linklocator.cpp Source File

kpimutils

 /*
   Copyright (c) 2002 Dave Corrie <kde@davecorrie.com>
 
   This library is free software; you can redistribute it and/or
   modify it under the terms of the GNU Library General Public
   License as published by the Free Software Foundation; either
   version 2 of the License, or (at your option) any later version.
 
   This library is distributed in the hope that it will be useful,
   but WITHOUT ANY WARRANTY; without even the implied warranty of
   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
   Library General Public License for more details.
 
   You should have received a copy of the GNU Library General Public License
   along with this library; see the file COPYING.LIB.  If not, write to
   the Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor,
   Boston, MA 02110-1301, USA.
 */
 #include "linklocator.h"
 
 #include <KEmoticons>
 
 #include <QtCore/QCoreApplication>
 #include <QtCore/QFile>
 #include <QtCore/QRegExp>
 #include <QTextDocument>
 
 #include <climits>
 
 using namespace KPIMUtils;
 
 //@cond PRIVATE
 class KPIMUtils::LinkLocator::Private
 {
   public:
     int mMaxUrlLen;
     int mMaxAddressLen;
 };
 //@endcond
 
 // Use a static for this as calls to the KEmoticons constructor are expensive.
 K_GLOBAL_STATIC( KEmoticons, sEmoticons )
 
 LinkLocator::LinkLocator( const QString &text, int pos )
   : mText( text ), mPos( pos ), d( new KPIMUtils::LinkLocator::Private )
 {
   d->mMaxUrlLen = 4096;
   d->mMaxAddressLen = 255;
 
   // If you change either of the above values for maxUrlLen or
   // maxAddressLen, then please also update the documentation for
   // setMaxUrlLen()/setMaxAddressLen() in the header file AND the
   // default values used for the maxUrlLen/maxAddressLen parameters
   // of convertToHtml().
 }
 
 LinkLocator::~LinkLocator()
 {
   delete d;
 }
 
 void LinkLocator::setMaxUrlLen( int length )
 {
   d->mMaxUrlLen = length;
 }
 
 int LinkLocator::maxUrlLen() const
 {
   return d->mMaxUrlLen;
 }
 
 void LinkLocator::setMaxAddressLen( int length )
 {
   d->mMaxAddressLen = length;
 }
 
 int LinkLocator::maxAddressLen() const
 {
   return d->mMaxAddressLen;
 }
 
 QString LinkLocator::getUrl()
 {
   QString url;
   if ( atUrl() ) {
     // NOTE: see http://tools.ietf.org/html/rfc3986#appendix-A and especially appendix-C
     // Appendix-C mainly says, that when extracting URLs from plain text, line breaks shall
     // be allowed and should be ignored when the URI is extracted.
 
     // This implementation follows this recommendation and
     // allows the URL to be enclosed within different kind of brackets/quotes
     // If an URL is enclosed, whitespace characters are allowed and removed, otherwise
     // the URL ends with the first whitespace
     // Also, if the URL is enclosed in brackets, the URL itself is not allowed
     // to contain the closing bracket, as this would be detected as the end of the URL
 
     QChar beforeUrl, afterUrl;
 
     // detect if the url has been surrounded by brackets or quotes
     if ( mPos > 0 ) {
       beforeUrl = mText[mPos - 1];
 
       /*if ( beforeUrl == '(' ) {
         afterUrl = ')';
       } else */if ( beforeUrl == QLatin1Char('[') ) {
         afterUrl = QLatin1Char(']');
       } else if ( beforeUrl == QLatin1Char('<') ) {
         afterUrl = QLatin1Char('>');
       } else if ( beforeUrl == QLatin1Char('>') ) { // for e.g. <link>http://.....</link>
         afterUrl = QLatin1Char('<');
       } else if ( beforeUrl == QLatin1Char('"') ) {
         afterUrl = QLatin1Char('"');
       }
     }
 
     url.reserve( maxUrlLen() );  // avoid allocs
     int start = mPos;
     while ( ( mPos < (int)mText.length() ) &&
             ( mText[mPos].isPrint() || mText[mPos].isSpace() ) &&
             ( ( afterUrl.isNull() && !mText[mPos].isSpace() ) ||
               ( !afterUrl.isNull() && mText[mPos] != afterUrl ) ) ) {
       if ( !mText[mPos].isSpace() ) {   // skip whitespace
         url.append( mText[mPos] );
         if ( url.length() > maxUrlLen() ) {
           break;
         }
       }
 
       mPos++;
     }
 
     if ( isEmptyUrl( url ) || ( url.length() > maxUrlLen() ) ) {
       mPos = start;
       url.clear();
     } else {
       --mPos;
     }
   }
 
   // HACK: This is actually against the RFC. However, most people don't properly escape the URL in
   //       their text with "" or <>. That leads to people writing an url, followed immediatley by
   //       a dot to finish the sentence. That would lead the parser to include the dot in the url,
   //       even though that is not wanted. So work around that here.
   //       Most real-life URLs hopefully don't end with dots or commas.
   QList<QChar> wordBoundaries;
   wordBoundaries << QLatin1Char('.') << QLatin1Char(',') << QLatin1Char(':') << QLatin1Char('!') << QLatin1Char('?') << QLatin1Char(')') << QLatin1Char('>');
   if ( url.length() > 1 ) {
     do {
       if ( wordBoundaries.contains( url.at( url.length() - 1 ) ) ) {
         url.chop( 1 );
         --mPos;
       } else {
         break;
       }
     } while( url.length() > 1 );
   }
 
   return url;
 }
 
 // keep this in sync with KMMainWin::slotUrlClicked()
 bool LinkLocator::atUrl() const
 {
   // the following characters are allowed in a dot-atom (RFC 2822):
   // a-z A-Z 0-9 . ! # $ % & ' * + - / = ? ^ _ ` { | } ~
   static const QString allowedSpecialChars = QLatin1String( ".!#$%&'*+-/=?^_`{|}~" );
 
   // the character directly before the URL must not be a letter, a number or
   // any other character allowed in a dot-atom (RFC 2822).
   if ( ( mPos > 0 ) &&
        ( mText[mPos-1].isLetterOrNumber() ||
          ( allowedSpecialChars.indexOf( mText[mPos-1] ) != -1 ) ) ) {
     return false;
   }
 
   QChar ch = mText[mPos];
   return
     ( ch == QLatin1Char('h') && ( mText.mid( mPos, 7 ) == QLatin1String( "http://" ) ||
                      mText.mid( mPos, 8 ) == QLatin1String( "https://" ) ) ) ||
     ( ch == QLatin1Char('v') && mText.mid( mPos, 6 ) == QLatin1String( "vnc://" ) ) ||
     ( ch == QLatin1Char('f') && ( mText.mid( mPos, 7 ) == QLatin1String( "fish://" ) ||
                      mText.mid( mPos, 6 ) == QLatin1String( "ftp://" ) ||
                      mText.mid( mPos, 7 ) == QLatin1String( "ftps://" ) ) ) ||
     ( ch == QLatin1Char('s') && ( mText.mid( mPos, 7 ) == QLatin1String( "sftp://" ) ||
                      mText.mid( mPos, 6 ) == QLatin1String( "smb://" ) ) ) ||
     ( ch == QLatin1Char('m') && mText.mid( mPos, 7 ) == QLatin1String( "mailto:" ) ) ||
     ( ch == QLatin1Char('w') && mText.mid( mPos, 4 ) == QLatin1String( "www." ) ) ||
     ( ch == QLatin1Char('f') && ( mText.mid( mPos, 4 ) == QLatin1String( "ftp." ) ||
                      mText.mid( mPos, 7 ) == QLatin1String( "file://" ) ) )||
     ( ch == QLatin1Char('n') && mText.mid( mPos, 5 ) == QLatin1String( "news:" ) );
 }
 
 bool LinkLocator::isEmptyUrl( const QString &url ) const
 {
   return url.isEmpty() ||
     url == QLatin1String( "http://" ) ||
     url == QLatin1String( "https://" ) ||
     url == QLatin1String( "fish://" ) ||
     url == QLatin1String( "ftp://" ) ||
     url == QLatin1String( "ftps://" ) ||
     url == QLatin1String( "sftp://" ) ||
     url == QLatin1String( "smb://" ) ||
     url == QLatin1String( "vnc://" ) ||
     url == QLatin1String( "mailto" ) ||
     url == QLatin1String( "www" ) ||
     url == QLatin1String( "ftp" ) ||
     url == QLatin1String( "news" ) ||
     url == QLatin1String( "news://" );
 }
 
 QString LinkLocator::getEmailAddress()
 {
   QString address;
 
   if ( mText[mPos] == QLatin1Char('@') ) {
     // the following characters are allowed in a dot-atom (RFC 2822):
     // a-z A-Z 0-9 . ! # $ % & ' * + - / = ? ^ _ ` { | } ~
     static const QString allowedSpecialChars = QLatin1String( ".!#$%&'*+-/=?^_`{|}~" );
 
     // determine the local part of the email address
     int start = mPos - 1;
     while ( start >= 0 && mText[start].unicode() < 128 &&
             ( mText[start].isLetterOrNumber() ||
               mText[start] == QLatin1Char('@') || // allow @ to find invalid email addresses
               allowedSpecialChars.indexOf( mText[start] ) != -1 ) ) {
       if ( mText[start] == QLatin1Char('@') ) {
         return QString(); // local part contains '@' -> no email address
       }
       --start;
     }
     ++start;
     // we assume that an email address starts with a letter or a digit
     while ( ( start < mPos ) && !mText[start].isLetterOrNumber() ) {
       ++start;
     }
     if ( start == mPos ) {
       return QString(); // local part is empty -> no email address
     }
 
     // determine the domain part of the email address
     int dotPos = INT_MAX;
     int end = mPos + 1;
     while ( end < (int)mText.length() &&
             ( mText[end].isLetterOrNumber() ||
               mText[end] == QLatin1Char('@') || // allow @ to find invalid email addresses
               mText[end] == QLatin1Char('.') ||
               mText[end] == QLatin1Char('-') ) ) {
       if ( mText[end] == QLatin1Char('@') ) {
         return QString(); // domain part contains '@' -> no email address
       }
       if ( mText[end] == QLatin1Char('.') ) {
         dotPos = qMin( dotPos, end ); // remember index of first dot in domain
       }
       ++end;
     }
     // we assume that an email address ends with a letter or a digit
     while ( ( end > mPos ) && !mText[end - 1].isLetterOrNumber() ) {
       --end;
     }
     if ( end == mPos ) {
       return QString(); // domain part is empty -> no email address
     }
     if ( dotPos >= end ) {
       return QString(); // domain part doesn't contain a dot
     }
 
     if ( end - start > maxAddressLen() ) {
       return QString(); // too long -> most likely no email address
     }
     address = mText.mid( start, end - start );
 
     mPos = end - 1;
   }
   return address;
 }
 
 QString LinkLocator::convertToHtml( const QString &plainText, int flags,
                                     int maxUrlLen, int maxAddressLen )
 {
   LinkLocator locator( plainText );
   locator.setMaxUrlLen( maxUrlLen );
   locator.setMaxAddressLen( maxAddressLen );
 
   QString str;
   QString result( (QChar*)0, (int)locator.mText.length() * 2 );
   QChar ch;
   int x;
   bool startOfLine = true;
 
   for ( locator.mPos = 0, x = 0; locator.mPos < (int)locator.mText.length();
         locator.mPos++, x++ ) {
     ch = locator.mText[locator.mPos];
     if ( flags & PreserveSpaces ) {
       if ( ch == QLatin1Char(' ') ) {
         if ( locator.mPos + 1 < locator.mText.length() ) {
           if ( locator.mText[locator.mPos + 1] != QLatin1Char(' ') ) {
 
             // A single space, make it breaking if not at the start or end of the line
             const bool endOfLine = locator.mText[locator.mPos + 1] == QLatin1Char('\n');
             if ( !startOfLine && !endOfLine ) {
               result += QLatin1Char(' ');
             } else {
               result += QLatin1String("&nbsp;");
             }
           } else {
 
             // Whitespace of more than one space, make it all non-breaking
             while ( locator.mPos < locator.mText.length() && locator.mText[locator.mPos] == QLatin1Char(' ') ) {
               result += QLatin1String("&nbsp;");
               locator.mPos++;
               x++;
             }
 
             // We incremented once to often, undo that
             locator.mPos--;
             x--;
           }
         } else {
           // Last space in the text, it is non-breaking
           result += QLatin1String("&nbsp;");
         }
 
         if ( startOfLine ) {
           startOfLine = false;
         }
         continue;
       } else if ( ch == QLatin1Char('\t') ) {
         do {
           result += QLatin1String("&nbsp;");
           x++;
         } while ( ( x & 7 ) != 0 );
         x--;
         startOfLine = false;
         continue;
       }
     }
     if ( ch == QLatin1Char('\n') ) {
       result += QLatin1String("<br />\n"); // Keep the \n, so apps can figure out the quoting levels correctly.
       startOfLine = true;
       x = -1;
       continue;
     }
 
     startOfLine = false;
     if ( ch == QLatin1Char('&') ) {
       result += QLatin1String("&amp;");
     } else if ( ch == QLatin1Char('"') ) {
       result += QLatin1String("&quot;");
     } else if ( ch == QLatin1Char('<') ) {
       result += QLatin1String("&lt;");
     } else if ( ch == QLatin1Char('>') ) {
       result += QLatin1String("&gt;");
     } else {
       const int start = locator.mPos;
       if ( !( flags & IgnoreUrls ) ) {
         str = locator.getUrl();
         if ( !str.isEmpty() ) {
           QString hyperlink;
           if ( str.left( 4 ) == QLatin1String("www.") ) {
             hyperlink = QLatin1String("http://") + str;
           } else if ( str.left( 4 ) == QLatin1String("ftp.") ) {
             hyperlink = QLatin1String("ftp://") + str;
           } else {
             hyperlink = str;
           }
 
           result += QLatin1String("<a href=\"") + hyperlink + QLatin1String("\">") + Qt::escape( str ) + QLatin1String("</a>");
           x += locator.mPos - start;
           continue;
         }
         str = locator.getEmailAddress();
         if ( !str.isEmpty() ) {
           // len is the length of the local part
           int len = str.indexOf( QLatin1Char('@') );
           QString localPart = str.left( len );
 
           // remove the local part from the result (as '&'s have been expanded to
           // &amp; we have to take care of the 4 additional characters per '&')
           result.truncate( result.length() -
                            len - ( localPart.count( QLatin1Char('&') ) * 4 ) );
           x -= len;
 
           result += QLatin1String("<a href=\"mailto:") + str + QLatin1String("\">") + str + QLatin1String("</a>");
           x += str.length() - 1;
           continue;
         }
       }
       if ( flags & HighlightText ) {
         str = locator.highlightedText();
         if ( !str.isEmpty() ) {
           result += str;
           x += locator.mPos - start;
           continue;
         }
       }
       result += ch;
     }
   }
 
   if ( flags & ReplaceSmileys ) {
     QStringList exclude;
     exclude << QLatin1String("(c)") << QLatin1String("(C)") << QLatin1String("&gt;:-(") << QLatin1String("&gt;:(") << QLatin1String("(B)") << QLatin1String("(b)") << QLatin1String("(P)") << QLatin1String("(p)");
     exclude << QLatin1String("(O)") << QLatin1String("(o)") << QLatin1String("(D)") << QLatin1String("(d)") << QLatin1String("(E)") << QLatin1String("(e)") << QLatin1String("(K)")<< QLatin1String("(k)");
     exclude << QLatin1String("(I)") << QLatin1String("(i)") << QLatin1String("(L)") << QLatin1String("(l)") << QLatin1String("(8)") << QLatin1String("(T)") << QLatin1String("(t)") << QLatin1String("(G)");
     exclude << QLatin1String("(g)") << QLatin1String("(F)") << QLatin1String("(f)") << QLatin1String("(H)");
     exclude << QLatin1String("8)") << QLatin1String("(N)") << QLatin1String("(n)") << QLatin1String("(Y)") << QLatin1String("(y)" )<< QLatin1String("(U)") << QLatin1String("(u)") << QLatin1String("(W)") << QLatin1String("(w)");
     static QString cachedEmoticonsThemeName;
     if ( cachedEmoticonsThemeName.isEmpty() ) {
       cachedEmoticonsThemeName = KEmoticons::currentThemeName();
     }
     result =
       sEmoticons->theme( cachedEmoticonsThemeName ).parseEmoticons(
         result, KEmoticonsTheme::StrictParse | KEmoticonsTheme::SkipHTML, exclude );
   }
 
   return result;
 }
 
 QString LinkLocator::pngToDataUrl( const QString &iconPath )
 {
   if ( iconPath.isEmpty() ) {
     return QString();
   }
 
   QFile pngFile( iconPath );
   if ( !pngFile.open( QIODevice::ReadOnly | QIODevice::Unbuffered ) ) {
     return QString();
   }
 
   QByteArray ba = pngFile.readAll();
   pngFile.close();
   return QString::fromLatin1( "data:image/png;base64,%1" ).arg( QLatin1String(ba.toBase64().constData()) );
 }
 
 QString LinkLocator::highlightedText()
 {
   // formating symbols must be prepended with a whitespace
   if ( ( mPos > 0 ) && !mText[mPos-1].isSpace() ) {
     return QString();
   }
 
   const QChar ch = mText[mPos];
   if ( ch != QLatin1Char('/') && ch != QLatin1Char('*') && ch != QLatin1Char('_') && ch != QLatin1Char('-') ) {
     return QString();
   }
 
   QRegExp re =
     QRegExp( QString::fromLatin1( "\\%1((\\w+)([\\s-']\\w+)*( ?[,.:\\?!;])?)\\%2" ).arg( ch ).arg( ch ) );
   re.setMinimal( true );
   if ( re.indexIn( mText, mPos ) == mPos ) {
     int length = re.matchedLength();
     // there must be a whitespace after the closing formating symbol
     if ( mPos + length < mText.length() && !mText[mPos + length].isSpace() ) {
       return QString();
     }
     mPos += length - 1;
     switch ( ch.toLatin1() ) {
     case '*':
       return QLatin1String("<b>*") + re.cap( 1 ) + QLatin1String("*</b>");
     case '_':
       return QLatin1String("<u>_") + re.cap( 1 ) + QLatin1String("_</u>");
     case '/':
       return QLatin1String("<i>/") + re.cap( 1 ) + QLatin1String("/</i>");
     case '-':
       return QLatin1String("<strike>-") + re.cap( 1 ) + QLatin1String("-</strike>");
     }
   }
   return QString();
 }
KDE's Doxygen guidelines are available online.
kpimutils

kpimutils

kdepimlibs API Reference

Search
