How to encode and decode Broken Chinese/Unicode ch

问题:

I've tried googling around but wasn't able to find what charset that this text below belongs to:

具有éœé›»ç”¢ç”Ÿè£ç½®ä¹‹å½±åƒè¼¸å…¥è£ç½®

But putting <meta http-equiv="Content-Type" Content="text/html; charset=utf-8"> and keeping that string into an HTML file, I was able to view the Chinese characters properly:

具有靜電產生裝置之影像輸入裝置

So my question is:

1. What tools can I use to detect the character set of this text?

2. And how do I convert/encode/decode them properly in C#?

Updates: For completion sake, i've updated this test.

   [TestMethod]
    public void TestMethod1()
    {
        string encodedText = "具有éœé›»ç”¢ç”Ÿè£ç½®ä¹‹å½±åƒè¼¸å…¥è£ç½®";
        Encoding utf8 = new UTF8Encoding();
        Encoding window1252 = Encoding.GetEncoding("Windows-1252");

        byte[] postBytes = window1252.GetBytes(encodedText);

        string decodedText = utf8.GetString(postBytes);
        string actualText = "具有靜電產生裝置之影像輸入裝置";
        Assert.AreEqual(actualText, decodedText);
    }
}

Thanks.

回答1:

What is happening when you save the "bad" string in a text file with a meta tag declaring the correct encoding is that your text editor is saving the file with Windows-1252 encoding, but the browser is reading the file and interpreting it as UTF-8. Since the "bad" string is incorrectly decoded UTF-8 bytes with the Windows-1252 encoding, you are reversing the process by encoding the file as Windows-1252 and decoding as UTF-8.

Here's an example:

using System.Text;
using System.Windows.Forms;

namespace Demo
{
    class Program
    {
        static void Main(string[] args)
        {
            string s = "具有靜電產生裝置之影像輸入裝置"; // Unicode
            Encoding Windows1252 = Encoding.GetEncoding("Windows-1252");
            Encoding Utf8 = Encoding.UTF8;
            byte[] utf8Bytes = Utf8.GetBytes(s); // Unicode -> UTF-8
            string badDecode = Windows1252.GetString(utf8Bytes); // Mis-decode as Latin1
            MessageBox.Show(badDecode,"Mis-decoded");  // Shows your garbage string.
            string goodDecode = Utf8.GetString(utf8Bytes); // Correctly decode as UTF-8
            MessageBox.Show(goodDecode, "Correctly decoded");

            // Recovering from bad decode...
            byte[] originalBytes = Windows1252.GetBytes(badDecode);
            goodDecode = Utf8.GetString(originalBytes);
            MessageBox.Show(goodDecode, "Re-decoded");
        }
    }
}

Even with correct decoding, you'll still need a font that supports the characters being displayed. If your default font doesn't support Chinese, you still might not see the correct characters.

The correct thing to do is figure out why the string you have was decoded as Windows-1252 in the first place. Sometimes, though, data in a database is stored incorrectly to begin with and you have to resort to these games to fix the problem.

回答2:

string test = "敭畳灴獩楫n"; //incoming data. must be mesutpiskin 

byte[] bytes = Encoding.Unicode.GetBytes(test);

string s = string.Empty;

for (int i = 0; i < bytes.Length; i++)
{
    s += (char)bytes[i];
}

s = s.Trim((char)0);

MessageBox.Show(s);
//s=mesutpiskin 

回答3:

I'm not really sure what you mean, but I'm guessing you want to convert between a string in a certain encoding in byte array form and a string. Let's assume the character encoding is called "FooBar":

This is how you encode and decode:

Encoding myEncoding = Encoding.GetEncoding("FooBar");
string myString = "lala";
byte[] myEncodedBytes = myEncoding.GetBytes(myString);
string myDecodedString = myEncoding.GetString(myEncodedBytes);

You can learn more about the Encoding class over at MSDN.

回答4:

Answering your question at the end of your post:

1. If you want to determine the text encoding on runtime you should look at that: http://code.google.com/p/ude/

2. for converting character sets you can use http://msdn.microsoft.com/en-us/library/system.text.encoding.convert(v=vs.100).aspx

回答5:

It's Windows Latin 1. I pasted the Chinese text as UTF-8 into BBEDIT (a text editor for Mac) and re-opened the file as Windows Latin 1 and bang, the exact diacritics appeared.