Change character encoding?

Question

Edit > Preferences > New Document > Character
encoding is set to UTF-8
However when I edit documents with non-standard extensions
(I'm working on PHP files with a .ctp extension) the document still
seems to save in iso-8859-1 format. This problem doesn't seem to
occur on files with .php/html extensions.
Does anyone know of a solution to this problem?
Thanks,
Emil

Answer 1

I'm not sure where you are getting %xx encoded UTF-8.... Is it cuz you have it in a GET method form and that's what you are seeing in the browser's location bar? ...
Let's assume you have a form on a page, and the page's charset is set to UTF-8, and you want to generate a URL encoded string (%xx format, although URLEncoder will not encode ASCII chars that way...).
In the page processing the form, you need to do this:
request.setCharacterEncoding("UTF-8"); // makes bytes read as UTF-8 strings(assumes that the form page was properly set to the UTF-8 charset)
String fieldValue = request.getParameter("fieldName"); // get value
// the value is now a Unicode String in Java, generated from reading the bytes submitted from the form as UTF-8 encoded text...
String utf8EncString = URLEncoder.encode(fieldValue, "UTF-8");
// now utf8EncString is a URL encoded (%xx) string of UTF-8 values
String euckrEncString = URLEncoder.encode(fieldValue, "EUC-KR");
// now euckrEncString is a URL encoded (%xx) string of EUC-KR valuesWhat is probably screwing things up for you mostly is this:
euckrValue = new String(utf8Value.getBytes(), "EUC-KR");
What this does is takes the bytes of the string utf8Value (which is not really UTF-8... see below) in the local encoding (possibly Cp1252 (Windows) or ISO8895-1 (Linux), or EUC-KR if it's Korean Windows), and then reads them as if they were EUC-KR... which they aren't.
The key here is that Strings in Java are not of any encoding. They are pure Unicode values. Encodings only matter when converting to or from bytes. The strings stored in a file or sent over the net have to convert to bytes since that's what is stored/sent, just bytes. The encoding defines how the characters can be encoded into 1 or more bytes, and thus reconstructed.