let b = document.body;

​

// length of child nodes is "1"

console.log(b.firstElementChild.childNodes.length);

​

// We split the child text nodes

b.firstElementChild.firstChild.splitText(15);

​

// Now the length is "2"

console.log(b.firstElementChild.childNodes.length);

​

// Read the text as a single text node

console.log(b.firstElementChild.childNodes[0].wholeText);

​

// It's still 2 nodes

console.log(b.firstElementChild.childNodes.length);

​

// Then we can normalize as needed

b.firstElementChild.normalize();

​

// Back to "1" text node

console.log(b.firstElementChild.childNodes.length);