Qt regular expressions!!

[notsonerdysunny]

I was wondering if there is are ways to acess the sub-matches apart from the complete match when we do regular-expression matching. for instance,

"^[a-z](\d+)" when mached with "abcd1243abc" should not only return the final match i.e "d1234" but also give me access to sub-match "1234" which corresponds to (\d+) which can be done in perl and other scripting languages which support regular expressions using variables like $1, $2 .. etc.

[jacek]

QRegExp::cap()

[fullmetalcoder]

QRegExp::capturedTexts() is another solution. The choice of either of these depends on what you want to do... the aforementioned returns a single item at a given index whereas this one returns a list of all captured texts...

[patrik08]

I was wondering if there is are ways to acess the sub-matches apart from the complete match when we do regular-expression matching. for etc.

Wath is sub-matches ... on italian , german is this like find each find image on tag .. src="image.png" ?? on text....?

like this smal piece to replace qt uncool text style css....

/*  find style,lang,class on a html file && append o a list && and end remove....*/
QString TidyExternalHtml( QString body )
{
    QString prehtml = TidyCleanhtml(body);  /* base clean to stay on minimal standard xhtml  and capable to find .... */
    QStringList notneed;
    notneed.clear();
    ///////////////width="456" lang  class
    QRegExp expression( "width=[\"\'](.*)[\"\']", Qt::CaseInsensitive );  /* table td tr width image amen */
    QRegExp expression2( "style=[\"\'](.*)[\"\']", Qt::CaseInsensitive );
    QRegExp expression3( "lang=[\"\'](.*)[\"\']", Qt::CaseInsensitive );
    QRegExp expression4( "class=[\"\'](.*)[\"\']", Qt::CaseInsensitive );
 
 
    expression.setMinimal(true);
    expression2.setMinimal(true);
    expression3.setMinimal(true);
    expression4.setMinimal(true);
 
    int iPosition = 0;
    while( (iPosition = expression.indexIn( prehtml , iPosition )) != -1 ) {
        QString semi1 = expression.cap( 1 );
        notneed.append(QString("width=\"%1\"").arg(semi1));
        notneed.append(QString("width='%1'").arg(semi1));
        iPosition += expression.matchedLength();
    }
 
    iPosition = 0;
       while( (iPosition = expression2.indexIn( prehtml , iPosition )) != -1 ) {
        QString semi2 = expression2.cap( 1 );
        notneed.append(QString("style=\"%1\"").arg(semi2));
        notneed.append(QString("style='%1'").arg(semi2));
        iPosition += expression2.matchedLength();
    }
 
    iPosition = 0;
       while( (iPosition = expression3.indexIn( prehtml , iPosition )) != -1 ) {
        QString semi3 = expression3.cap( 1 );
        notneed.append(QString("lang=\"%1\"").arg(semi3));
        notneed.append(QString("lang='%1'").arg(semi3));
        iPosition += expression3.matchedLength();
    }
 
     iPosition = 0;
       while( (iPosition = expression4.indexIn( prehtml , iPosition )) != -1 ) {
        QString semi4 = expression4.cap( 1 );
        notneed.append(QString("class=\"%1\"").arg(semi4));
        notneed.append(QString("class='%1'").arg(semi4));
        iPosition += expression4.matchedLength();
    }
 
    for (int i = 0; i < notneed.size(); ++i)  {
          const QString fluteremove = notneed.at(i);
          prehtml = prehtml.replace(fluteremove,"", Qt::CaseInsensitive );
    }
 
    return prehtml;
}

/*  find style,lang,class on a html file && append o a list && and end remove....*/
QString TidyExternalHtml( QString body )
{
    QString prehtml = TidyCleanhtml(body);  /* base clean to stay on minimal standard xhtml  and capable to find .... */
    QStringList notneed;
    notneed.clear();
    ///////////////width="456" lang  class
    QRegExp expression( "width=[\"\'](.*)[\"\']", Qt::CaseInsensitive );  /* table td tr width image amen */
    QRegExp expression2( "style=[\"\'](.*)[\"\']", Qt::CaseInsensitive );
    QRegExp expression3( "lang=[\"\'](.*)[\"\']", Qt::CaseInsensitive );
    QRegExp expression4( "class=[\"\'](.*)[\"\']", Qt::CaseInsensitive );
    
    
    expression.setMinimal(true);
    expression2.setMinimal(true);
    expression3.setMinimal(true);
    expression4.setMinimal(true);
    
    int iPosition = 0;
    while( (iPosition = expression.indexIn( prehtml , iPosition )) != -1 ) {
        QString semi1 = expression.cap( 1 );
        notneed.append(QString("width=\"%1\"").arg(semi1));
        notneed.append(QString("width='%1'").arg(semi1));
        iPosition += expression.matchedLength();
    }
    
    iPosition = 0;
       while( (iPosition = expression2.indexIn( prehtml , iPosition )) != -1 ) {
        QString semi2 = expression2.cap( 1 );
        notneed.append(QString("style=\"%1\"").arg(semi2));
        notneed.append(QString("style='%1'").arg(semi2));
        iPosition += expression2.matchedLength();
    }
    
    iPosition = 0;
       while( (iPosition = expression3.indexIn( prehtml , iPosition )) != -1 ) {
        QString semi3 = expression3.cap( 1 );
        notneed.append(QString("lang=\"%1\"").arg(semi3));
        notneed.append(QString("lang='%1'").arg(semi3));
        iPosition += expression3.matchedLength();
    }
    
     iPosition = 0;
       while( (iPosition = expression4.indexIn( prehtml , iPosition )) != -1 ) {
        QString semi4 = expression4.cap( 1 );
        notneed.append(QString("class=\"%1\"").arg(semi4));
        notneed.append(QString("class='%1'").arg(semi4));
        iPosition += expression4.matchedLength();
    }
    
    for (int i = 0; i < notneed.size(); ++i)  {
          const QString fluteremove = notneed.at(i);
          prehtml = prehtml.replace(fluteremove,"", Qt::CaseInsensitive );
    }
    
    return prehtml;
}

To copy to clipboard, switch view to plain text mode 

[fullmetalcoder]

You're parsing HTML right? So why don't you use Qt XML processing facilities??? HTML is nothing but specialized XML...

[elcuco]

HTML is not XML. See this example:

wrong nesting of nodes:

<div><b>a</div></b>

<div><b>a</div></b>

To copy to clipboard, switch view to plain text mode 

not closing nodes (fint with html 4.01)

<ul>
   <li>one
   <li>two
</il>
<br>

<ul>
   <li>one
   <li>two
</il>
<br>

To copy to clipboard, switch view to plain text mode 

not using quotes on properties:

<img src=test.jpg border=0></img>

<img src=test.jpg border=0></img>

To copy to clipboard, switch view to plain text mode 

[patrik08]

HTML is not XML. See this example:
<img src=test.jpg border=0></img>
[/code]

Right .... so many format html the best is XHTML .... tidy can handle .....

qtclass
https://qt-webdav.svn.sourceforge.ne...T4_doc/qtidy.h
static lib .... svn co
https://qt-webdav.svn.sourceforge.ne.../lib_tidy_src/ path...

if i grab image on regex or on dom xml as xhtml is work all two.....

But have a look on word or openoffice or other programm... the copy fragment...
this is horror...

Wenn i import on copy paste extern text ... evryting go before on tidy .... otherwise i can not find image and link ..... && i remove extern class && table widht .... from A4 or other
and reformat to new ....


tidiconfigfile.append("output-xhtml: yes");
tidiconfigfile.append("clean: yes");
tidiconfigfile.append("wrap: 550");
tidiconfigfile.append("indent-spaces: 1");
tidiconfigfile.append("char-encoding: utf8");
tidiconfigfile.append("output-encoding: utf8");
tidiconfigfile.append("wrap-attributes: yes");
tidiconfigfile.append("doctype: yes");
tidiconfigfile.append("hide-comments: yes");
tidiconfigfile.append("numeric-entities: yes");
tidiconfigfile.append("drop-proprietary-attributes: yes");
tidiconfigfile.append("word-2000: yes");
tidiconfigfile.append("bare: yes");
//////tidiconfigfile.append("show-body-only: yes"); /* only body checks */

void QVimedit::insertFromMimeData ( const QMimeData * source )
{
 
    ////////////////QTextEdit::insertFromMimeData(source);
    if ( source->formats().contains("text/html") ) {
        ////////qDebug() << "### incomming paste text/html  ";
        const  QString tidicaches = QString("%2/.qtidy/").arg(QDir::homePath());
        QString draghtml = source->html();
        /* fwriteutf8(QString fullFileName,QString xml) */
        QTidy *tidy = new QTidy();   /* QTidy  *tidy; */
               tidy->Init(tidicaches);  /* tidy cache remove on last event */
        const  QString xhtmlnew = tidy->TidyExternalHtml(draghtml);
        ///////fwriteutf8("copy_in.html",xhtmlnew);
         QTextDocumentFragment fragment = QTextDocumentFragment::fromHtml(xhtmlnew);
         textCursor().insertFragment(fragment);   
         emit IncommingHTML();  /* signal to reload image on resource...  QTextDocument */
        return;
    }
}
 
if hasimage ........
 
if url .....

void QVimedit::insertFromMimeData ( const QMimeData * source )
{

    ////////////////QTextEdit::insertFromMimeData(source);
    if ( source->formats().contains("text/html") ) {
        ////////qDebug() << "### incomming paste text/html  ";
        const  QString tidicaches = QString("%2/.qtidy/").arg(QDir::homePath());
        QString draghtml = source->html();
        /* fwriteutf8(QString fullFileName,QString xml) */
        QTidy *tidy = new QTidy();   /* QTidy  *tidy; */
               tidy->Init(tidicaches);  /* tidy cache remove on last event */
        const  QString xhtmlnew = tidy->TidyExternalHtml(draghtml);
        ///////fwriteutf8("copy_in.html",xhtmlnew);
         QTextDocumentFragment fragment = QTextDocumentFragment::fromHtml(xhtmlnew);
         textCursor().insertFragment(fragment);   
         emit IncommingHTML();  /* signal to reload image on resource...  QTextDocument */
        return;
    }
}

if hasimage ........

if url .....

To copy to clipboard, switch view to plain text mode 

[wysota]

AFAIK Qt uses xhtml.

[patrik08]

AFAIK Qt uses xhtml.

<br> this here is xhtml ? QTexEdit you can reload 2 or 6 time

QTextBrowser (on edit modus) or QTexEdit not handle to <br /> ....
only tidy handle that .... && asci or other tag....

[kiker99]

if you just want to use the captured groups in a replace, you can also use QString::replace ( const QRegExp & rx, const QString & after ). After can have '\\1', '\\2', etc items, which will be replaced with the corresponding capture groups.