Unicode

From Linuxintro
Revision as of 07:10, 16 May 2009 by imported>ThorstenStaerk

Understanding

Clearly, every text file has an encoding, that means, you must know if two bytes form one character to display, one byte, or the characters have mixed byte length. Unicode defines every character in the world.

Here is some practice: Store a file containing

hellö world

in file.txt. Do:

tweedleburg:~ # cat >file.txt
hellö world
tweedleburg:~ # cat file.txt
hellö world
tweedleburg:~ # hexdump -C file.txt
00000000  68 65 6c 6c c3 b6 20 77  6f 72 6c 64 0a           |hell.. world.|
0000000d

This means, every "normal" character has been stored in 1 byte, every umlaut in 2 bytes. That is unicode's UTF-8 encoding

To show what Qt understands when it reads UTF8, we store a file with the content

ü

nothing else. The following code outputs the code:

QFile inputfile(args->url(0).fileName());
inputfile.open(QIODevice::ReadOnly);
inputfilecontent = inputfile.read(inputfile.bytesAvailable());
kDebug() << "inputfilecontent.data()[0]"<<(byte)inputfilecontent.data()[0];
kDebug() << "inputfilecontent.data()[1]"<<(byte)inputfilecontent.data()[1];

For little endian systems, ü UTF8 encoded delivers

195 
188

Doing

Convert a file to UTF-8

convmv -f iso-8859-1 -t utf8 -r --notest <datei>
recode latin1..u8 <datei>

Unicode text editor:

  • yudit