How To: Get encoding of a remote file

Question

How To: Get encoding of a remote file
Java EE
URL url = new URL ("http://www.someSite.com/myCsvFile.csv"); // comma separated
InputStream is = url.openConnection().getInputStream();
InputStreamReader reader = new InputStreamReader(is);
System.out.println("reader.getEncoding(): " + reader.getEncoding());
For both an ISO-8859-1 file and a UTF-8 file I get the following print out:
reader.getEncoding(): Cp1252
Could it have something to do with this warning during boot the .war?
[WARNING] Using platform encoding (Cp1252 actually) to copy filtered resources, i.e. build is platform dependent!
If I use local files it prints ISO8859_1 for one of the files.

Answer 1

All of that is because the HTTP server attaches a charset to every response. That's what you are seeing. The server may be using some logic to determine the actual encoding of the file it returns, or it may simply be using a hard-coded charset which may or may not be suitable for reading the file. The latter is unfortunately more likely.
By the way if you receive an XML file over HTTP, and the HTTP charset differs from the encoding declared in the XML document, there's a rule which says the HTTP charset takes precedence. (I don't know where that rule is documented, but I have encountered that situation in real life -- the data came from a Google application -- and that rule was indeed the right thing to do.)
If you're still under the impression that there's something which can look at a file and determine what encoding was used to produce it, let me tell you that there isn't. Sure, there's that XML prolog thing which works for XML files (if they weren't botched by the producer), but for text files in general there's no way to determine their encoding. Short of asking the person who created them, that is.