QString to ASCII QByteArray

[kshegunov]

Hello,
I need to convert a QString data into an ASCII containing QByteArray (I need to have the ^Z character in it). In Qt 4 there was the QString::toAscii method however it was deprecated in favor of QString::toLatin1. There's note put in the documentation about said method, though:

The returned byte array is undefined if the string contains non-Latin1 characters. Those characters may be suppressed or replaced with a question mark.

As far as I know, the latin1 charset (ISO/IEC 8859-1:1998) doesn't specify anything about the codes for non-printable characters under 0x20. So my question boils down to:
Can I use QString::toLatin1, and will my ^Z character be preserved?

Kind regards.

[Paul Colby]

@kshegunov said:

I need to convert a QString data into an ASCII containing QByteArray (I need to have the ^Z character in it).

Just an idea: why not use QString::toUtf8? Since all UTF-8 characters less than 128 are the same as ASCII, then that should give you your ^Z character correctly.

If your string happens to contain characters greater than 127, then what would you expect to get in the QByteArray? Latin-1? UTF-8? Something else? Where is the QByteArray going / being used?

Depending on your answers, its possible that what you really want is one or more of:

• QString::toLocal8Bit
• QString::toUtf8
• not use QString at all, but start and end with QByteArray (if you're dealing with binary data).

In Qt 4 there was the QString::toAscii method however it was deprecated in favor of QString::toLatin1.

For what its worth, here's what source code for QString::toAscii looked like in Qt4:

QByteArray QString::toAscii() const
{
#ifndef QT_NO_TEXTCODEC
    if (codecForCStrings)
        return codecForCStrings->fromUnicode(*this);
#endif // QT_NO_TEXTCODEC
    return toLatin1();
}

So you were probably getting either equivalent behaviour to QString::toLocal8Bit, or QString::toLatin1? No wonder the function was deprecated... its behaviour is pretty ambiguous, and host-configuration dependant (when you look into the default text codec).

Cheers.

[kshegunov]

@Paul-Colby
That was fast! ;)

Just an idea: why not use QString::toUtf8? Since all UTF-8 characters less than 128 are the same as ASCII, then that should give you your ^Z character correctly.

Works for me, I don't know why I didn't think of that in the first place ... :D

If your string happens to contain characters greater than 127

It doesn't.

Something else? Where is the QByteArray going / being used?

It's going through the serial port, but I am dealing with text data. (QSerialPort expects a QByteArray, and the device accepts ASCII data, possibly other charsets but I don't want to deal with them ...)

No wonder the function was deprecated... its behaviour is pretty ambiguous, and host-configuration dependant

QString::toLocal8Bit is host-configuration dependent as well, but as I said I'm dealing with latin text only + a few non-printables (under 0x20).

Thanks for the suggestion(s)! I'll stick to utf8 and everything should be just fine.
Kind regards.

[andr]

If .toAscii() worked for you before, you can savely use .toLatin1(), too, and ^Z will be preserved.