Re: Croatian characters not surviving LOAD DATA INFILE from UTF-8 .txt file

[Rick James]

GIGO

Kupreškić <-- This has the letters encoded according to some character set (probably utf8 on this forum)
Kupre%C5%A1ki%C4%87 <-- This is "html encoding". It is understood only by browsers.

You have CHARACTER SET latin2 on the table.
What arguments did you use when doing the LOAD DATA? That makes a big difference.
Are the values in the table correct? That is hard to test, SELECT HEX may help. See this for further discussion:
http://mysql.rjweb.org/doc.php/charcoll

Kupre??kiÄ? <-- This is very likely due to incorrect settings either during loading of the data, or during fetching the data. Again see my web page.

I think 5.5.12 has a utf8_croatian_ci; it might be better in the long run to use utf8 instead of latin2 -- this would let you use character sets from all languages, and might avoid the need for different columns having different charsets.
SHOW COLLATION;

"%C5%A1" -- This smells a lot like it came from a utf8 encoding (hex C5A1), not latin2. Perhaps your data is really utf8? Another argument for switching from latin1/2 to utf8.