Best solution for getting the word count?

[shustudios]

Hi, I'm looking for an optimized way for getting the number of words in the editor. The problem is that text between nodes (multiple paragraphs) get "joined" when converting the editor's contents into plain text. Right now the only solution i see is to recursively loop through all the nodes to get the text.

Anyone have a better idea? Thanks!

[sereneinserenade]

I don't know if this is the best solution, but here's what I've tried.

https://codesandbox.io/s/reverent-babycat-g4yuk

[shustudios]

Hi,
Your solution has the same problem as mine - if you write:

this
is
a
test

You wil see it only counts as 1 word. This is because when you read as innertext, the paragraphs gets joined without any spaces between them.

[shustudios]

Will probably have to look at using a combination of the existing solution, plus counting all the paragraphs, but I was hoping to find a more efficient method.

[sereneinserenade]

Hey there @ironshu , I did some adjustments and it works fine for me. https://codesandbox.io/s/reverent-babycat-g4yuk

    countTheWords() {
      const proseMirror = document.querySelector(".ProseMirror");

      const innerText = proseMirror.innerText
        .replaceAll("\n", " ")
        .split(" ")
        .filter((s) => s !== "" && s !== " ");

      this.length = innerText.length;
    },

[shustudios]

hmmm, that didn't work for me, but it was probably because I was using the editor textContent property... I'll try your approach. Still wondering if it is efficient, but it is a nice and simple solution 👍

[shustudios]

Thanks! Also I found that the issue for me was not the source string, but the regex I was using the split the words. So instead of parsing the html output, seems you can use editor.state.doc.textContent - not sure which is faster though

Edit - scratch that - had my variables mixed up... need to parse the dom as per your solution..