This code can be used as an eleventy filter to create the reading time (in minutes) of a piece of content.

readingtime.js

/*
 * Calculate the reading time, in minutes, of a post
 *
 * Assumptions:
 * - average reading time is 240 words per minute
 *   source: https://bit.ly/3HCogSr, "Most Comprehensive
 *   Review To Date Finds The Average Person’s Reading
 *   Speed Is Slower Than Previously Thought"
 *
 * Output:
 * - reading time is rounded to the nearest minute
 * - in the case of less than 1 minute, reading time is
 *   displayed as "less than a minute"
 *
 * @param {String} text
 */

module.exports = function (text) {
  var content = new String(text);
  const speed = 240; // reading speed in words per minute

  // remove all html elements
  var re = /(&lt;.*?&gt;)|(<[^>]+>)/gi;
  var plain = content.replace(re, "");

  // replace all newlines and 's with spaces
  var plain = plain.replace(/\s+|'s/g, " ");

  // create array of all the words in the post & count them
  var words = plain.split(" ");
  var count = words.length;

  // calculate the reading time
  var readingTime = Math.round(count / speed);
  if (readingTime === 0) {
    return "Less than 1 minute to read";
  } else if (readingTime === 1) {
    return "1 minute to read";
  } else {
    return readingTime + " minutes to read";
  }
};

A couple of unsolicited suggestions as I just came across this:

1. Replace var re = /(&lt;.*?&gt;)|(<.*?>)/gi; with var re = /(&lt;.*?&gt;)|(<[^>]+>)/gi;. .*? and other non-greedy matching operators can be faster, but not in this case. Is checking for escaped HTML appropriate, BTW?
2. Normalize all whitespace: var plain = plain.replace(/\s+|'s/g, " ");. You might also want to think about Unicode apostrophes and things like 't, 're, and so on. Maybe plain.replace(/\s+|(?:(\p{Letter})\p{Punctuation}(\p{Letter}))/gu, "$1 $2");, although that’s just me trying to write it on the fly here.

Thanks for your thoughts. I had to think for a second about how I originally came up with that gist. I found a blog post by Phil Hawksworth of Netlify where he was showing how he created a fast search feature for his blog. He has a "sqash" filter that he created and the code above is based on that. I recall spending a lot of time wanting to understand how the regex worked. They are not my strong suit. I'll take a close look at your suggestion and see if I can understand it. I really appreciate you taking the time. Here's a link to Phil's squash filter.

https://github.com/philhawksworth/hawksworx.com/blob/8c96ba2541c8fd6fe6f521cdb5e17848c231636c/src/site/_filters/squash.js

I understand. Regexes can be quite opaque. I recommend using a tool like regex101. It’ll explain every character of the regex and let you try it on different strings in various modes. And, of course, please feel free to ask me—whether here or on Discord—if you have any questions: my first programming language was Perl, so I have an unhealthy degree of comfort with regexes.

He has a "sqash" filter that he created and the code above is based on that.

Thanks for the link. This works as a quick-and-dirty solution for their use case, but I think the changes I suggested will make it more generally useful and accurate. If you want to ignore small words like in the original code, I would suggest defining a Set of those words and filtering the array before counting (var count = words.filter((w) => !WORDS_TO_IGNORE.has(w)).length;) instead of trying to ignore them via regex. In that case, you’d also need var content = text.toLowerCase();.

Cool. Thanks for the link to regex101. Looks very handy. I'm going to make an update to my Calculating blog using your suggestions (at some point). Right now, I'm scouring the net for the next issue of The 11ty Bundler. I am thinking that if I continue to generate the Bundler, it will probably turn into its own website. I broke down and bought the 11tybundler.com domain yesterday. Not that that necessarily signals commitment as I have a handful of other unused domains that started as ideas in my head. At least with this one, I've already started something that could make more sense for me to head down that path. What I like about the idea is that it will give me more reasons to keep learning about Eleventy, JS, and CSS.

Sounds like a great idea to me. I look forward to reading more, if you do put together more issues (so to speak). 😊

This gist has been updated with the help of @Aankhen. Many thanks. The resulting word counts are comparable to the earlier regex I used.