std::string convertDVBUTF8(const char *data, int len, int table, int tsidonid)
{
int newtable = 0;
bool twochar = false;
if (!len)
return "";
int i = 0, t = 0;
if ( tsidonid )
{
std::map<int, int>::iterator it = TransponderDefaultMapping.find(tsidonid);
if ( it != TransponderDefaultMapping.end() )
table = it->second;
twochar = TransponderUseTwoCharMapping.find(tsidonid) != TransponderUseTwoCharMapping.end();
}
//printf("table %d tsidonid %04x twochar %d : %20s\n", table, tsidonid, twochar, data);
switch(data[0])
{
case 1 ... 12:
newtable=data[i++]+4;
//printf("(1..12)text encoded in ISO-8859-%d\n", table);
break;
case 0x10:
{
//printf("(0x10)text encoded in ISO-8859-%d\n",n);
int n=(data[i+1]<<8)|(data[i+2]);
i += 3;
switch(n)
{
case 12:
{} //eDebug("unsup. ISO8859-12 enc.", n);
default:
newtable=n;
break;
}
break;
}
case 0x11:// Basic Multilingual Plane of ISO/IEC 10646-1 enc (UTF-16... Unicode)
table = 65;
tsidonid = 0;
++i;
break;
case 0x12:
++i;
{} //printf("unsup. KSC 5601 enc.\n");
break;
case 0x13:
++i;
{} //printf("unsup. GB-2312-1980 enc.\n");
break;
case 0x14:
++i;
{} //printf("unsup. Big5 subset of ISO/IEC 10646-1 enc.\n");
break;
case 0x15: // UTF-8 encoding of ISO/IEC 10646-1
return std::string((char*)data+1, len-1);
case 0x1F:
{
std::string decoded_string = huffmanDecoder.decode((const unsigned char *)data, len);
if (!decoded_string.empty())
return decoded_string;
}
++i;
break;
case 0x0:
case 0xD ... 0xF:
case 0x16 ... 0x1E:
{} //printf("reserved %d\n", data[0]);
++i;
break;
}
if(!table)
table = newtable;
if(table == 64 && (newtable != 0 ))
{
//for ISO6937
table = newtable;
}
//printf("recode:::: tsidonid %X table %d two-char %d len %d\n", tsidonid, table, twochar, len);
unsigned char res[2048];
while (i < len)
{
unsigned long code = 0;
if ( i + 1 < len && twochar && (code = doVideoTexSuppl(data[i], data[i+1])) )
{
i += 2;
}
if (!code)
{
if (table == 65)
{
// unicode
if (i+1 < len)
{
code = (data[i] << 8) | data[i+1];
i += 2;
}
}
else
code = recode(data[i++], table);
}
if (!code)
continue;
// Unicode->UTF8 encoding
if (code < 0x80) // identity ascii <-> utf8 mapping
res[t++] = char(code);
else if((table == 5) && (code == 0x8A))
res[t++]= 0x20;
else if ((code == 0x8A))
res[t++]= '\n'; // 0x8a is vertical tab. Just use newline for now.
else if((code >= 0x80) && (code <= 0x9F))
continue;
else if (code < 0x800) // two byte mapping
{
res[t++] = (code>>6)|0xC0;
res[t++] = (code&0x3F)|0x80;
}
else if (code < 0x10000) // three bytes mapping
std::string convertDVBUTF8(const unsigned char *data, int len, int table, int tsidonid)
{
if (!len)
return "";
int i=0, t=0;
if ( tsidonid )
encodingHandler.getTransponderDefaultMapping(tsidonid, table);
switch(data[0])
{
case 1 ... 11:
// For Thai providers, encoding char is present but faulty.
if (table != 11)
table=data[i]+4;
++i;
// eDebug("(1..11)text encoded in ISO-8859-%d",table);
break;
case 0x10:
{
int n=(data[++i]<<8);
n |= (data[++i]);
// eDebug("(0x10)text encoded in ISO-8859-%d",n);
++i;
switch(n)
{
case 12:
eDebug("unsup. ISO8859-12 enc.");
break;
default:
table=n;
break;
}
break;
}
case 0x11: // Basic Multilingual Plane of ISO/IEC 10646-1 enc (UTF-16... Unicode)
table = 65;
tsidonid = 0;
++i;
break;
case 0x12:
++i;
eDebug("unsup. KSC 5601 enc.");
break;
case 0x13:
++i;
eDebug("unsup. GB-2312-1980 enc.");
break;
case 0x14:
++i;
eDebug("unsup. Big5 subset of ISO/IEC 10646-1 enc.");
break;
case 0x15: // UTF-8 encoding of ISO/IEC 10646-1
return std::string((char*)data+1, len-1);
case 0x1F:
{
// Attempt to decode Freesat Huffman encoded string
unsigned char *temp = (unsigned char *) huffmanDecoder.decode(data, len);
if (temp)
{
int newlen = strlen((char*) temp);
std::string decoded_string = convertDVBUTF8(temp, newlen, table, tsidonid);
free(temp);
return decoded_string;
}
}
i++;
eDebug("failed to decode bbc freesat huffman");
break;
case 0x0:
case 0xC ... 0xF:
case 0x16 ... 0x1E:
eDebug("reserved %d", data[0]);
++i;
break;
}
bool useTwoCharMapping = !table || (tsidonid && encodingHandler.getTransponderUseTwoCharMapping(tsidonid));
if (useTwoCharMapping && table == 5) { // i hope this dont break other transponders which realy use ISO8859-5 and two char byte mapping...
// eDebug("Cyfra / Cyfrowy Polsat HACK... override given ISO8859-5 with ISO6397");
table = 0;
}
unsigned char res[2048];
while (i < len)
{
unsigned long code=0;
if ( useTwoCharMapping && i+1 < len && (code=doVideoTexSuppl(data[i], data[i+1])) )
i+=2;
if (!code) {
if (table == 65) { // unicode
if (i+1 < len) {
code=(data[i] << 8) | data[i+1];
i += 2;
}
}
else
code=recode(data[i++], table);
}
if (!code)
continue;
// Unicode->UTF8 encoding
if (code < 0x80) // identity ascii <-> utf8 mapping
res[t++]=char(code);
else if (code < 0x800) // two byte mapping
{
res[t++]=(code>>6)|0xC0;
res[t++]=(code&0x3F)|0x80;
} else if (code < 0x10000) // three bytes mapping
