QHebrewCodec Class Reference

Detailed Description

The QHebrewCodec class provides conversion to and from visually ordered Hebrew.

Hebrew as a semitic language is written from right to left. Because older computer systems couldn't handle reordering a string so that the first letter appears on the right, many older documents were encoded in visual order, so that the first letter of a line is the rightmost one in the string.

In contrast to this, Unicode defines characters to be in logical order (the order you would read the string). This codec tries to convert visually ordered Hebrew (8859-8) to Unicode. This might not always work perfectly, because reversing the bidi (bi-directional) algorithm that transforms from logical to visual order is non-trivial.

Transformation from Unicode to visual Hebrew (8859-8) is done using the bidi algorithm in Qt, and will produce correct results, so long as the codec is given the text a whole paragraph at a time. Places where newlines are supposed to go can be indicated by a newline character ('\n'). Note that these newline characters change the reordering behaviour of the algorithm, since the bidi reordering only takes place within one line of text, whereas line breaks are determined in visual order.

Visually ordered Hebrew is still used quite often in some places, mainly in email communication (since most email programs still don't understand logically ordered Hebrew) and on web pages. The use on web pages is rapidly decreasing, due to the availability of browsers that correctly support logically ordered Hebrew.

This codec has the name "iso8859-8". If you don't want any bidi reordering to happen during conversion, use the "iso8859-8-i" codec, which assumes logical order for the 8-bit string.

See also Internationalization with Qt.

Member Function Documentation

QCString QHebrewCodec::fromUnicode ( const QString & uc, int & lenInOut ) const [virtual]

The algorithm is designed to work on whole paragraphs of text, so processing a line at a time may produce incorrect results. This approach is taken because the reordering of the contents of a particular line in a paragraph may depend on the previous line in the same paragraph.

Some encodings (for example Japanese or UTF-8) are multibyte (so one input character is mapped to two output characters). The lenInOut argument specifies the number of QChars that should be converted and is set to the number of characters returned.

Reimplemented from QTextCodec.

const char * QHebrewCodec::mimeName () const [virtual]

Reimplemented from QTextCodec.

This file is part of the Qt toolkit. Copyright © 1995-2005 Trolltech. All Rights Reserved.