convert_cyr_string()

This chapter explains what convert_cyr_string() did, why it existed, why it was removed, and what to use in its place. It is reference material for reading or maintaining old PHP code — not a function you should call in anything new.

What convert_cyr_string() did

The function converted a string from one single-byte Cyrillic character set to another. Before UTF-8 became the default, Cyrillic text was stored in several incompatible 8-bit encodings (Windows-1251, KOI8-R, ISO 8859-5, DOS code page 866, and a couple of Mac/main-frame sets). convert_cyr_string() remapped the bytes from one of those legacy sets to another.

Syntax

string convert_cyr_string ( string $str , string $from , string $to )

It took three parameters and returned the converted string:

The $from and $to arguments were single-letter codes, not full encoding names:

Example

<?php
// Legacy code only — this throws a fatal error on PHP 8+.
$converted = convert_cyr_string($str, "w", "k"); // Windows-1251 -> KOI8-R
echo $converted;
?>

Here $str is assumed to contain raw Windows-1251 bytes ("w"), and the function rewrites them as KOI8-R bytes ("k"). The key thing to understand is that convert_cyr_string() operated on raw bytes and assumed the source string was already in the $from encoding. If you fed it a UTF-8 string (the modern default), the output was meaningless garbage (mojibake), because UTF-8 bytes do not correspond to any single legacy Cyrillic set.

Why it was removed

• It only knew a fixed list of legacy 8-bit Cyrillic encodings — UTF-8 was never supported, so it could not participate in modern, Unicode-based text handling.
• It silently mangled any input that was not actually in the declared $from encoding.
• Its job is fully covered by the general-purpose iconv() and mb_convert_encoding() functions, which support hundreds of encodings including UTF-8.

For these reasons it was deprecated in PHP 7.4 and deleted in PHP 8.0. Calling it today raises a fatal Error.

What to use instead

Use mb_convert_encoding() (from the mbstring extension) or iconv(). Both take real encoding names and both support UTF-8:

<?php
// Convert Windows-1251 bytes to UTF-8.
$utf8 = mb_convert_encoding($legacy, "UTF-8", "Windows-1251");

// The same thing with iconv().
$utf8 = iconv("Windows-1251", "UTF-8", $legacy);

// Convert KOI8-R to UTF-8.
$utf8 = mb_convert_encoding($legacy, "UTF-8", "KOI8-R");
?>

Converting to UTF-8 (rather than between two legacy sets) is almost always what you want: once text is UTF-8 it works everywhere in modern PHP, databases, and browsers.

Summary

convert_cyr_string() was a pre-Unicode helper for swapping bytes between legacy Cyrillic encodings. It is gone from PHP 8.0 onward. If you encounter it in old code, replace each call with iconv() or mb_convert_encoding() and migrate your data to UTF-8.

For more on working with text in PHP, see PHP Strings and htmlentities().

Practice