UTF-8, Java, and text output to an RTF

[montag]

I am coding my first commercial application and I'm working through bugs slowly but surely. Here's a problem I've run into:

I am having the character '▓' and others output to an RTF by this method:

FileWriter outputStream = new FileWriter("output.rtf");

When I open the output file in WordPad it prints a 'â–ˆ'. I have to use WordPad for this and I also have to have the block character. I'm coding a medical text-based application where the user will eventually look at a printout of a 96 well plate and line it up and the dark boxes represent a positive read and the light boxes represent negative reads. I have to use those because other characters aren't dark enough.

Any advice? I'm coding in eclipse.

Also, can you make it output with default font other than Times New?

I also have noticed in WordPad when I open the output file instead of showing a picture of a colored rtf in the tile bar it's just a regular text picture next to "output - Wordpad" on the top left. This only happens when I open this rtf and not another rtf I create in WordPad.

[Norm]

What is the hex/binary value of the block character?
Does Wordpad have any characters that it will display that looks like the block character?

[montag]

U+2588 █ e2 96 88 FULL BLOCK

I guess the hex value is e2 96 88 according to Unicode/UTF-8-character table

WordPad will display this character only if I copy and paste it in or change my outputStream to .doc and then save it as an rtf.

[Norm]

I copy and pasted the first line of your last post into Wordpad and saved it as Unicode text
and got the following in the .txt file:

FFFE 55002B0032003500380038000900
8825 0900650032002000390036002000
3800 38000900460055004C004C002000
4200 4C004F0043004B00

It looks like the block character was written as 0x8825
Which is 2588 with reversed bytes.

[montag]

What does that mean exactly? It's reversing my unicode characters?

[Norm]

Here's some code that will write a block to a txt file:

      // UTF-8 rules
      // 1110xxxx	10xxxxxx	10xxxxxx   for 16 bits
      // 11101234	10567890	10123456   the bit positions(1 based)   takes 3 bytes for 1 char

      // UTF-8 character U+2588	¦	e2 96 88	FULL BLOCK
      char[] data = {'\u3e00', '\u2588', '<'}; 
      // UTF16 - FEFF 003E 8825 003C       <<< This shows the BLOCK in Wordpad with sqrs on each side
      // UTF16 - FEFF 3E00 8825 003C       <<<< Shows > on left side of BLOCK
      // UTF8 -  3E  E8 A0A5 3C            <<< This shows: >è*¥<
      // UTF8 -  FEFF 3E00 8825 003C       <<<< This shows as > Block and a sqr !!!!

      OutputStreamWriter out  = new OutputStreamWriter(new FileOutputStream("BlockUnicode.txt"), "UTF8");
      out.write(data, 0, data.length);
      out.close();

[montag]

Thanks man. That helped a lot, although when I open it and run it in WordPad it still gives me those random characters. I ended up looking around thinking WordPad was the problem and I found AbiWord. I'm going to use AbiWord which is open source and seems a lot better than WordPad. The formatting is the exact same (unlike OpenOffice and MS Word) so I'm a fan, and it prints the blocks. Thanks again! Kudos.

[montag]

Also how do I set the default output font to Courier New at 12 pt?

[montag]

So I've come back with a new problem. I have virtually completed the backend and made it possible to open a text file with the correct UTF-8 formatting so I can see the black squares. I've since started my first GUI frontend for this. Now I'm trying to have it open in a JTextPane and I am not sure how to specify UTF-8 formatting again so I'm running into the same problem. Any advice? I'm going to keep doing research after lunch to see if I can figure it out. Would I have the FileReader specify UTF-8?

Thanks again,

montag

[montag]

I figured it out. It was all due to the reader and me not specifying UTF8.

Reader in = new InputStreamReader(new FileInputStream(file), "UTF-8");
               
               //Reader in = new FileReader(file);
               
               char[] buff = new char[40000096];
               int nch;
               while((nch = in.read(buff, 0, buff.length)) != -1)
               {
                  //jTextPane0.setDocument(new StyledDocument());
                  jTextPane0.setText(new String(buff, 0, nch));
               }

I have yet to fix the buffer size. Thanks again Norm.