str_word_count doesn't return the right number of words

The issue comes from how str_word_count() defines a "word" and which characters it treats as word boundaries. By default, str_word_count() does not handle Unicode apostrophes (like ’) or accented characters properly, and it may split words unexpectedly, especially with French text.

For example, in your string:

L’Auvergne, cœur volcanique de la France, se distingue par ses chaînes de puys.

• The character ’ (Unicode RIGHT SINGLE QUOTATION MARK) is not treated as part of a word.
• Accented characters (like é, è, ç) may not be included as word characters unless specified.

Solution:

To handle this properly, you can use a regular expression with preg_match_all() that supports Unicode and includes apostrophes and accented characters.

Here's how you can do it:

$text = "L’Auvergne, cœur volcanique de la France, se distingue par ses chaînes de puys.";

// This regex matches words with Unicode letters and allows for apostrophes inside words.
preg_match_all('/\p{L}+(?:[’\'-]\p{L}+)*/u', $text, $matches);

$count = count($matches[0]);

echo $count; // Should output 14

Explanation:

• \p{L}+ matches one or more Unicode letters.
• (?:[’\'-]\p{L}+)* allows for apostrophes (both ' and ’) or hyphens inside words.
• The u modifier enables Unicode support.

Summary:
str_word_count() is limited for non-English or accented text. For accurate word counts in languages like French, use preg_match_all() with a Unicode-aware regular expression as shown above.