int IptcData::load(const byte* buf, long len)
{
#ifdef DEBUG
std::cerr << "IptcData::load, len = " << len << "\n";
#endif
const byte* pRead = buf;
iptcMetadata_.clear();
uint16_t record = 0;
uint16_t dataSet = 0;
uint32_t sizeData = 0;
byte extTest = 0;
while (pRead + 3 < buf + len) {
// First byte should be a marker. If it isn't, scan forward and skip
// the chunk bytes present in some images. This deviates from the
// standard, which advises to treat such cases as errors.
if (*pRead++ != marker_) continue;
record = *pRead++;
dataSet = *pRead++;
extTest = *pRead;
if (extTest & 0x80) {
// extended dataset
uint16_t sizeOfSize = (getUShort(pRead, bigEndian) & 0x7FFF);
if (sizeOfSize > 4) return 5;
pRead += 2;
sizeData = 0;
for (; sizeOfSize > 0; --sizeOfSize) {
sizeData |= *pRead++ << (8 *(sizeOfSize-1));
}
}
else {
// standard dataset
sizeData = getUShort(pRead, bigEndian);
pRead += 2;
}
if (pRead + sizeData <= buf + len) {
int rc = 0;
if ((rc = readData(dataSet, record, pRead, sizeData)) != 0) {
#ifndef SUPPRESS_WARNINGS
std::cerr << "Warning: "
<< "Failed to read IPTC dataset "
<< IptcKey(dataSet, record)
<< " (rc = " << rc << "); skipped.\n";
#endif
}
}
#ifndef SUPPRESS_WARNINGS
else {
std::cerr << "Warning: "
<< "IPTC dataset " << IptcKey(dataSet, record)
<< " has invalid size " << sizeData << "; skipped.\n";
}
#endif
pRead += sizeData;
}
return 0;
} // IptcData::load
const char *IptcData::detectCharset() const
{
const_iterator pos = findKey(IptcKey("Iptc.Envelope.CharacterSet"));
if (pos != end()) {
const std::string value = pos->toString();
if (pos->value().ok()) {
if (value == "\033%G") return "UTF-8";
// other values are probably not practically relevant
}
}
bool ascii = true;
bool utf8 = true;
for (pos = begin(); pos != end(); ++pos) {
std::string value = pos->toString();
if (pos->value().ok()) {
int seqCount = 0;
std::string::iterator i;
for (i = value.begin(); i != value.end(); ++i) {
char c = *i;
if (seqCount) {
if ((c & 0xc0) != 0x80) {
utf8 = false;
break;
}
--seqCount;
}
else {
if (c & 0x80) ascii = false;
else continue; // ascii character
if ((c & 0xe0) == 0xc0) seqCount = 1;
else if ((c & 0xf0) == 0xe0) seqCount = 2;
else if ((c & 0xf8) == 0xf0) seqCount = 3;
else if ((c & 0xfc) == 0xf8) seqCount = 4;
else if ((c & 0xfe) == 0xfc) seqCount = 5;
else {
utf8 = false;
break;
}
}
}
if (seqCount) utf8 = false; // unterminated seq
if (!utf8) break;
}
}
if (ascii) return "ASCII";
if (utf8) return "UTF-8";
return NULL;
}
