll use Google Cloud’s Natural Language API to do sentiment analysis on tickets submitted to a Google Tables Support Issue Tracker.

We’ll use Google Tables as the platform for our Support Tracker and Apps Script to connect to the Cloud Natural Language API.

Here’s a GIF showing a portion of the Issue tracker to illustrate what happens when an issue ticket is submitted:

Sentiment Analysis Google Tables

The description (which someone fills out when they submit a support ticket) is sent to the Natural Language API for sentiment analysis, to determine if it’s positive (i.e. good) or negative (i.e. bad) feedback.

This information is returned to our Google Tables table and displayed as a tag alongside each row, so that we can group tickets by sentiment.

For example, we may want to prioritize tickets that are either extremely positive or extremely negative, as these are probably the most important items to double down on or fix!

Sentiment Analysis Google Tables Overview

Here’s the architecture of the system:

Google Tables Sentiment Analyzer architecture

In words:

  • User submits support ticket
  • New row of data added to our table
  • Bot is triggered and sends data to our Apps Script webhook
  • The Apps Script receives data and parses it
  • Apps Script sends data to Natural Language API
  • And receives the sentiment scores back from API
  • Apps Script parses score to create sentiment tag
  • And updates the original row of data in our table

This all happens in a under a second, so it feels almost simultaneous.

Inspiration for this idea came from the excellent sentiment analysis in Google Sheets post originally published by Alicia Williams.

Note: Since I’ve set this up using a webhook triggered by each new row, the API is called once for every new row. If we were building a high volume ticket system, we’d want to consider a different set up where we send the data through in batches and have the Apps Script running on a timer trigger instead.

Sentiment Analysis Google Tables Set Up

For this example, I’m starting with the default “New ? Support Ticket Queue” template from the Google Tables team.

1. Open Google Tables, login and select Templates > New ? Support Ticket Queue

2. Add three new columns: 1) Sentiment Score (number), 2) Sentiment Magnitude (number) and 3) Sentiment Tag (Tags).

3. Edit the new Sentiment Tag column and add the following tags: Super happy!, Happy, Satisfied, No opinion, Frustrated, Angry, Super angry!

Sentiment analysis tags in Google Tables

(You can edit these categories to whatever you want, but you’ll need them to match the tags in your Apps Script.)

Sentiment Analysis Apps Script Set Up

4. Create a blank Apps Script file in Drive or through the Apps Script dashboard

5. Clear out the boiler plate code and add the following code to declare the two global variables we need for this project (we’ll fill them in soon):

/**
 * global variables
 */
const API_KEY = ''; // <-- enter google cloud project API key
const TABLE_NAME = ''; // <-- enter google tables table ID

Since we just created our Table, let’s copy in the Table ID.

6. We find our table’s ID, by looking at the URL and copying the string right after /table/.

Table URLs can take the following form:

https://tables.area120.google.com/u/0/workspace/abcdefghijklmnop/table/TABLE_ID
https://tables.area120.google.com/u/0/table/TABLE_ID
https://tables.area120.google.com/u/0/table/TABLE_ID/view/abcedfghijk

Look for the string represented by the TABLE_ID in these fictional examples.

7. Paste this ID string into our Apps Script project, between the quotes in the line where we declare the variable TABLE_NAME:

const TABLE_NAME = ''; // <-- enter google tables table ID

Sentiment Analysis Google Cloud Set Up

This is probably the most difficult part of this whole project! 😉

For this to work, we need a Google Cloud account with billing set up. Don’t worry though, the Natural Language API is free for the first 5k “units” we use (each unit is worth 1000 characters). This is way more than we need to set this project up and test it out.

The full details of how to set up Google Cloud Natural Language API can be found in the documentation.

The steps to take are:

8. If you don’t already have a Cloud account, register for a Google Cloud account and set up billing.

9. Create a project in the Google Cloud account to use for this project.

10. Enable the Cloud Natural Language API, by clicking the link half-way down this page.

11. Create credentials for the Cloud Natural Language API. From the Cloud console choose the project we created in step 5 and navigate to APIs & Services > Credentials

Generate a new API Key through +CREATE CREDENTIALS > API key

12. Restrict the API key we generated to the Cloud Natural Language API

13. Copy the API key and paste it into our Apps Script file, between the quotes in the line where we declare the variable API_KEY (see code above):

const API_KEY = ''; // <-- enter google cloud project API key

Sentiment Analysis Google Tables Code

Staying in the Apps Script file, let’s add the webhook listener and main control function code for our program.

We use a special function called doPost(e) { } so that we can (eventually) publish our script as a web app. The doPost function sits there listening for a ping from our Google Tables bot (which we’ll set up later).

When a new row is added to our Google Table by a Form submission, the bot is triggered and sends the data we need through to our webhook.

This doPost function receives that data, parses it and sends it to the Natural Language API for sentiment analysis.

The returned sentiment data is parsed and sent back to our table to update the new row, using the patch method of the Area120Tables service.

14. Add the following doPost code:

/**
 * doPost webhook to catch data from Google Tables
 */
function doPost(e) {
  
  if (typeof e !== 'undefined') {

    // parse data
    const data = JSON.parse(e.postData.contents);

    // get the id and description
    const rowId = data.id
    const description = data.description;

    // analyze sentiment
    const sentiment = analyzeFeedback(description); // [nlScore,nlMagnitude,emotion]
    
    // combine arrays
    const sentimentArray = [rowId,description].concat(sentiment);

    // send score back to Google Tables
    const rowName = 'tables/' + TABLE_NAME + '/rows/' + rowId;
    const sentimentValues = {
      'Sentiment Score': sentiment[0],
      'Sentiment Magnitude': sentiment[1],
      'Sentiment Tag': sentiment[2]
    };
    Area120Tables.Tables.Rows.patch({values: sentimentValues}, rowName);

    return null;
  }
}

In the code above, we call a function called analyzeFeedback, so we had better declare it.

This function handles the logic around the NL scores and how to interpret them as human readable tags. Feel free to play around with the boundaries. The sentiment score is bounded between -1 (max negative) and 1 (max positive), but the magnitude only has a lower bound of 0, so can be any positive number.

For more about the interpretation of the sentiment scoring, have a read of this page in the documentation.

15. Add the following code to our Apps Script file to analyze the sentiment scores:

/**
 * Get each new row of form data and retrieve the sentiment 
 * scores from the NL API for text in the feedback column.
 */
function analyzeFeedback(description) {

  if (description !== '') {
      
      // call the NL API
      const nlData = retrieveSentiment(description);
      nlMagnitude = nlData.documentSentiment.magnitude ? nlData.documentSentiment.magnitude : 0; // set to 0 if nothing returned by api
      nlScore = nlData.documentSentiment.score ? nlData.documentSentiment.score : 0; // set to 0 if nothing returned by api
      //console.log(nlMagnitude);
      //console.log(nlScore);
    }
    else {
      
      // set to zero if the description is blank
      nlMagnitude = 0;
      nlScore = 0;

    }

    // turn sentiment numbers into tags
    let emotion = '';
    
    // happy
    if (nlScore > 0.5) { 
      
      if (nlMagnitude > 2) { emotion = 'Super happy!'; } // higher magnitude gets higher emotion tag
      else { emotion = 'Happy'; }

    }
    
    // satisfied
    else if (nlScore > 0) {  emotion = 'Satisfied'; }

    // frustrated
    else if (nlScore < 0 && nlScore >= -0.5) { emotion = 'Frustrated'; }

    // angry
    else if (nlScore < -0.5) { 
      
      if (nlMagnitude > 2) { emotion = 'Super angry!'; } // higher magnitude gets higher emotion tag
      else { emotion = 'Angry'; }
    
    }

    // if score is 0
    else { emotion = 'No opinion' }

    return [nlScore,nlMagnitude,emotion];
}

Finally we need to declare the function called retrieveSentiment to actually call the API.

16. Add the code to call the NL API:

/**
 * Calls Google Cloud Natural Language API with string from Tables
 */
function retrieveSentiment(description) {
  
  //console.log(description);

  const apiEndpoint = 'https://language.googleapis.com/v1/documents:analyzeSentiment?key=' + API_KEY;
  
  // Create our json request, w/ text, language, type & encoding
  const nlData = {
    document: {
      language: 'en-us',
      type: 'PLAIN_TEXT',
      content: description
    },
    encodingType: 'UTF8'
  };
  
  //  Package all of the options and the data together for the call
  const nlOptions = {
    method : 'post',
    contentType: 'application/json',  
    payload : JSON.stringify(nlData)
  };
  
  //  Try fetching the natural language api
  try {
    
    // return the parsed JSON data if successful
    const response = UrlFetchApp.fetch(apiEndpoint, nlOptions);
    return JSON.parse(response);
    
  } catch(e) {
    
    // log the error message and return null if not successful
    console.log("Error fetching the Natural Language API: " + e);
    return null;
  } 
}

Press save!

The full code is available here on GitHub.

Since we’re using the Area 120 Tables Apps Script service, we need to enable it for this project.

17. Go to Resources > Advanced Google services… and switch on Area120 Tables API:

Apps Script Advanced Services

Publish As Web App

18. Publish this file as a web app via the menu: Publish > Deploy as a web app…

19. Set the access to Anyone, even anonymous, as shown in this image:

Apps Script deploy as Web App

We’ll be prompted to review permissions:

Google Tables Sentiment Analyzer authorization

followed by a review of the project scopes:

Apps Script scopes

Click Allow.

This is a one-time step the first time we publish to the web or run our script (unless we add additional services in the future).

20. Copy the URL of the web app so we can paste that into our Tables bot, which we’ll create next!

Sentiment Analysis Tables Bot Set Up

The final piece of the puzzle is the bot in Google Tables.

When the issue tracker form is submitted it creates a new row of data in our table, which triggers the bot. The bot sends the data to the webhook (i.e. the code above) that handles the rest.

21. Create a new bot with the following specification:

Trigger: row added
Action: Send to webhook
Webhook URL: Our Apps Script web app URL from step 20
Webhook format: POST with JSON
Request parameters:
id : [[record_id]]
description : {{description}}

Visually, this is the bot:

Google Tables bot

The red arrow indicates where we paste the Apps Script web app URL.

Test The Sentiment Analysis Google Tables Tool

Finally, we’re ready to submit the form.

22. From the Google Table, click on the Support Ticket Form to open it in a new tab:

Google Tables Support Tracker Form

23. Submit it with a strong positive or negative sentiment in the description field (which is the one we send to the Natural Language API) to test out the Natural Language scores.

You’ll see the row of data arrive when we submit the form and then, a few moments later, the sentiment analysis columns get automatically populated too!

Sentiment Analysis Google Tables

That’s it! Let me know how you get on in the comments.