<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:dc="http://purl.org/dc/elements/1.1/">
  <channel>
    <title>DEV Community: Medleyo</title>
    <description>The latest articles on DEV Community by Medleyo (@medleyo).</description>
    <link>https://dev.to/medleyo</link>
    <image>
      <url>https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F4246405%2F0b1a1cb3-54c3-45fe-9f86-58af8eacf046.png</url>
      <title>DEV Community: Medleyo</title>
      <link>https://dev.to/medleyo</link>
    </image>
    <atom:link rel="self" type="application/rss+xml" href="https://dev.to/feed/medleyo"/>
    <language>en</language>
    <item>
      <title>Counting words in the browser: why split(/\s+/) is not enough</title>
      <dc:creator>Medleyo</dc:creator>
      <pubDate>Mon, 16 Nov 2026 19:04:30 +0000</pubDate>
      <link>https://dev.to/medleyo/counting-words-in-the-browser-why-splits-is-not-enough-l8f</link>
      <guid>https://dev.to/medleyo/counting-words-in-the-browser-why-splits-is-not-enough-l8f</guid>
      <description>&lt;p&gt;Counting words looks like a one-liner: &lt;code&gt;text.split(/\s+/).length&lt;/code&gt;. It works for English prose and fails quietly for most of the world. Here is what I ran into while building a word counter that runs entirely in the browser, and what I ended up doing instead.&lt;/p&gt;

&lt;h2&gt;
  
  
  Where &lt;code&gt;split(/\s+/)&lt;/code&gt; breaks
&lt;/h2&gt;

&lt;p&gt;Many writing systems do not put spaces between words. Take a Japanese sentence:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight javascript"&gt;&lt;code&gt;&lt;span class="s2"&gt;"今日は天気がいいですね。"&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nf"&gt;split&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="o"&gt;/&lt;/span&gt;\&lt;span class="n"&gt;s&lt;/span&gt;&lt;span class="o"&gt;+/&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="n"&gt;length&lt;/span&gt;&lt;span class="p"&gt;;&lt;/span&gt; &lt;span class="c1"&gt;// 1&lt;/span&gt;
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;One "word". Chinese, Thai, Lao, Khmer and Myanmar do not put spaces between words either. A counter based on whitespace is wrong for a large share of users, and nothing tells them it is wrong.&lt;/p&gt;

&lt;h2&gt;
  
  
  Let the platform find word boundaries
&lt;/h2&gt;

&lt;p&gt;&lt;code&gt;Intl.Segmenter&lt;/code&gt; does locale-sensitive text segmentation: you pick a granularity (graphemes, words or sentences) and iterate over the segments, and it knows that Japanese needs different rules from English:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight javascript"&gt;&lt;code&gt;&lt;span class="kd"&gt;const&lt;/span&gt; &lt;span class="n"&gt;words&lt;/span&gt; &lt;span class="o"&gt;=&lt;/span&gt; &lt;span class="k"&gt;new&lt;/span&gt; &lt;span class="nc"&gt;Intl&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nc"&gt;Segmenter&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="kc"&gt;undefined&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt; &lt;span class="n"&gt;granularity&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"word"&lt;/span&gt; &lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;&lt;span class="p"&gt;;&lt;/span&gt;

&lt;span class="kd"&gt;function&lt;/span&gt; &lt;span class="nf"&gt;countWords&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="n"&gt;text&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt;
  &lt;span class="kd"&gt;let&lt;/span&gt; &lt;span class="n"&gt;count&lt;/span&gt; &lt;span class="o"&gt;=&lt;/span&gt; &lt;span class="mi"&gt;0&lt;/span&gt;&lt;span class="p"&gt;;&lt;/span&gt;
  &lt;span class="k"&gt;for&lt;/span&gt; &lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="kd"&gt;const&lt;/span&gt; &lt;span class="n"&gt;part&lt;/span&gt; &lt;span class="k"&gt;of&lt;/span&gt; &lt;span class="n"&gt;words&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nf"&gt;segment&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="n"&gt;text&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt;
    &lt;span class="k"&gt;if&lt;/span&gt; &lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="n"&gt;part&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="n"&gt;isWordLike&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt; &lt;span class="n"&gt;count&lt;/span&gt; &lt;span class="o"&gt;+=&lt;/span&gt; &lt;span class="mi"&gt;1&lt;/span&gt;&lt;span class="p"&gt;;&lt;/span&gt;
  &lt;span class="p"&gt;}&lt;/span&gt;
  &lt;span class="k"&gt;return&lt;/span&gt; &lt;span class="n"&gt;count&lt;/span&gt;&lt;span class="p"&gt;;&lt;/span&gt;
&lt;span class="p"&gt;}&lt;/span&gt;
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;The &lt;code&gt;isWordLike&lt;/code&gt; flag matters: the segmenter also returns spaces and punctuation as segments, and you do not want to count those. On Node 24 this gives:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight javascript"&gt;&lt;code&gt;&lt;span class="p"&gt;[&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nf"&gt;segment&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="s2"&gt;"Hello, world! ... ?"&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;&lt;span class="p"&gt;]&lt;/span&gt;   &lt;span class="c1"&gt;// ["Hello", "world"]&lt;/span&gt;
&lt;span class="p"&gt;[&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nf"&gt;segment&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="s2"&gt;"今日は天気がいいですね。"&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;&lt;span class="p"&gt;]&lt;/span&gt;  &lt;span class="c1"&gt;// ["今日","は","天気","が","いい","です","ね"]&lt;/span&gt;
&lt;span class="p"&gt;[&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nf"&gt;segment&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="s2"&gt;"สวัสดีครับ ผมชื่อสมชาย"&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;&lt;span class="p"&gt;]&lt;/span&gt;   &lt;span class="c1"&gt;// ["สวัสดี","ครับ","ผม","ชื่อ","สมชาย"]&lt;/span&gt;
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;(&lt;code&gt;segment&lt;/code&gt; here stands for the word segmenter above, filtered to word-like parts.) The split is not perfect and can differ between engines and versions, so I treat the numbers as "good estimates", not as exact truths. Hyphenated words are also worth knowing about: &lt;code&gt;e-mail&lt;/code&gt; is counted as two words, and &lt;code&gt;don’t&lt;/code&gt; and &lt;code&gt;example.com&lt;/code&gt; as one.&lt;/p&gt;

&lt;h2&gt;
  
  
  Sentences are not "split on a dot"
&lt;/h2&gt;

&lt;p&gt;Splitting sentences on &lt;code&gt;.&lt;/code&gt; miscounts decimals and domains. The sentence granularity handles them:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight javascript"&gt;&lt;code&gt;&lt;span class="p"&gt;[&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="k"&gt;new&lt;/span&gt; &lt;span class="nc"&gt;Intl&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nc"&gt;Segmenter&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="s2"&gt;"en"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt; &lt;span class="n"&gt;granularity&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"sentence"&lt;/span&gt; &lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nf"&gt;segment&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="s2"&gt;"Pi is 3.14. See example.com! Done?"&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;&lt;span class="p"&gt;]&lt;/span&gt;
&lt;span class="c1"&gt;// ["Pi is 3.14. ", "See example.com! ", "Done?"]&lt;/span&gt;
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;I keep only segments that contain at least one letter or digit (&lt;code&gt;/[\p{L}\p{N}]/u&lt;/code&gt;), so a stray &lt;code&gt;...&lt;/code&gt; does not become a sentence.&lt;/p&gt;

&lt;h2&gt;
  
  
  "Characters" has three answers
&lt;/h2&gt;

&lt;p&gt;The same emoji, a family of three people, measured three ways:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight javascript"&gt;&lt;code&gt;&lt;span class="kd"&gt;const&lt;/span&gt; &lt;span class="n"&gt;s&lt;/span&gt; &lt;span class="o"&gt;=&lt;/span&gt; &lt;span class="s2"&gt;"👨‍👩‍👧"&lt;/span&gt;&lt;span class="p"&gt;;&lt;/span&gt;
&lt;span class="n"&gt;s&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="n"&gt;length&lt;/span&gt;&lt;span class="p"&gt;;&lt;/span&gt;                 &lt;span class="c1"&gt;// 8  (UTF-16 code units)&lt;/span&gt;
&lt;span class="p"&gt;[&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="n"&gt;s&lt;/span&gt;&lt;span class="p"&gt;]&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="n"&gt;length&lt;/span&gt;&lt;span class="p"&gt;;&lt;/span&gt;            &lt;span class="c1"&gt;// 5  (code points, three emoji joined by two zero-width joiners)&lt;/span&gt;
&lt;span class="p"&gt;[&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="k"&gt;new&lt;/span&gt; &lt;span class="nc"&gt;Intl&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nc"&gt;Segmenter&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="kc"&gt;undefined&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt; &lt;span class="n"&gt;granularity&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"grapheme"&lt;/span&gt; &lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nf"&gt;segment&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="n"&gt;s&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;&lt;span class="p"&gt;]&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="n"&gt;length&lt;/span&gt;&lt;span class="p"&gt;;&lt;/span&gt; &lt;span class="c1"&gt;// 1 (what a reader sees)&lt;/span&gt;
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;&lt;code&gt;string.length&lt;/code&gt; counts UTF-16 code units, so anything outside the Basic Multilingual Plane counts twice. Spreading the string iterates by code point, which fixes plain emoji and is what my counter uses. If you need "what a person would call one character", use the grapheme granularity instead. Which one is right depends on the limit you are checking: a database column, an SMS and a social post each count differently.&lt;/p&gt;

&lt;h2&gt;
  
  
  A fallback for older browsers
&lt;/h2&gt;

&lt;p&gt;MDN lists &lt;code&gt;Intl.Segmenter&lt;/code&gt; as widely available since April 2024, but older browsers and runtimes still exist. I feature-detect it and fall back to whitespace words and punctuation-based sentences, and I test both paths:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight javascript"&gt;&lt;code&gt;&lt;span class="kd"&gt;const&lt;/span&gt; &lt;span class="n"&gt;supported&lt;/span&gt; &lt;span class="o"&gt;=&lt;/span&gt; &lt;span class="k"&gt;typeof&lt;/span&gt; &lt;span class="nc"&gt;Intl&lt;/span&gt; &lt;span class="o"&gt;!==&lt;/span&gt; &lt;span class="s2"&gt;"undefined"&lt;/span&gt; &lt;span class="o"&gt;&amp;amp;&amp;amp;&lt;/span&gt; &lt;span class="s2"&gt;"Segmenter"&lt;/span&gt; &lt;span class="k"&gt;in&lt;/span&gt; &lt;span class="nc"&gt;Intl&lt;/span&gt;&lt;span class="p"&gt;;&lt;/span&gt;
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;The fallback is worse for Japanese, Chinese and Thai, but it never crashes and it is right for space-separated languages.&lt;/p&gt;

&lt;h2&gt;
  
  
  Takeaways
&lt;/h2&gt;

&lt;ul&gt;
&lt;li&gt;Do not count words with &lt;code&gt;split(/\s+/)&lt;/code&gt; unless you only ever see space-separated text.&lt;/li&gt;
&lt;li&gt;Use &lt;code&gt;Intl.Segmenter&lt;/code&gt; with &lt;code&gt;granularity: "word"&lt;/code&gt; and check &lt;code&gt;isWordLike&lt;/code&gt;.&lt;/li&gt;
&lt;li&gt;Decide what "character" means for your use case: code units, code points or graphemes.&lt;/li&gt;
&lt;li&gt;Keep a fallback and test it, and expect small differences between engines.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;The counter I built this for runs in the browser without sending text anywhere: &lt;a href="https://medleyo.com/word-counter/" rel="noopener noreferrer"&gt;https://medleyo.com/word-counter/&lt;/a&gt;&lt;/p&gt;

</description>
      <category>javascript</category>
      <category>webdev</category>
      <category>unicode</category>
      <category>i18n</category>
    </item>
  </channel>
</rss>
