Question

Like the title

1.what’s the difference between QString and QLatin1String??

2.when and where do I need to use one of them??

3.following:

QString str;
str = "";
str = QLatin1String("");

Is "" == QLatin1String("")??

Answer 1

QString holds unicode. A string literal “foo” is a byte sequence that could contain text in any encoding. When assigning a string literal to a QString, QString str = "foo", you implicitely convert from a byte sequence in undefined encoding to a QString holding unicode. The QString(const char*) constructor assumes ASCII and will convert as if you typed QString str = QString::fromAscii("foo"). That would break if you use non-ascii literals in your source files (e.g., japanese string literals in UTF-8) or pass character data from a char* or QByteArray you read from elsewhere (a file, socket, etc.). Thus it’s good practice to keep the unicode QString world and the byte array QByteArray/char* world separated and only convert between those two explicitly, clearly stating which encoding you want to use to convert between those two. One can define QT_NO_CAST_FROM_ASCII and QT_NO_CAST_TO_ASCII to enforce explicit conversions (I would always enable them when writing a parser of any sort).
Now, to assign a latin1 string literal to a QString variable using explicit conversion, one can use

QString foo = QString::fromLatin1("föö");

or

QString foo = QLatin1String("föö");

Both state that the literal is encoded in latin1 and allow “encoding-safe” conversions to unicode.
I find QLatin1String nicer to read and the QLatin1String docs explain why it will be also faster in some situations.

Wrapping string literals, or in some cases QByteArray or char* variables, holding latin1 data for conversion is the main use for QLatin1String, one wouldn’t use QLatin1String as method arguments, member variables or temporaries (all QString).