<glacius:metadata>

    <title>string.Format() in PHP</title>

    <description>Implementation of .NET-style string.Format() in PHP</description>

    <category>PHP</category>

    <category>C#</category>

    <category>Programming</category>

    <category>Legacy blog posts</category>

</glacius:metadata>

<glacius:macro name="legacy blargh banner">

    <properties>

        <originalUrl>https://tmont.com/blargh/2010/1/string-format-in-php</originalUrl>

        <originalDate>2010-01-09T22:22:08.000Z</originalDate>

    </properties>

</glacius:macro>

<p>

  .NET's <a href="https://www.php.net/sprintf"><code>sprintf()</code></a> 

  <a href="https://docs.microsoft.com/en-us/dotnet/api/system.string.format?view=net-6.0">equivalent</a> 

  has ordered parameters.

</p>

<glacius:code lang="csharp"><![CDATA[var s = string.Format("My name is {1}, {0}", "John", "Doe");

//s = "My name is Doe, John"

var s = string.Format("My name is {{{0}}}", "John");

//s = "My name is {John}"]]></glacius:code>

<p>I needed that. Here it is, PHPified:</p>

<glacius:code lang="php"><![CDATA[function format($format) {

	$args = func_get_args();

	$format = array_shift($args);

	preg_match_all('/(?=\{)\{(\d+)\}(?!\})/', $format, $matches, PREG_OFFSET_CAPTURE);

	$offset = 0;

	foreach ($matches[1] as $data) {

		$i = $data[0];

		$format = substr_replace($format, @$args[$i], $offset + $data[1] - 1, 2 + strlen($i));

		$offset += strlen(@$args[$i]) - 2 - strlen($i);

	}

	return $format;

}]]></glacius:code>

<p>

  There's a lot going on in here that isn't used very often in PHP. Here is a terse 

  explanation of the not-so-obvious features:

</p>

<ul>

  <li>

    The <a href="https://www.php.net/manual/en/pcre.constants.php#constant.preg-offset-capture"><code>PREG_OFFSET_CAPTURE</code></a> 

    constant tells the regex engine to capture the string index offset of each match.

  </li>

  <li>

    <code>(?=\{)</code> and <code>(?!\})</code> in the regular expression are 

    <a href="https://www.regular-expressions.info/lookaround.html">positive and negative lookarounds</a>, 

    respectively. The nice thing about that is they're also zero-width matching, which means they 

    won't be returned as part of the match result. These are needed because to output a 

    literal <code>{</code> in a C# string you precede it with another <code>{</code> (a little known 

    fact about C#).

  </li>

  <li>

    <a href="https://www.php.net/manual/en/function.substr-replace.php"><code>substr_replace</code></a> is 

    an infrequently used, extremely useful function, that replaces a substring within a string, 

    expanding to the given width.

  </li>

  <li>

    The <code>$offset</code> variable tracks the string length delta. Since we're modifying the

    string, the offsets captured in the regex match becomes out of date.

  </li>

</ul>