Extract a QString with a regular expression.

Patou355

Hi all!

I'm trying to do the followin :
I have a QString str containing exactly what follows :
[1] "Johnny Clegg" "B" "C" "D" "E" "F" "G" "H"
And I want to get a QStringList containing :

Johnny Clegg
B
C
D
E
F
G
H

the split function doesn't match what I need because I want to extract a pattern and not split from a pattern. The problem with splitting is that the first part containing [1] " will be with Johnny Clegg in the first item of the QStringList

So the question: how to make a QStringList from a pattern extraction in a simple way? maybe as simple as QString::split ?

mrjj

@Patou355
There is no way you can get a comma between the fields?

Patou355

I could search and replace looking for " " but it won't solve this problem of the [1] at the beginning. I't a solution I explored before... no success :/

mrjj

@Patou355
well im not sure how complex the input will be.
For a fast solution, i would just split at space and check the entries.
If not ending in ' it was falsely split and should be merged with next.
if format always is like shown. its very easy to fix up.

J.Hilk

@Patou355
Do exactly that and from the resulting QStringlist, simply remove the first entry.

myList.removeFirst();

SGaist

Hi,

Do you mean the pattern is one uppercased letter between two single quotes ?

Patou355

@SGaist the pattern can be anything between the double quotes. That's why I replaced the "A" with "Johnny Clegg" ;)
I could split on the spaces but the issue here is that I would get
Johnny
Clegg
B
C
and so on

instead of the
Johnny Clegg
B
C
and so on
that I want.

Patou355

@J.Hilk
The problem is that I don't always have a [1] at the beginning. Sometimes I have nothing. The cleanest solution in my case is actually extracting instead of splitting...

J.Hilk

@Patou355
ok, how about this:

Search the string for [ and ] and remove those and everything in between

int start = myString.indexOf('[');
int end = myString.indexOf(']');

if(start != -1 && end != -1){
myString.replace(start, end-start, '');
}

Now split over ' '.

SGaist

Something like ?

QRegularExpression rx("\"(\\w*\\s?\\w*)\"");
QRegularExpressionMatchIterator rxIterator = rx.globalMatch(yourCoolString);
QStringList words;
while (rxIterator.hasNext()) {
    QRegularExpressionMatch match = rxIterator.next();
    QString word = match.captured(1);
    words << word;
}

[edit: Fixed double quote escaping SGaist]

Patou355

I wrote this function which looks a lot like @SGaist's one and that works:

    /*Create a QStringList from the extraction of a QRegularExpression*/
    static QStringList extractStr(const QString &s, QRegularExpression delim){
        QStringList output;
        QString buff;
        QRegularExpressionMatchIterator i = delim.globalMatch(s);
        while (i.hasNext()) {
            QRegularExpressionMatch match = i.next();
            if (match.hasMatch()) {
                buff = match.captured(0);
                output.push_back(buff);
            }
        }
        return output;
    }

Thanks to all of you for your help :)
Patrick.

SGaist

One small note, you are wasting CPU cycle here. The QRegularExpressionMatchIterator returns only match objects that contains a match thus the if is useless.