QString::fromUcs4() eats ZERO WIDTH NO-BREAK SPACE

QString::fromUcs4() internally calls Utf32::convertToUnicode(), which defaults to detect endianess from the data (which eats leading BOM).

But QString::fromUcs4() always uses the machine endianess, so no header parsing is needed.

I suggest to call it with QTextCodec::IgnoreHeader flag set.

Test case:

uint chr = 0xFEFF;
qDebug << QString::fromUcs4(&chr, 1).toUtf8();

shows an empty QByteArray.