John Davidson

Message:

I am trying to convert a UTF-16LE txt file to UTF-8.

I found a possible solution here on Stackoverflow and tried the following two:

 iconv($in_charset = 'UTF-16LE' , $out_charset = 'UTF-8' , $str);

and

 mb_convert_encoding($str, 'UTF8', 'UTF-16LE');

This seems to be working, unfortunately only partially. The problem seems to be that I am trying to convert a txt file that is in Czech, which uses Latin extended characters, such as š, č, ř, ž, Š, Č, Ř, ě etc., i.e. letters from the Czech alphabet, which is much longer than the English one. And unfortunately all these non-English characters are encoded incorrectly. Is there any way around this to convert these correctly as well?

I thought about str_replace for the converted text, but the problem is that some of the above are not converted to "weird symbols", but instead to other letters such as "q" or "+" sign, which means that if I replace them, it would also replace all "q" letters or "+" symbols, which is not what I want.

Here are all the possible non-English characters that we use in Czech:

 ěščřžýáíéťůúňĚŠČŘŽÝÁÍÉŮÚŤŇ

---

My Answer:
UTF-8 characters at all, so this approach wouldn't work for all cases.

One possible solution is to use the `mb_convert_encoding` function with the `UTF-16LE` encoding and specify the `UTF-8` encoding along with the `//IGNORE` flag. This flag will ignore any characters that cannot be converted to UTF-8. Here's an example:

php
$utf8_str = mb_convert_encoding($str, 'UTF-8', 'UTF-16LE//IGNORE');


This should help in converting the Latin extended characters correctly to UTF-8. You can also try different combinations of encoding and flags to see which one works best for your specific case.