In this post from the old Forum, @TIMAI2 shows us how we can create a PDF file with JavaScript using the jsPDF.js library.
Thank you TIM!

HOWTO: Create a PDF, offline, "within" AI2 using jsPDF and base64 extn by Juan Antonio

The code uses the jsPDF.js JavaScript library and an extension to work with Base64.

We will try to adapt the code without using the extension. [spoiler: currently not working]

I have used two methods, one with extension and the other without extension:

Method 1:

Method 2:

The problem is that atob doesn't decode the image well.
javascript_pdf1 702×550 70.4 KB eval(content); var returnOutput = doc.output('datauri'); // .split("base64,")[1]) to remove data:*;base64, AppInventor.setWebViewString(returnOutput.split("base64,")[1]); </script>

genpdf_atob.html

<script src="jspdf.min.js"></script>
<script>
var content = AppInventor.getWebViewString();
var doc = new jsPDF();
eval(content);
var returnOutput = doc.output('datauri');
 // .split("base64,")[1]) to remove data:*;base64,
    AppInventor.setWebViewString(atob(returnOutput.split("base64,")[1]));
</script>
ToBase64.htm
<script>
async function readTextFileAsBlob(file) {
    const response = await fetch(file);
    const blob = await response.blob();
    return blob;
const blobToBase64DataURL = blob => new Promise(  
    resolvePromise => {
        const reader = new FileReader();
        reader.onload = () => resolvePromise(reader.result);
        reader.readAsDataURL(blob);
my_file = AppInventor.getWebViewString(); // INPUT
readTextFileAsBlob(my_file).then(
    async blob => {
        const base64URL = await blobToBase64DataURL(blob);
               // .split("base64,")[1]) to remove data:image/png;base64,
	     AppInventor.setWebViewString(base64URL.split("base64,")[1]); // OUTPUT    
</script>
              
- Documentation.
http://raw.githack.com/MrRio/jsPDF/master/docs/index.html
https://mrrio.github.io/
https://artskydj.github.io/jsPDF/docs/jsPDF.html
http://kio4.com/appinventor/277_extension_imagen_string.htm