kmime_util.cpp

/*
  kmime_util.cpp

  KMime, the KDE Internet mail/usenet news message library.
  Copyright (c) 2001 the KMime authors.
  See file AUTHORS for details

  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 "kmime_util.h"
#include "kmime_util_p.h"

#include "kmime_charfreq.h"
#include "kmime_codecs.h"
#include "kmime_header_parsing.h"
#include "kmime_message.h"
#include "kmime_warning.h"

#include <config-kmime.h>
#include <kdefakes.h> // for strcasestr
#include <kglobal.h>
#include <klocale.h>
#include <klocalizedstring.h>
#include <kcharsets.h>
#include <kcodecs.h>
#include <kdebug.h>

#include <QtCore/QList>
#include <QtCore/QString>
#include <QtCore/QTextCodec>

#include <ctype.h>
#include <time.h>
#include <stdlib.h>
#include <unistd.h>

using namespace KMime;

namespace KMime {

QList<QByteArray> c_harsetCache;
QList<QByteArray> l_anguageCache;
QString f_allbackCharEnc;
bool u_seOutlookEncoding = false;

QByteArray cachedCharset( const QByteArray &name )
{
  foreach ( const QByteArray& charset, c_harsetCache ) {
    if ( qstricmp( name.data(), charset.data() ) == 0 ) {
      return charset;
    }
  }

  c_harsetCache.append( name.toUpper() );
  //kDebug() << "KNMimeBase::cachedCharset() number of cs" << c_harsetCache.count();
  return c_harsetCache.last();
}

QByteArray cachedLanguage( const QByteArray &name )
{
  foreach ( const QByteArray& language, l_anguageCache ) {
    if ( qstricmp( name.data(), language.data() ) == 0 ) {
      return language;
    }
  }

  l_anguageCache.append( name.toUpper() );
  //kDebug() << "KNMimeBase::cachedCharset() number of cs" << c_harsetCache.count();
  return l_anguageCache.last();
}

bool isUsAscii( const QString &s )
{
  uint sLength = s.length();
  for ( uint i=0; i<sLength; i++ ) {
    if ( s.at( i ).toLatin1() <= 0 ) { // c==0: non-latin1, c<0: non-us-ascii
      return false;
    }
  }
  return true;
}

QString nameForEncoding( Headers::contentEncoding enc )
{
  switch ( enc ) {
    case Headers::CE7Bit: return QString::fromLatin1( "7bit" );
    case Headers::CE8Bit: return QString::fromLatin1( "8bit" );
    case Headers::CEquPr: return QString::fromLatin1( "quoted-printable" );
    case Headers::CEbase64: return QString::fromLatin1( "base64" );
    case Headers::CEuuenc: return QString::fromLatin1( "uuencode" );
    case Headers::CEbinary: return QString::fromLatin1( "binary" );
    default: return QString::fromLatin1( "unknown" );
  }
}

QList<Headers::contentEncoding> encodingsForData( const QByteArray &data )
{
  QList<Headers::contentEncoding> allowed;
  CharFreq cf( data );

  switch ( cf.type() ) {
    case CharFreq::SevenBitText:
      allowed << Headers::CE7Bit;
    case CharFreq::EightBitText:
      allowed << Headers::CE8Bit;
    case CharFreq::SevenBitData:
      if ( cf.printableRatio() > 5.0/6.0 ) {
        // let n the length of data and p the number of printable chars.
        // Then base64 \approx 4n/3; qp \approx p + 3(n-p)
        // => qp < base64 iff p > 5n/6.
        allowed << Headers::CEquPr;
        allowed << Headers::CEbase64;
      } else {
        allowed << Headers::CEbase64;
        allowed << Headers::CEquPr;
      }
      break;
    case CharFreq::EightBitData:
      allowed << Headers::CEbase64;
      break;
    case CharFreq::None:
    default:
      Q_ASSERT( false );
  }

  return allowed;
}

// "(),.:;<>@[\]
const uchar specialsMap[16] = {
  0x00, 0x00, 0x00, 0x00, // CTLs
  0x20, 0xCA, 0x00, 0x3A, // SPACE ... '?'
  0x80, 0x00, 0x00, 0x1C, // '@' ... '_'
  0x00, 0x00, 0x00, 0x00  // '`' ... DEL
};

// "(),:;<>@[\]/=?
const uchar tSpecialsMap[16] = {
  0x00, 0x00, 0x00, 0x00, // CTLs
  0x20, 0xC9, 0x00, 0x3F, // SPACE ... '?'
  0x80, 0x00, 0x00, 0x1C, // '@' ... '_'
  0x00, 0x00, 0x00, 0x00  // '`' ... DEL
};

// all except specials, CTLs, SPACE.
const uchar aTextMap[16] = {
  0x00, 0x00, 0x00, 0x00,
  0x5F, 0x35, 0xFF, 0xC5,
  0x7F, 0xFF, 0xFF, 0xE3,
  0xFF, 0xFF, 0xFF, 0xFE
};

// all except tspecials, CTLs, SPACE.
const uchar tTextMap[16] = {
  0x00, 0x00, 0x00, 0x00,
  0x5F, 0x36, 0xFF, 0xC0,
  0x7F, 0xFF, 0xFF, 0xE3,
  0xFF, 0xFF, 0xFF, 0xFE
};

// none except a-zA-Z0-9!*+-/
const uchar eTextMap[16] = {
  0x00, 0x00, 0x00, 0x00,
  0x40, 0x35, 0xFF, 0xC0,
  0x7F, 0xFF, 0xFF, 0xE0,
  0x7F, 0xFF, 0xFF, 0xE0
};

void setFallbackCharEncoding(const QString& fallbackCharEnc)
{
  f_allbackCharEnc = fallbackCharEnc;
}

QString fallbackCharEncoding()
{
  return f_allbackCharEnc;
}

void setUseOutlookAttachmentEncoding( bool violateStandard )
{
  u_seOutlookEncoding = violateStandard;
}

bool useOutlookAttachmentEncoding()
{
  return u_seOutlookEncoding;
}


QString decodeRFC2047String( const QByteArray &src, QByteArray &usedCS,
                             const QByteArray &defaultCS, bool forceCS )
{
  QByteArray result;
  result.reserve(64);
  QByteArray spaceBuffer;
  spaceBuffer.reserve(64);
  const char *scursor = src.constData();
  const char *send = scursor + src.length();
  bool onlySpacesSinceLastWord = false;

  while ( scursor != send ) {
     // space
    if ( isspace( *scursor ) && onlySpacesSinceLastWord ) {
      spaceBuffer += *scursor++;
      continue;
    }

    // possible start of an encoded word
    if ( *scursor == '=' ) {
      QByteArray language;
      QString decoded;
      ++scursor;
      const char *start = scursor;
      if ( HeaderParsing::parseEncodedWord( scursor, send, decoded, language, usedCS, defaultCS, forceCS ) ) {
        result += decoded.toUtf8();
        onlySpacesSinceLastWord = true;
        spaceBuffer.clear();
      } else {
        if ( onlySpacesSinceLastWord ) {
          result += spaceBuffer;
          onlySpacesSinceLastWord = false;
        }
        result += '=';
        scursor = start; // reset cursor after parsing failure
      }
      continue;
    } else {
      // unencoded data
      if ( onlySpacesSinceLastWord ) {
        result += spaceBuffer;
        onlySpacesSinceLastWord = false;
      }
      result += *scursor;
      ++scursor;
    }
  }
  // If there are any chars that couldn't be decoded in UTF-8,
  //  use the fallback charset if it exists
  const QString tryUtf8 = QString::fromUtf8( result );
  if ( tryUtf8.contains( 0xFFFD ) && !f_allbackCharEnc.isEmpty() ) {
    QTextCodec* codec = KGlobal::charsets()->codecForName( f_allbackCharEnc );
    return codec->toUnicode( result );
  } else {
    return tryUtf8;
  }
}

QString decodeRFC2047String( const QByteArray &src )
{
  QByteArray usedCS;
  return decodeRFC2047String( src, usedCS, "utf-8", false );
}

static const char *reservedCharacters = "\"()<>@,.;:\\[]=";

QByteArray encodeRFC2047String( const QString &src, const QByteArray &charset,
                                bool addressHeader, bool allow8BitHeaders )
{
  QByteArray result;
  int start=0, end=0;
  bool nonAscii=false, ok=true, useQEncoding=false;

  // fromLatin1() is safe here, codecForName() uses toLatin1() internally
  const QTextCodec *codec = KGlobal::charsets()->codecForName( QString::fromLatin1( charset ), ok );

  QByteArray usedCS;
  if ( !ok ) {
    //no codec available => try local8Bit and hope the best ;-)
    usedCS = KGlobal::locale()->encoding();
    codec = KGlobal::charsets()->codecForName( QString::fromLatin1( usedCS ), ok );
  } else {
    Q_ASSERT( codec );
    if ( charset.isEmpty() ) {
      usedCS = codec->name();
    } else {
      usedCS = charset;
    }
  }

  QTextCodec::ConverterState converterState( QTextCodec::IgnoreHeader );
  QByteArray encoded8Bit = codec->fromUnicode( src.constData(), src.length(), &converterState );
  if ( converterState.invalidChars > 0 ) {
    usedCS = "utf-8";
    codec = QTextCodec::codecForName( usedCS );
    encoded8Bit = codec->fromUnicode( src );
  }

  if ( usedCS.contains( "8859-" ) ) { // use "B"-Encoding for non iso-8859-x charsets
    useQEncoding = true;
  }

  if ( allow8BitHeaders ) {
    return encoded8Bit;
  }

  uint encoded8BitLength = encoded8Bit.length();
  for ( unsigned int i=0; i<encoded8BitLength; i++ ) {
    if ( encoded8Bit[i] == ' ' ) { // encoding starts at word boundaries
      start = i + 1;
    }

    // encode escape character, for japanese encodings...
    if ( ( (signed char)encoded8Bit[i] < 0 ) || ( encoded8Bit[i] == '\033' ) ||
         ( addressHeader && ( strchr( "\"()<>@,.;:\\[]=", encoded8Bit[i] ) != 0 ) ) ) {
      end = start;   // non us-ascii char found, now we determine where to stop encoding
      nonAscii = true;
      break;
    }
  }

  if ( nonAscii ) {
    while ( ( end < encoded8Bit.length() ) && ( encoded8Bit[end] != ' ' ) ) {
      // we encode complete words
      end++;
    }

    for ( int x=end; x<encoded8Bit.length(); x++ ) {
      if ( ( (signed char)encoded8Bit[x] < 0 ) || ( encoded8Bit[x] == '\033' ) ||
           ( addressHeader && ( strchr( reservedCharacters, encoded8Bit[x] ) != 0 ) ) ) {
        end = x;     // we found another non-ascii word

        while ( ( end < encoded8Bit.length() ) && ( encoded8Bit[end] != ' ' ) ) {
          // we encode complete words
          end++;
        }
      }
    }

    result = encoded8Bit.left( start ) + "=?" + usedCS;

    if ( useQEncoding ) {
      result += "?Q?";

      char c, hexcode;// "Q"-encoding implementation described in RFC 2047
      for ( int i=start; i<end; i++ ) {
        c = encoded8Bit[i];
        if ( c == ' ' ) { // make the result readable with not MIME-capable readers
          result += '_';
        } else {
          if ( ( ( c >= 'a' ) && ( c <= 'z' ) ) || // paranoid mode, encode *all* special chars to avoid problems
              ( ( c >= 'A' ) && ( c <= 'Z' ) ) ||  // with "From" & "To" headers
              ( ( c >= '0' ) && ( c <= '9' ) ) ) {
            result += c;
          } else {
            result += '=';                 // "stolen" from KMail ;-)
            hexcode = ( ( c & 0xF0 ) >> 4 ) + 48;
            if ( hexcode >= 58 ) {
              hexcode += 7;
            }
            result += hexcode;
            hexcode = ( c & 0x0F ) + 48;
            if ( hexcode >= 58 ) {
              hexcode += 7;
            }
            result += hexcode;
          }
        }
      }
    } else {
      result += "?B?" + encoded8Bit.mid( start, end - start ).toBase64();
    }

    result +="?=";
    result += encoded8Bit.right( encoded8Bit.length() - end );
  } else {
    result = encoded8Bit;
  }

  return result;
}

QByteArray encodeRFC2047Sentence(const QString& src, const QByteArray& charset )
{
  QByteArray result;
  QList<QChar> splitChars;
  splitChars << QLatin1Char( ',' ) << QLatin1Char( '\"' ) << QLatin1Char( ';' ) << QLatin1Char( '\\' );
  const QChar *ch = src.constData();
  const int length = src.length();
  int pos = 0;
  int wordStart = 0;

  //qDebug() << "Input:" << src;
  // Loop over all characters of the string.
  // When encountering a split character, RFC-2047-encode the word before it, and add it to the result.
  while ( pos < length ) {
    //qDebug() << "Pos:" << pos << "Result:" << result << "Char:" << ch->toLatin1();
    const bool isAscii = ch->unicode() < 127;
    const bool isReserved = ( strchr( reservedCharacters, ch->toLatin1() ) != 0 );
    if ( isAscii && isReserved ) {
      const int wordSize = pos - wordStart;
      if ( wordSize > 0 ) {
        const QString word = src.mid( wordStart, wordSize );
        result += encodeRFC2047String( word, charset );
      }

      result += ch->toLatin1();
      wordStart = pos + 1;
    }
    ch++;
    pos++;
  }

  // Encode the last word
  const int wordSize = pos - wordStart;
  if ( wordSize > 0 ) {
    const QString word = src.mid( wordStart, pos - wordStart );
    result += encodeRFC2047String( word, charset );
  }

  return result;
}



//-----------------------------------------------------------------------------
QByteArray encodeRFC2231String( const QString& str, const QByteArray& charset )
{
  if ( str.isEmpty() ) {
    return QByteArray();
  }

  const QTextCodec *codec = KGlobal::charsets()->codecForName( QString::fromLatin1( charset ) );
  QByteArray latin;
  if ( charset == "us-ascii" ) {
    latin = str.toLatin1();
  } else if ( codec ) {
    latin = codec->fromUnicode( str );
  } else {
    latin = str.toLocal8Bit();
  }

  char *l;
  for ( l = latin.data(); *l; ++l ) {
    if ( ( ( *l & 0xE0 ) == 0 ) || ( *l & 0x80 ) ) {
      // *l is control character or 8-bit char
      break;
    }
  }
  if ( !*l ) {
    return latin;
  }

  QByteArray result = charset + "''";
  for ( l = latin.data(); *l; ++l ) {
    bool needsQuoting = ( *l & 0x80 ) || ( *l == '%' );
    if ( !needsQuoting ) {
      const QByteArray especials = "()<>@,;:\"/[]?.= \033";
      int len = especials.length();
      for ( int i = 0; i < len; i++ ) {
        if ( *l == especials[i] ) {
          needsQuoting = true;
          break;
        }
      }
    }
    if ( needsQuoting ) {
      result += '%';
      unsigned char hexcode;
      hexcode = ( ( *l & 0xF0 ) >> 4 ) + 48;
      if ( hexcode >= 58 ) {
        hexcode += 7;
      }
      result += hexcode;
      hexcode = ( *l & 0x0F ) + 48;
      if ( hexcode >= 58 ) {
        hexcode += 7;
      }
      result += hexcode;
    } else {
      result += *l;
    }
  }
  return result;
}


//-----------------------------------------------------------------------------
QString decodeRFC2231String( const QByteArray &str, QByteArray &usedCS, const QByteArray &defaultCS,
  bool forceCS )
{
  int p = str.indexOf( '\'' );
  if ( p < 0 ) {
    return KGlobal::charsets()->codecForName( QString::fromLatin1( defaultCS ) )->toUnicode( str );
  }


  QByteArray charset = str.left( p );

  QByteArray st = str.mid( str.lastIndexOf( '\'' ) + 1 );

  char ch, ch2;
  p = 0;
  while ( p < (int)st.length() ) {
    if ( st.at( p ) == 37 ) {
      // Only try to decode the percent-encoded character if the percent sign
      // is really followed by two other characters, see testcase at bug 163024
      if ( p + 2 < st.length() ) {
        ch = st.at( p + 1 ) - 48;
        if ( ch > 16 ) {
          ch -= 7;
        }
        ch2 = st.at( p + 2 ) - 48;
        if ( ch2 > 16 ) {
          ch2 -= 7;
        }
        st[p] = ch * 16 + ch2;
        st.remove( p + 1, 2 );
      }
    }
    p++;
  }
  kDebug() << "Got pre-decoded:" << st;
  QString result;
  const QTextCodec * charsetcodec = KGlobal::charsets()->codecForName( QString::fromLatin1( charset ) );
  if ( !charsetcodec || forceCS ) {
    charsetcodec = KGlobal::charsets()->codecForName( QString::fromLatin1( defaultCS ) );
  }

  usedCS = charsetcodec->name();
  return charsetcodec->toUnicode( st );
}

QString decodeRFC2231String( const QByteArray &src )
{
  QByteArray usedCS;
  return decodeRFC2231String( src, usedCS, "utf-8", false );
}

QByteArray uniqueString()
{
  static char chars[] = "0123456789abcdefghijklmnopqrstuvxyzABCDEFGHIJKLMNOPQRSTUVWXYZ";
  time_t now;
  char p[11];
  int pos, ran;
  unsigned int timeval;

  p[10] = '\0';
  now = time( 0 );
  ran = 1 + (int)( 1000.0 * rand() / ( RAND_MAX + 1.0 ) );
    timeval = ( now / ran ) + getpid();

  for ( int i = 0; i < 10; i++ ) {
    pos = (int) ( 61.0 * rand() / ( RAND_MAX + 1.0 ) );
    //kDebug() << pos;
    p[i] = chars[pos];
  }

  QByteArray ret;
  ret.setNum( timeval );
  ret += '.';
  ret += p;

  return ret;
}

QByteArray multiPartBoundary()
{
  return "nextPart" + uniqueString();
}

QByteArray unfoldHeader( const QByteArray &header )
{
  QByteArray result;
  if ( header.isEmpty() ) {
    return result;
  }

  int pos = 0, foldBegin = 0, foldMid = 0, foldEnd = 0;
  while ( ( foldMid = header.indexOf( '\n', pos ) ) >= 0 ) {
    foldBegin = foldEnd = foldMid;
    // find the first space before the line-break
    while ( foldBegin > 0 ) {
      if ( !QChar::fromLatin1( header[foldBegin - 1] ).isSpace() ) {
        break;
      }
      --foldBegin;
    }
    // find the first non-space after the line-break
    while ( foldEnd <= header.length() - 1 ) {
      if ( QChar::fromLatin1( header[foldEnd] ).isSpace() ) {
        ++foldEnd;
      } else if ( foldEnd > 0 && header[foldEnd - 1] == '\n' &&
                header[foldEnd] == '=' && foldEnd + 2 < header.length() &&
                ( ( header[foldEnd + 1] == '0' &&
                    header[foldEnd + 2] == '9' ) ||
                  ( header[foldEnd + 1] == '2' &&
                    header[foldEnd + 2] == '0' ) ) ) {
        // bug #86302: malformed header continuation starting with =09/=20
        foldEnd += 3;
      }
      else {
        break;
      }
    }

    result += header.mid( pos, foldBegin - pos );
    if ( foldEnd < header.length() - 1 ) {
      result += ' ';
    }
    pos = foldEnd;
  }
  const int len = header.length();
  if ( len > pos ) {
    result += header.mid( pos, len - pos );
  }
  return result;
}

int findHeaderLineEnd( const QByteArray &src, int &dataBegin, bool *folded )
{
  int end = dataBegin;
  int len = src.length() - 1;

  if ( folded ) {
    *folded = false;
  }

  if ( dataBegin < 0 ) {
    // Not found
    return -1;
  }

  if ( dataBegin > len ) {
    // No data available
    return len + 1;
  }

  // If the first line contains nothing, but the next line starts with a space
  // or a tab, that means a stupid mail client has made the first header field line
  // entirely empty, and has folded the rest to the next line(s).
  if ( src.at( end ) == '\n' && end + 1 < len &&
       ( src[end + 1] == ' ' || src[end + 1] == '\t' ) ) {

    // Skip \n and first whitespace
    dataBegin += 2;
    end += 2;
  }

  if ( src.at( end ) != '\n' ) {  // check if the header is not empty
    while ( true ) {
      end = src.indexOf( '\n', end + 1 );
      if ( end == -1 || end == len ) {
        // end of string
        break;
      } else if ( src[end + 1] == ' ' || src[end + 1] == '\t' ||
                ( src[end + 1] == '=' && end + 3 <= len &&
                  ( ( src[end + 2] == '0' && src[end + 3] == '9' ) ||
                    ( src[end + 2] == '2' && src[end + 3] == '0' ) ) ) ) {
        // next line is header continuation or starts with =09/=20 (bug #86302)
        if ( folded ) {
          *folded = true;
        }
      } else {
        // end of header (no header continuation)
        break;
      }
    }
  }

  if ( end < 0 ) {
    end = len + 1; //take the rest of the string
  }
  return end;
}

int indexOfHeader( const QByteArray &src, const QByteArray &name, int &end, int &dataBegin, bool *folded )
{
  QByteArray n = name;
  n.append( ':' );
  int begin = -1;

  if ( qstrnicmp( n.constData(), src.constData(), n.length() ) == 0 ) {
    begin = 0;
  } else {
    n.prepend( '\n' );
    const char *p = strcasestr( src.constData(), n.constData() );
    if ( !p ) {
      begin = -1;
    } else {
      begin = p - src.constData();
      ++begin;
    }
  }

  if ( begin > -1 ) {     //there is a header with the given name
    dataBegin = begin + name.length() + 1; //skip the name
    // skip the usual space after the colon
    if ( src.at( dataBegin ) == ' ' ) {
      ++dataBegin;
    }
    end = findHeaderLineEnd( src, dataBegin, folded );
    return begin;

  } else {
    end = -1;
    dataBegin = -1;
    return -1; //header not found
  }
}

QByteArray extractHeader( const QByteArray &src, const QByteArray &name )
{
  int begin, end;
  bool folded;
  QByteArray result;

  if ( src.isEmpty() || indexOfHeader( src, name, end, begin, &folded ) < 0 ) {
    return result;
  }

  if ( begin >= 0 ) {
    if ( !folded ) {
      result = src.mid( begin, end - begin );
    } else {
      if ( end > begin ) {
        QByteArray hdrValue = src.mid( begin, end - begin );
        result = unfoldHeader( hdrValue );
      }
    }
  }
  return result;
}

QList<QByteArray> extractHeaders( const QByteArray &src, const QByteArray &name )
{
  int begin, end;
  bool folded;
  QList<QByteArray> result;
  QByteArray copySrc( src );

  if ( indexOfHeader( copySrc, name, end, begin, &folded ) < 0 ) {
    return result;
  }

  while ( begin >= 0 ) {
    if ( !folded ) {
      result.append( copySrc.mid( begin, end - begin ) );
    } else {
      QByteArray hdrValue = copySrc.mid( begin, end - begin );
      result.append( unfoldHeader( hdrValue ) );
    }

    // get the next one, a tiny bit ugly, but we don't want the previous to be found again...
    copySrc = copySrc.mid( end );
    if ( indexOfHeader( copySrc, name, end, begin, &folded ) < 0 ) {
      break;
    }
  }
  return result;
}

void removeHeader( QByteArray &header, const QByteArray &name )
{
  int begin, end, dummy;
  begin = indexOfHeader( header, name, end, dummy );
  if ( begin >= 0 ) {
    header.remove( begin, end - begin + 1 );
  }
}

QByteArray CRLFtoLF( const QByteArray &s )
{
  QByteArray ret = s;
  ret.replace( "\r\n", "\n" );
  return ret;
}

QByteArray CRLFtoLF( const char *s )
{
  QByteArray ret = s;
  return CRLFtoLF( ret );
}

QByteArray LFtoCRLF( const QByteArray &s )
{
  QByteArray ret = s;
  ret.replace( '\n', "\r\n" );
  return ret;
}

QByteArray LFtoCRLF( const char *s )
{
  QByteArray ret = s;
  return LFtoCRLF( ret );
}

namespace {
template < typename StringType, typename CharType > void removeQuotesGeneric( StringType & str )
{
  bool inQuote = false;
  for ( int i = 0; i < str.length(); ++i ) {
    if ( str[i] == CharType( '"' ) ) {
      str.remove( i, 1 );
      i--;
      inQuote = !inQuote;
    } else {
      if ( inQuote && ( str[i] == CharType( '\\' ) ) ) {
        str.remove( i, 1 );
      }
    }
  }
}
}

void removeQuots( QByteArray &str )
{
  removeQuotesGeneric<QByteArray, char>( str );
}

void removeQuots( QString &str )
{
  removeQuotesGeneric<QString, QLatin1Char>( str );
}

template<class StringType,class CharType,class CharConverterType,class StringConverterType,class ToString>
void addQuotes_impl( StringType &str, bool forceQuotes )
{
  bool needsQuotes=false;
  for ( int i=0; i < str.length(); i++ ) {
    const CharType cur = str.at( i );
    if ( QString( ToString( str ) ).contains( QRegExp( QLatin1String( "\"|\\\\|=|\\]|\\[|:|;|,|\\.|,|@|<|>|\\)|\\(" ) ) ) ) {
      needsQuotes = true;
    }
    if ( cur == CharConverterType( '\\' ) || cur == CharConverterType( '\"' ) ) {
      str.insert( i, CharConverterType( '\\' ) );
      i++;
    }
  }

  if ( needsQuotes || forceQuotes ) {
    str.insert( 0, CharConverterType( '\"' ) );
    str.append( StringConverterType( "\"" ) );
  }
}

void addQuotes( QByteArray &str, bool forceQuotes )
{
  addQuotes_impl<QByteArray, char, char, char*, QLatin1String>( str, forceQuotes );
}

void addQuotes( QString &str, bool forceQuotes )
{
  addQuotes_impl<QString, QChar, QLatin1Char, QLatin1String, QString>( str, forceQuotes );
}

KMIME_EXPORT QString balanceBidiState( const QString &input )
{
  const int LRO = 0x202D;
  const int RLO = 0x202E;
  const int LRE = 0x202A;
  const int RLE = 0x202B;
  const int PDF = 0x202C;

  QString result = input;

  int openDirChangers = 0;
  int numPDFsRemoved = 0;
  for ( int i = 0; i < input.length(); i++ ) {
    const ushort &code = input.at( i ).unicode();
    if ( code == LRO || code == RLO || code == LRE || code == RLE ) {
      openDirChangers++;
    } else if ( code == PDF ) {
      if ( openDirChangers > 0 ) {
        openDirChangers--;
      } else {
        // One PDF too much, remove it
        kWarning() << "Possible Unicode spoofing (unexpected PDF) detected in" << input;
        result.remove( i - numPDFsRemoved, 1 );
        numPDFsRemoved++;
      }
    }
  }

  if ( openDirChangers > 0 ) {
    kWarning() << "Possible Unicode spoofing detected in" << input;

    // At PDF chars to the end until the correct state is restored.
    // As a special exception, when encountering quoted strings, place the PDF before
    // the last quote.
    for ( int i = openDirChangers; i > 0; i-- ) {
      if ( result.endsWith( QLatin1Char( '"' ) ) ) {
        result.insert( result.length() - 1, QChar( PDF ) );
      } else {
        result += QChar( PDF );
      }
    }
  }

  return result;
}

QString removeBidiControlChars( const QString &input )
{
  const int LRO = 0x202D;
  const int RLO = 0x202E;
  const int LRE = 0x202A;
  const int RLE = 0x202B;
  QString result = input;
  result.remove( LRO );
  result.remove( RLO );
  result.remove( LRE );
  result.remove( RLE );
  return result;
}

static bool isCryptoPart( Content* content )
{
  if ( !content->contentType( false ) ) {
    return false;
  }

  if ( content->contentType()->subType().toLower() == "octet-stream" &&
      !content->contentDisposition( false ) ) {
    return false;
  }

  const Headers::ContentType *contentType = content->contentType();
  const QByteArray lowerSubType = contentType->subType().toLower();
  return ( contentType->mediaType().toLower() == "application" &&
         ( lowerSubType == "pgp-encrypted" ||
           lowerSubType == "pgp-signature" ||
           lowerSubType == "pkcs7-mime" ||
           lowerSubType == "x-pkcs7-mime" ||
           lowerSubType == "pkcs7-signature" ||
           lowerSubType == "x-pkcs7-signature" ||
           ( lowerSubType == "octet-stream" &&
             content->contentDisposition()->filename().toLower() == QLatin1String( "msg.asc" ) ) ) );
}

bool hasAttachment( Content* content )
{
  if ( !content ) {
    return false;
  }

  bool emptyFilename = true;
  if ( content->contentDisposition( false ) &&
       !content->contentDisposition()->filename().isEmpty() ) {
    emptyFilename = false;
  }

  if ( emptyFilename &&
       content->contentType( false ) &&
       !content->contentType()->name().isEmpty() ) {
    emptyFilename = false;
  }

  // ignore crypto parts
  if ( !emptyFilename && !isCryptoPart( content ) ) {
    return true;
  }

  // Ok, content itself is not an attachment. now we deal with multiparts
  if ( content->contentType()->isMultipart() ) {
    Q_FOREACH ( Content *child, content->contents() ) {
      if ( hasAttachment( child ) ) {
        return true;
      }
    }
  }
  return false;
}

bool hasInvitation( Content *content )
{
  if ( !content ) {
    return false;
  }

  if ( isInvitation(content) ) {
    return true;
  }

  // Ok, content itself is not an invitation. now we deal with multiparts
  if ( content->contentType()->isMultipart() ) {
    Q_FOREACH ( Content *child, content->contents() ) {
      if ( hasInvitation( child ) ) {
        return true;
      }
    }
  }
  return false;
}

bool isSigned( Message *message )
{
  if ( !message ) {
    return false;
  }

  const KMime::Headers::ContentType* const contentType = message->contentType();
  if ( contentType->isSubtype( "signed" ) ||
       contentType->isSubtype( "pgp-signature" ) ||
       contentType->isSubtype( "pkcs7-signature" ) ||
       contentType->isSubtype( "x-pkcs7-signature" ) ||
       message->mainBodyPart( "multipart/signed" ) ||
       message->mainBodyPart( "application/pgp-signature" ) ||
       message->mainBodyPart( "application/pkcs7-signature" ) ||
       message->mainBodyPart( "application/x-pkcs7-signature" ) ) {
    return true;
  }
  return false;
}

bool isEncrypted( Message *message )
{
  if ( !message ) {
    return false;
  }

  const KMime::Headers::ContentType* const contentType = message->contentType();
  if ( contentType->isSubtype( "encrypted" ) ||
       contentType->isSubtype( "pgp-encrypted" ) ||
       contentType->isSubtype( "pkcs7-mime" ) ||
       contentType->isSubtype( "x-pkcs7-mime" ) ||
       message->mainBodyPart( "multipart/encrypted" ) ||
       message->mainBodyPart( "application/pgp-encrypted" ) ||
       message->mainBodyPart( "application/pkcs7-mime" ) ||
       message->mainBodyPart( "application/x-pkcs7-mime" ) ) {
    return true;
  }

  return false;
}

bool isInvitation( Content *content )
{
  if ( !content ) {
    return false;
  }

  const KMime::Headers::ContentType* const contentType = content->contentType( false );

  if ( contentType && contentType->isMediatype( "text" ) && contentType->isSubtype( "calendar" ) ) {
    return true;
  }

  return false;
}

} // namespace KMime