Find any things except one words by QRegularExpression

tham

String want to parse

//there are many similar strings in the file, this is one of them
<div class="rg_meta">{"id":"IuDOvhrPwcs2WM:","isu":"healthline.com","itg":0,"ity":"jpg","oh":728,"ou":"http://www.healthline.com/hlcmsresource/images/News/general-health/071916_unitsmoking_BODY.jpg","ow":1296,"pt":"Secondhand Smoke

Expected result

<div class="rg_meta">{"id":"IuDOvhrPwcs2WM:","isu":"healthline.com","itg":0,"ity":"jpg","oh":728,"ou"

Real result is no match

My solution

QRegularExpression re("<div class=\"rg_meta\">{^(?!\"ou\").*\"ou\"");
auto iter = re.globalMatch(contents);
while(iter.hasMatch()){
  auto match = iter.next();
  qDebug()<<match.match(0);
}else{
  qDebug()<<"no match";
}

How should I fix this error?Thanks

tham

There are work around for my requirement, my ultimate goal is get the images address(ou and tu)

QRegularExpression reg("<div class=\"rg_meta\">{[^}]*}");
auto iter = reg.globalMatch(contents);        
while(iter.hasNext()){  
    QRegularExpressionMatch match = iter.next();
    QRegularExpression link_big("\"ou\":\"([^\"]*)");
    QRegularExpression link_small("\"tu\":\"([^\"]*)");
    auto const bm = link_big.match(match.captured(0));
    auto const sm = link_small.match(match.captured(0));
    qDebug()<<bm.captured(0)<<","<<sm.captured(0);
}

This solution work, but I want to know why look ahead solution fail, how could I do it correctly in Qt?Thanks

SGaist

Hi,

Shouldn't it rather be something like <div class=\"rg_meta\".+(?>"ou") or <div class=\"rg_meta\">[a-zA-Z0-9{":,.\/]+(?>"ou") ?

tham

Thanks, you are right, I have wrong understanding about negative look ahead of regex.

Looks like there are not easy way to express

match some string > match anything but a pattern

by regular expression. The "work around" I use maybe a nice choice for this problem.

tham

I mark this post as solved, if anybody come out better solution, please post at here, thanks