C++ Codec::fromUnicode方法代码示例

void Multipart::appendTextPart( EString & r, const Bodypart * bp,
                                ContentType * ct ) const
{
    Codec * c = 0;

    EString::Encoding e = EString::Binary;
    ContentTransferEncoding * cte
        = bp->header()->contentTransferEncoding();
    if ( cte )
        e = cte->encoding();

    if ( ct && !ct->parameter( "charset" ).isEmpty() )
        c = Codec::byName( ct->parameter( "charset" ) );
    if ( !c )
        c = Codec::byString( bp->text() );

    EString body = c->fromUnicode( bp->text() );

    r.append( body.encoded( e, 72 ) );
}

EString Bodypart::asText( bool avoidUtf8 ) const
{
    EString r;
    Codec *c = 0;

    ContentType *ct = header()->contentType();
    if ( ct && !ct->parameter( "charset" ).isEmpty() )
        c = Codec::byName( ct->parameter( "charset" ) );
    if ( !c )
        c = new AsciiCodec;

    if ( !children()->isEmpty() )
        appendMultipart( r, avoidUtf8 );
    else if ( !header()->contentType() ||
              header()->contentType()->type() == "text" )
        r = c->fromUnicode( text() );
    else
        r = d->data.e64( 72 );

    return r;
}

//.........这里部分代码省略.........

        bp->d->hasText = true;
        bp->d->text = c->toUnicode( body.crlf() );

        if ( c->name() == "GB2312" || c->name() == "ISO-2022-JP" ||
             c->name() == "KS_C_5601-1987" ) {
            // undefined code point usage in GB2312 spam is much too
            // common. (GB2312 spam is much too common, but that's
            // another matter.) Gb2312Codec turns all undefined code
            // points into U+FFFD, so here, we can take the unicode
            // form and say it's the canonical form. when a client
            // later reads the message, it gets the text in unicode,
            // including U+FFFD.

            bool bad = !c->valid();

            // the header may contain some unencoded gb2312. we bang
            // it by hand, ignoring errors.
            List<HeaderField>::Iterator hf( h->fields() );
            while ( hf ) {
                if ( !hf->valid() &&
                     hf->type() == HeaderField::Subject ) {
                    // is it right to bang only Subject?
                    c->reset();
                    hf->setValue( c->toUnicode( hf->unparsedValue() ) );
                }
                ++hf;
            }

            // if the body was bad, we prefer the (unicode) in
            // bp->d->text and pretend it arrived as UTF-8:
            if ( bad ) {
                c = new Utf8Codec;
                body = c->fromUnicode( bp->d->text );
            }
        }

        if ( ( !specified && ( !c->wellformed() ||
                               ct->subtype() == "html" ) ) ||
             ( specified &&  ( !c->valid() ) ) ) {
            Codec * g = 0;
            if ( ct->subtype() == "html" )
                g = guessHtmlCodec( body );
            else
                g = guessTextCodec( body );
            UString guessed;
            if ( g )
                guessed = g->toUnicode( body.crlf() );
            if ( !g ) {
                // if we couldn't guess anything, keep what we had if
                // it's valid or explicitly specified, else use
                // unknown-8bit.
                if ( !specified && !c->valid() ) {
                    c = new Unknown8BitCodec;
                    bp->d->text = c->toUnicode( body.crlf() );
                }
            }
            else {
                // if we could guess something, is our guess better
                // than what we had?
                if ( g->wellformed() && !c->wellformed() ) {
                    c = g;
                    bp->d->text = guessed;
                }
            }
        }