Google Apps Script, Use PDF Data with AI

Convert PDF Content into Something Useful

Working with PDFs can be annoying.

But, we are going to make it easy. First you need to:

1. Make a folder in Google Drive

2. Drop a PDF in the folder

3. Go to the sharing and grab the URL, it should have a /file/d/ in the string

In Google Apps Script, make a new script, and add these services:

The Apps Script Primer is here if you have no idea what that meant.

The process works like this:

1. Insert the url

2. Script converts the data to a Google Doc

3. Google Doc is read into a container; The script allows you to set a word limit; default is 2000

4. Data is accessible

5. Paid subscribers get the next level, sending the data to OpenAI API

The PDF Conversion Code

function convertPdfAndStoreLimitedContents(aiRequest, inputUrl) {
  // Default PDF URL if none is provided

  var pdfUrl = inputUrl || 'https://drive.google.com/file/d/YOUR ID GOES HERE/view?usp=drive_link';

  var theAiRequest = aiRequest || "Summarize this into three paragraphs.";
  
  try {
    // Extract the file ID from the URL
    var pdfFileId = extractFileIdFromUrl(pdfUrl);
    
    // Convert the PDF to a Google Doc
    var googleDocId = convertPdfToGoogleDoc(pdfFileId);
    
    // Store the limited contents of the Google Doc
    var documentContents = storeLimitedGoogleDocContents(googleDocId, 2000);
    
    // Log the contents
    Logger.log('Converted Document Contents:');
    Logger.log(documentContents);
    
   
    
  } catch (error) {
    Logger.log('Error: ' + error.message);
  }
}



function extractFileIdFromUrl(url) {
  // Extract the file ID from the full URL
  var regex = /\/d\/(.*?)\//;
  var match = url.match(regex);
  if (match && match[1]) {
    return match[1];
  } else {
    throw new Error('Invalid URL. Unable to extract file ID.');
  }
}

function convertPdfToGoogleDoc(pdfFileId) {
  // Get the PDF file
  var pdfFile = DriveApp.getFileById(pdfFileId);
  
  // Get the parent folder of the PDF
  var parentFolder = pdfFile.getParents().next();
  
  // Create a new Google Doc from the PDF
  var convertedFile = Drive.Files.copy(
    {
      title: pdfFile.getName().replace('.pdf', '') + ' (Converted)',
      mimeType: MimeType.GOOGLE_DOCS
    },
    pdfFileId
  );
  
  // Move the converted file to the same location as the PDF
  var googleDocFile = DriveApp.getFileById(convertedFile.id);
  parentFolder.addFile(googleDocFile);
  DriveApp.getRootFolder().removeFile(googleDocFile); // Remove from "My Drive"

  // Log the new file's details
  Logger.log('Converted File Name: ' + googleDocFile.getName());
  Logger.log('Converted File URL: ' + googleDocFile.getUrl());
  
  // Return the Google Doc ID
  return googleDocFile.getId();
}

function storeLimitedGoogleDocContents(googleDocId, wordLimit) {
  // Access the Google Document
  var document = DocumentApp.openById(googleDocId);
  
  // Get the body text
  var bodyText = document.getBody().getText();
  
  // Split the text into words
  var words = bodyText.split(/\s+/);
  
  // Check if the text exceeds the word limit
  if (words.length > wordLimit) {
    // Truncate the text to the word limit
    words = words.slice(0, wordLimit);
  }
  
  // Join the words back into a single string
  var limitedText = words.join(' ');
  
  // Store text in a container
  var textContainer = {
    fullText: limitedText,
    wordCount: words.length
  };
  
  // Return the container
  return textContainer;
}

The Output should look like this:

If you are a paid subscriber, there is an additional script below that incorporates the AI function and the AI function itself.

Consider signing up to fuel my dedication to serverless programming.