You cannot select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
exiv2/src/bmffimage.cpp

808 lines
28 KiB
C++

// SPDX-License-Identifier: GPL-2.0-or-later
// included header files
#include "bmffimage.hpp"
#include "basicio.hpp"
#include "config.h"
#include "enforce.hpp"
#include "error.hpp"
#include "futils.hpp"
#include "image.hpp"
#include "image_int.hpp"
#include "safe_op.hpp"
#include "tiffimage.hpp"
#include "tiffimage_int.hpp"
#include "types.hpp"
#include "utils.hpp"
#ifdef EXV_HAVE_BROTLI
#include <brotli/decode.h> // for JXL brob
#endif
// + standard includes
#include <cinttypes>
#include <cstdio>
#include <cstring>
#include <iostream>
#include <string>
#define TAG_ftyp 0x66747970 /**< "ftyp" File type box */
#define TAG_avif 0x61766966 /**< "avif" AVIF */
#define TAG_avio 0x6176696f /**< "avio" AVIF */
#define TAG_avis 0x61766973 /**< "avis" AVIF */
#define TAG_heic 0x68656963 /**< "heic" HEIC */
#define TAG_heif 0x68656966 /**< "heif" HEIF */
#define TAG_heim 0x6865696d /**< "heim" HEIC */
#define TAG_heis 0x68656973 /**< "heis" HEIC */
#define TAG_heix 0x68656978 /**< "heix" HEIC */
#define TAG_mif1 0x6d696631 /**< "mif1" HEIF */
#define TAG_crx 0x63727820 /**< "crx " Canon CR3 */
#define TAG_jxl 0x6a786c20 /**< "jxl " JPEG XL file type */
#define TAG_moov 0x6d6f6f76 /**< "moov" Movie */
#define TAG_meta 0x6d657461 /**< "meta" Metadata */
#define TAG_mdat 0x6d646174 /**< "mdat" Media data */
#define TAG_uuid 0x75756964 /**< "uuid" UUID */
#define TAG_dinf 0x64696e66 /**< "dinf" Data information */
#define TAG_iprp 0x69707270 /**< "iprp" Item properties */
#define TAG_ipco 0x6970636f /**< "ipco" Item property container */
#define TAG_iinf 0x69696e66 /**< "iinf" Item info */
#define TAG_iloc 0x696c6f63 /**< "iloc" Item location */
#define TAG_ispe 0x69737065 /**< "ispe" Image spatial extents */
Spelling (#2067) * spelling: accelerometer Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: address Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: anonymous Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: available Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: baby Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: buffer Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: build Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: builds Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: by inclusion Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: byte order Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: camera Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: command Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: consistency Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: constructor Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: continuous Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: continuously Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: correspond Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: decompression Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: dedicated Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: denominator Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: dependencies Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: detector Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: dlighting Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: dragging Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: duplicate Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: dynamic Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: empty Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: enable Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: endianness Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: examining Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: exceed Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: extension Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: fact Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: fails Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: fixed Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: fluorescent Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: focuspoint Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: focuspoints Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: from Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: happened Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: hierarchy Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: initiate Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: interesting Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: internal Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: javascript Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: language Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: libiconv Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: libraries Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: library Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: localization Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: macros Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: member Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: members Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: metadata Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: miscellaneous Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: mnemonic Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: modified Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: multiple Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: multithreaded Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: mysterious Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: photoshop Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: platforms Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: preliminary Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: priority Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: properties Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: qualifier Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: recognition Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: recognized Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: reentrancy Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: remaining Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: remotetest Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: requested Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: sensitivity Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: standardize Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: successful Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: such Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: tags Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: temperature Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: test Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: the Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: these Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: third Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: this Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: transparent Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: unable Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: underwater Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: uninterpreted Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: using Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: utc - coordinated universal time Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: which Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> * spelling: with Signed-off-by: Josh Soref <jsoref@users.noreply.github.com> Co-authored-by: Josh Soref <jsoref@users.noreply.github.com>
3 years ago
#define TAG_infe 0x696e6665 /**< "infe" Item Info Extension */
#define TAG_ipma 0x69706d61 /**< "ipma" Item Property Association */
#define TAG_cmt1 0x434d5431 /**< "CMT1" ifd0Id */
#define TAG_cmt2 0x434D5432 /**< "CMD2" exifID */
#define TAG_cmt3 0x434D5433 /**< "CMT3" canonID */
#define TAG_cmt4 0x434D5434 /**< "CMT4" gpsID */
#define TAG_colr 0x636f6c72 /**< "colr" Colour information */
#define TAG_exif 0x45786966 /**< "Exif" Used by JXL */
#define TAG_xml 0x786d6c20 /**< "xml " Used by JXL */
#define TAG_brob 0x62726f62 /**< "brob" Used by JXL (brotli box) */
#define TAG_thmb 0x54484d42 /**< "THMB" Canon thumbnail */
#define TAG_prvw 0x50525657 /**< "PRVW" Canon preview image */
// *****************************************************************************
// class member definitions
#ifdef EXV_ENABLE_BMFF
namespace Exiv2 {
static bool enabled = false;
EXIV2API bool enableBMFF(bool enable) {
enabled = enable;
return true;
}
std::string Iloc::toString() const {
return Internal::stringFormat("ID = %u from,length = %u,%u", ID_, start_, length_);
}
BmffImage::BmffImage(BasicIo::UniquePtr io, bool /* create */) :
Image(ImageType::bmff, mdExif | mdIptc | mdXmp, std::move(io)) {
} // BmffImage::BmffImage
std::string BmffImage::toAscii(uint32_t n) {
const auto p = reinterpret_cast<const char*>(&n);
std::string result;
for (int i = 0; i < 4; i++) {
char c = p[isBigEndianPlatform() ? i : (3 - i)];
result += [c]() {
if (32 <= c && c < 127)
return c; // only allow 7-bit printable ascii
if (c == 0)
return '_'; // show 0 as _
return '.'; // others .
}();
}
return result;
}
bool BmffImage::superBox(uint32_t box) {
return box == TAG_moov || box == TAG_dinf || box == TAG_iprp || box == TAG_ipco || box == TAG_meta ||
box == TAG_iinf || box == TAG_iloc;
}
bool BmffImage::fullBox(uint32_t box) {
return box == TAG_meta || box == TAG_iinf || box == TAG_iloc || box == TAG_thmb || box == TAG_prvw;
}
static bool skipBox(uint32_t box) {
// Allows boxHandler() to optimise the reading of files by identifying
// box types that we're not interested in. Box types listed here must
// not appear in the cases in switch (box_type) in boxHandler().
return box == TAG_mdat; // mdat is where the main image lives and can be huge
}
std::string BmffImage::mimeType() const {
switch (fileType_) {
case TAG_avif:
case TAG_avio:
case TAG_avis:
return "image/avif";
case TAG_heic:
case TAG_heim:
case TAG_heis:
case TAG_heix:
return "image/heic";
case TAG_heif:
case TAG_mif1:
return "image/heif";
case TAG_crx:
return "image/x-canon-cr3";
case TAG_jxl:
return "image/jxl"; // https://github.com/novomesk/qt-jpegxl-image-plugin/issues/1
default:
return "image/generic";
}
}
uint32_t BmffImage::pixelWidth() const {
auto imageWidth = exifData_.findKey(Exiv2::ExifKey("Exif.Photo.PixelXDimension"));
if (imageWidth != exifData_.end() && imageWidth->count() > 0) {
return imageWidth->toUint32();
}
return pixelWidth_;
}
uint32_t BmffImage::pixelHeight() const {
auto imageHeight = exifData_.findKey(Exiv2::ExifKey("Exif.Photo.PixelYDimension"));
if (imageHeight != exifData_.end() && imageHeight->count() > 0) {
return imageHeight->toUint32();
}
return pixelHeight_;
}
std::string BmffImage::uuidName(const Exiv2::DataBuf& uuid) {
const char* uuidCano = "\x85\xC0\xB6\x87\x82\xF\x11\xE0\x81\x11\xF4\xCE\x46\x2B\x6A\x48";
const char* uuidXmp = "\xBE\x7A\xCF\xCB\x97\xA9\x42\xE8\x9C\x71\x99\x94\x91\xE3\xAF\xAC";
const char* uuidCanp = "\xEA\xF4\x2B\x5E\x1C\x98\x4B\x88\xB9\xFB\xB7\xDC\x40\x6E\x4D\x16";
if (uuid.cmpBytes(0, uuidCano, 16) == 0)
return "cano";
if (uuid.cmpBytes(0, uuidXmp, 16) == 0)
return "xmp";
if (uuid.cmpBytes(0, uuidCanp, 16) == 0)
return "canp";
return "";
}
#ifdef EXV_HAVE_BROTLI
// Wrapper class for BrotliDecoderState that automatically calls
// BrotliDecoderDestroyInstance in its destructor.
class BrotliDecoderWrapper {
BrotliDecoderState* decoder_;
public:
BrotliDecoderWrapper() : decoder_(BrotliDecoderCreateInstance(nullptr, nullptr, nullptr)) {
if (!decoder_) {
throw Error(ErrorCode::kerMallocFailed);
}
}
~BrotliDecoderWrapper() {
BrotliDecoderDestroyInstance(decoder_);
}
BrotliDecoderWrapper(const BrotliDecoderWrapper&) = delete;
BrotliDecoderWrapper& operator=(const BrotliDecoderWrapper&) = delete;
BrotliDecoderState* get() const {
return decoder_;
}
};
void BmffImage::brotliUncompress(const byte* compressedBuf, size_t compressedBufSize, DataBuf& arr) {
BrotliDecoderWrapper decoder;
size_t uncompressedLen = compressedBufSize * 2; // just a starting point
BrotliDecoderResult result;
int dos = 0;
size_t available_in = compressedBufSize;
const byte* next_in = compressedBuf;
size_t available_out;
byte* next_out;
size_t total_out = 0;
do {
arr.alloc(uncompressedLen);
available_out = uncompressedLen - total_out;
next_out = arr.data() + total_out;
result =
BrotliDecoderDecompressStream(decoder.get(), &available_in, &next_in, &available_out, &next_out, &total_out);
if (result == BROTLI_DECODER_RESULT_SUCCESS) {
arr.resize(total_out);
} else if (result == BROTLI_DECODER_RESULT_NEEDS_MORE_OUTPUT) {
uncompressedLen *= 2;
// DoS protection - can't be bigger than 128k
if (uncompressedLen > 131072) {
if (++dos > 1)
break;
uncompressedLen = 131072;
}
} else if (result == BROTLI_DECODER_RESULT_NEEDS_MORE_INPUT) {
// compressed input buffer in incomplete
throw Error(ErrorCode::kerFailedToReadImageData);
} else {
// something bad happened
throw Error(ErrorCode::kerErrorMessage, BrotliDecoderErrorString(BrotliDecoderGetErrorCode(decoder.get())));
}
} while (result != BROTLI_DECODER_RESULT_SUCCESS);
if (result != BROTLI_DECODER_RESULT_SUCCESS) {
throw Error(ErrorCode::kerFailedToReadImageData);
}
}
#endif
uint64_t BmffImage::boxHandler(std::ostream& out /* = std::cout*/, Exiv2::PrintStructureOption option /* = kpsNone */,
uint64_t pbox_end, size_t depth) {
const size_t address = io_->tell();
// never visit a box twice!
if (depth == 0)
visits_.clear();
if (visits_.find(address) != visits_.end() || visits_.size() > visits_max_) {
throw Error(ErrorCode::kerCorruptedMetadata);
}
visits_.insert(address);
#ifdef EXIV2_DEBUG_MESSAGES
bool bTrace = true;
#else
bool bTrace = option == kpsBasic || option == kpsRecursive;
#endif
// 8-byte buffer for parsing the box length and type.
byte hdrbuf[2 * sizeof(uint32_t)];
size_t hdrsize = sizeof(hdrbuf);
Internal::enforce(hdrsize <= static_cast<size_t>(pbox_end - address), Exiv2::ErrorCode::kerCorruptedMetadata);
if (io_->read(reinterpret_cast<byte*>(&hdrbuf), sizeof(hdrbuf)) != sizeof(hdrbuf))
return pbox_end;
// The box length is encoded as a uint32_t by default, but the special value 1 means
// that it's a uint64_t.
uint64_t box_length = getULong(reinterpret_cast<byte*>(&hdrbuf[0]), endian_);
uint32_t box_type = getULong(reinterpret_cast<byte*>(&hdrbuf[sizeof(uint32_t)]), endian_);
bool bLF = true;
if (bTrace) {
bLF = true;
out << Internal::indent(depth) << "Exiv2::BmffImage::boxHandler: " << toAscii(box_type)
<< Internal::stringFormat(" %8ld->%" PRIu64 " ", address, box_length);
}
if (box_length == 1) {
// The box size is encoded as a uint64_t, so we need to read another 8 bytes.
hdrsize += 8;
Internal::enforce(hdrsize <= static_cast<size_t>(pbox_end - address), Exiv2::ErrorCode::kerCorruptedMetadata);
DataBuf data(8);
io_->read(data.data(), data.size());
box_length = data.read_uint64(0, endian_);
}
// read data in box and restore file position
const size_t restore = io_->tell();
Internal::enforce(box_length >= hdrsize, Exiv2::ErrorCode::kerCorruptedMetadata);
Internal::enforce(box_length - hdrsize <= pbox_end - restore, Exiv2::ErrorCode::kerCorruptedMetadata);
const auto buffer_size = box_length - hdrsize;
if (skipBox(box_type)) {
if (bTrace) {
out << std::endl;
}
// The enforce() above checks that restore + buffer_size won't
// exceed pbox_end, and by implication, won't exceed LONG_MAX
return restore + buffer_size;
}
DataBuf data(static_cast<size_t>(buffer_size));
const size_t box_end = restore + data.size();
io_->read(data.data(), data.size());
io_->seek(restore, BasicIo::beg);
size_t skip = 0; // read position in data.pData_
uint8_t version = 0;
uint32_t flags = 0;
if (fullBox(box_type)) {
Internal::enforce(data.size() - skip >= 4, Exiv2::ErrorCode::kerCorruptedMetadata);
flags = data.read_uint32(skip, endian_); // version/flags
version = static_cast<uint8_t>(flags >> 24);
flags &= 0x00ffffff;
skip += 4;
}
switch (box_type) {
// See notes in skipBox()
case TAG_ftyp: {
Internal::enforce(data.size() >= 4, Exiv2::ErrorCode::kerCorruptedMetadata);
fileType_ = data.read_uint32(0, endian_);
if (bTrace) {
out << "brand: " << toAscii(fileType_);
}
} break;
// 8.11.6.1
case TAG_iinf: {
if (bTrace) {
out << std::endl;
bLF = false;
}
Internal::enforce(data.size() - skip >= 2, Exiv2::ErrorCode::kerCorruptedMetadata);
uint16_t n = data.read_uint16(skip, endian_);
skip += 2;
io_->seek(skip, BasicIo::cur);
while (n-- > 0) {
io_->seek(boxHandler(out, option, box_end, depth + 1), BasicIo::beg);
}
} break;
// 8.11.6.2
case TAG_infe: { // .__._.__hvc1_ 2 0 0 1 0 1 0 0 104 118 99 49 0
Internal::enforce(data.size() - skip >= 8, Exiv2::ErrorCode::kerCorruptedMetadata);
/* getULong (data.pData_+skip,endian_) ; */ skip += 4;
uint16_t ID = data.read_uint16(skip, endian_);
skip += 2;
/* getShort(data.pData_+skip,endian_) ; */ skip += 2; // protection
std::string id;
// Check that the string has a '\0' terminator.
const char* str = data.c_str(skip);
const size_t maxlen = data.size() - skip;
Internal::enforce(maxlen > 0 && strnlen(str, maxlen) < maxlen, Exiv2::ErrorCode::kerCorruptedMetadata);
std::string name(str);
if (Internal::contains(name, "Exif")) { // "Exif" or "ExifExif"
exifID_ = ID;
id = " *** Exif ***";
} else if (Internal::contains(name, "mime\0xmp") || Internal::contains(name, "mime\0application/rdf+xml")) {
xmpID_ = ID;
id = " *** XMP ***";
}
if (bTrace) {
out << Internal::stringFormat("ID = %3d ", ID) << name << " " << id;
}
} break;
case TAG_moov:
case TAG_iprp:
case TAG_ipco:
case TAG_meta: {
if (bTrace) {
out << std::endl;
bLF = false;
}
io_->seek(skip, BasicIo::cur);
while (io_->tell() < box_end) {
io_->seek(boxHandler(out, option, box_end, depth + 1), BasicIo::beg);
}
// post-process meta box to recover Exif and XMP
if (box_type == TAG_meta) {
if (ilocs_.find(exifID_) != ilocs_.end()) {
const Iloc& iloc = ilocs_.find(exifID_)->second;
if (bTrace) {
out << Internal::indent(depth) << "Exiv2::BMFF Exif: " << iloc.toString() << std::endl;
}
parseTiff(Internal::Tag::root, iloc.length_, iloc.start_);
}
if (ilocs_.find(xmpID_) != ilocs_.end()) {
const Iloc& iloc = ilocs_.find(xmpID_)->second;
if (bTrace) {
out << Internal::indent(depth) << "Exiv2::BMFF XMP: " << iloc.toString() << std::endl;
}
parseXmp(iloc.length_, iloc.start_);
}
ilocs_.clear();
}
} break;
// 8.11.3.1
case TAG_iloc: {
Internal::enforce(data.size() - skip >= 2, Exiv2::ErrorCode::kerCorruptedMetadata);
uint8_t u = data.read_uint8(skip++);
uint16_t offsetSize = u >> 4;
uint16_t lengthSize = u & 0xF;
#if 0
uint16_t indexSize = 0 ;
u = data.read_uint8(skip++);
if ( version == 1 || version == 2 ) {
indexSize = u & 0xF ;
}
#else
skip++;
#endif
Internal::enforce(data.size() - skip >= (version < 2u ? 2u : 4u), Exiv2::ErrorCode::kerCorruptedMetadata);
uint32_t itemCount = version < 2 ? data.read_uint16(skip, endian_) : data.read_uint32(skip, endian_);
skip += version < 2 ? 2 : 4;
if (itemCount && itemCount < box_length / 14 && offsetSize == 4 && lengthSize == 4 &&
((box_length - 16) % itemCount) == 0) {
if (bTrace) {
out << std::endl;
bLF = false;
}
size_t step = (static_cast<size_t>(box_length) - 16) / itemCount; // length of data per item.
size_t base = skip;
for (uint32_t i = 0; i < itemCount; i++) {
skip = base + i * step; // move in 14, 16 or 18 byte steps
Internal::enforce(data.size() - skip >= (version > 2u ? 4u : 2u), Exiv2::ErrorCode::kerCorruptedMetadata);
Internal::enforce(data.size() - skip >= step, Exiv2::ErrorCode::kerCorruptedMetadata);
uint32_t ID = version > 2 ? data.read_uint32(skip, endian_) : data.read_uint16(skip, endian_);
auto offset = [this, &data, skip, step] {
if (step == 14 || step == 16)
return data.read_uint32(skip + step - 8, endian_);
if (step == 18)
return data.read_uint32(skip + 4, endian_);
return 0u;
}();
uint32_t ldata = data.read_uint32(skip + step - 4, endian_);
if (bTrace) {
out << Internal::indent(depth)
<< Internal::stringFormat("%8ld | %8ld | ID | %4u | %6u,%6u", address + skip, step, ID, offset, ldata)
<< std::endl;
}
// save data for post-processing in meta box
if (offset && ldata && ID != unknownID_) {
ilocs_[ID] = Iloc(ID, offset, ldata);
}
}
}
} break;
case TAG_ispe: {
Internal::enforce(data.size() - skip >= 12, Exiv2::ErrorCode::kerCorruptedMetadata);
skip += 4;
uint32_t width = data.read_uint32(skip, endian_);
skip += 4;
uint32_t height = data.read_uint32(skip, endian_);
skip += 4;
if (bTrace) {
out << "pixelWidth_, pixelHeight_ = " << Internal::stringFormat("%d, %d", width, height);
}
// HEIC files can have multiple ispe records
// Store largest width/height
if (width > pixelWidth_ && height > pixelHeight_) {
pixelWidth_ = width;
pixelHeight_ = height;
}
} break;
// 12.1.5.2
case TAG_colr: {
if (data.size() >= (skip + 4 + 8)) { // .____.HLino..__mntrR 2 0 0 0 0 12 72 76 105 110 111 2 16 ...
// https://www.ics.uci.edu/~dan/class/267/papers/jpeg2000.pdf
uint8_t meth = data.read_uint8(skip + 0);
uint8_t prec = data.read_uint8(skip + 1);
uint8_t approx = data.read_uint8(skip + 2);
auto colour_type = std::string(data.c_str(), 4);
skip += 4;
if (colour_type == "rICC" || colour_type == "prof") {
DataBuf profile(data.c_data(skip), data.size() - skip);
setIccProfile(std::move(profile));
} else if (meth == 2 && prec == 0 && approx == 0) {
// JP2000 files have a 3 byte head // 2 0 0 icc......
skip -= 1;
DataBuf profile(data.c_data(skip), data.size() - skip);
setIccProfile(std::move(profile));
}
}
} break;
case TAG_uuid: {
DataBuf uuid(16);
io_->read(uuid.data(), uuid.size());
std::string name = uuidName(uuid);
if (bTrace) {
out << " uuidName " << name << std::endl;
bLF = false;
}
if (name == "cano" || name == "canp") {
if (name == "canp") {
// based on
// https://github.com/lclevy/canon_cr3/blob/7be75d6/parse_cr3.py#L271
io_->seek(8, BasicIo::cur);
}
while (io_->tell() < box_end) {
io_->seek(boxHandler(out, option, box_end, depth + 1), BasicIo::beg);
}
} else if (name == "xmp") {
parseXmp(box_length, io_->tell());
}
} break;
case TAG_cmt1:
parseTiff(Internal::Tag::root, box_length);
break;
case TAG_cmt2:
parseTiff(Internal::Tag::cmt2, box_length);
break;
case TAG_cmt3:
parseTiff(Internal::Tag::cmt3, box_length);
break;
case TAG_cmt4:
parseTiff(Internal::Tag::cmt4, box_length);
break;
case TAG_exif:
parseTiff(Internal::Tag::root, buffer_size, io_->tell());
break;
case TAG_xml:
parseXmp(buffer_size, io_->tell());
break;
case TAG_brob: {
Internal::enforce(data.size() >= 4, Exiv2::ErrorCode::kerCorruptedMetadata);
uint32_t realType = data.read_uint32(0, endian_);
if (bTrace) {
out << "type: " << toAscii(realType);
}
#ifdef EXV_HAVE_BROTLI
DataBuf arr;
brotliUncompress(data.c_data(4), data.size() - 4, arr);
if (realType == TAG_exif) {
uint32_t offset = Safe::add(arr.read_uint32(0, endian_), 4u);
Internal::enforce(Safe::add(offset, 4u) < arr.size(), Exiv2::ErrorCode::kerCorruptedMetadata);
Internal::TiffParserWorker::decode(exifData(), iptcData(), xmpData(), arr.c_data(offset), arr.size() - offset,
Internal::Tag::root, Internal::TiffMapping::findDecoder);
} else if (realType == TAG_xml) {
try {
Exiv2::XmpParser::decode(xmpData(), std::string(arr.c_str(), arr.size()));
} catch (...) {
throw Error(ErrorCode::kerFailedToReadImageData);
}
}
#endif
} break;
case TAG_thmb:
switch (version) {
case 0: // JPEG
parseCr3Preview(data, out, bTrace, version, skip, skip + 2, skip + 4, skip + 12);
break;
case 1: // HDR
parseCr3Preview(data, out, bTrace, version, skip + 2, skip + 4, skip + 8, skip + 12);
break;
default:
break;
}
break;
case TAG_prvw:
switch (version) {
case 0: // JPEG
case 1: // HDR
parseCr3Preview(data, out, bTrace, version, skip + 2, skip + 4, skip + 8, skip + 12);
break;
default:
break;
}
break;
default:
break; /* do nothing */
}
if (bLF && bTrace)
out << std::endl;
// return address of next box
return box_end;
}
void BmffImage::parseTiff(uint32_t root_tag, uint64_t length, uint64_t start) {
Internal::enforce(start <= io_->size(), ErrorCode::kerCorruptedMetadata);
Internal::enforce(length <= io_->size() - start, ErrorCode::kerCorruptedMetadata);
Internal::enforce(start <= static_cast<uint64_t>(std::numeric_limits<int64_t>::max()),
ErrorCode::kerCorruptedMetadata);
Internal::enforce(length <= std::numeric_limits<size_t>::max(), ErrorCode::kerCorruptedMetadata);
// read and parse exif data
const size_t restore = io_->tell();
DataBuf exif(static_cast<size_t>(length));
io_->seek(static_cast<int64_t>(start), BasicIo::beg);
if (exif.size() > 8 && io_->read(exif.data(), exif.size()) == exif.size()) {
// hunt for "II" or "MM"
const size_t eof = std::numeric_limits<size_t>::max(); // impossible value for punt
size_t punt = eof;
for (size_t i = 0; i < exif.size() - 9 && punt == eof; ++i) {
if (exif.read_uint8(i) == exif.read_uint8(i + 1))
if (exif.read_uint8(i) == 'I' || exif.read_uint8(i) == 'M')
punt = i;
}
if (punt != eof) {
Internal::TiffParserWorker::decode(exifData(), iptcData(), xmpData(), exif.c_data(punt), exif.size() - punt,
root_tag, Internal::TiffMapping::findDecoder);
}
}
io_->seek(restore, BasicIo::beg);
}
void BmffImage::parseTiff(uint32_t root_tag, uint64_t length) {
if (length > 8) {
Internal::enforce(length - 8 <= io_->size() - io_->tell(), ErrorCode::kerCorruptedMetadata);
Internal::enforce(length - 8 <= std::numeric_limits<size_t>::max(), ErrorCode::kerCorruptedMetadata);
DataBuf data(static_cast<size_t>(length - 8u));
const size_t bufRead = io_->read(data.data(), data.size());
if (io_->error())
throw Error(ErrorCode::kerFailedToReadImageData);
if (bufRead != data.size())
throw Error(ErrorCode::kerInputDataReadFailed);
Internal::TiffParserWorker::decode(exifData(), iptcData(), xmpData(), data.c_data(), data.size(), root_tag,
Internal::TiffMapping::findDecoder);
}
}
void BmffImage::parseXmp(uint64_t length, uint64_t start) {
Internal::enforce(start <= io_->size(), ErrorCode::kerCorruptedMetadata);
Internal::enforce(length <= io_->size() - start, ErrorCode::kerCorruptedMetadata);
const size_t restore = io_->tell();
io_->seek(static_cast<int64_t>(start), BasicIo::beg);
auto lengthSizeT = static_cast<size_t>(length);
DataBuf xmp(lengthSizeT + 1);
xmp.write_uint8(lengthSizeT, 0); // ensure xmp is null terminated!
if (io_->read(xmp.data(), lengthSizeT) != lengthSizeT)
throw Error(ErrorCode::kerInputDataReadFailed);
if (io_->error())
throw Error(ErrorCode::kerFailedToReadImageData);
try {
Exiv2::XmpParser::decode(xmpData(), std::string(xmp.c_str()));
} catch (...) {
throw Error(ErrorCode::kerFailedToReadImageData);
}
io_->seek(restore, BasicIo::beg);
}
/// \todo instead of passing the last 4 parameters, pass just one and build the different offsets inside
void BmffImage::parseCr3Preview(DataBuf& data, std::ostream& out, bool bTrace, uint8_t version, size_t width_offset,
size_t height_offset, size_t size_offset, size_t relative_position) {
// Derived from https://github.com/lclevy/canon_cr3
const size_t here = io_->tell();
Internal::enforce(here <= std::numeric_limits<size_t>::max() - relative_position, ErrorCode::kerCorruptedMetadata);
NativePreview nativePreview;
nativePreview.position_ = here + relative_position;
nativePreview.width_ = data.read_uint16(width_offset, endian_);
nativePreview.height_ = data.read_uint16(height_offset, endian_);
nativePreview.size_ = data.read_uint32(size_offset, endian_);
nativePreview.filter_ = "";
nativePreview.mimeType_ = [version] {
if (version == 0)
return "image/jpeg";
return "application/octet-stream";
}();
nativePreviews_.push_back(nativePreview);
if (bTrace) {
out << Internal::stringFormat("width,height,size = %zu,%zu,%zu", nativePreview.width_, nativePreview.height_,
nativePreview.size_);
}
}
void BmffImage::setExifData(const ExifData& /*exifData*/) {
throw(Error(ErrorCode::kerInvalidSettingForImage, "Exif metadata", "BMFF"));
}
void BmffImage::setIptcData(const IptcData& /*iptcData*/) {
throw(Error(ErrorCode::kerInvalidSettingForImage, "IPTC metadata", "BMFF"));
}
void BmffImage::setXmpData(const XmpData& /*xmpData*/) {
throw(Error(ErrorCode::kerInvalidSettingForImage, "XMP metadata", "BMFF"));
}
void BmffImage::setComment(const std::string&) {
// bmff files are read-only
throw(Error(ErrorCode::kerInvalidSettingForImage, "Image comment", "BMFF"));
}
void BmffImage::openOrThrow() {
if (io_->open() != 0) {
throw Error(ErrorCode::kerDataSourceOpenFailed, io_->path(), strError());
}
// Ensure that this is the correct image type
if (!isBmffType(*io_, false)) {
if (io_->error() || io_->eof())
throw Error(ErrorCode::kerFailedToReadImageData);
throw Error(ErrorCode::kerNotAnImage, "BMFF");
}
} // Bmff::openOrThrow();
void BmffImage::readMetadata() {
openOrThrow();
IoCloser closer(*io_);
clearMetadata();
ilocs_.clear();
visits_max_ = io_->size() / 16;
unknownID_ = 0xffff;
exifID_ = unknownID_;
xmpID_ = unknownID_;
uint64_t address = 0;
const auto file_end = io_->size();
while (address < file_end) {
io_->seek(address, BasicIo::beg);
address = boxHandler(std::cout, kpsNone, file_end, 0);
}
bReadMetadata_ = true;
} // BmffImage::readMetadata
void BmffImage::printStructure(std::ostream& out, Exiv2::PrintStructureOption option, size_t depth) {
if (!bReadMetadata_)
readMetadata();
switch (option) {
default:
break; // do nothing
case kpsIccProfile: {
out.write(iccProfile_.c_str(), iccProfile_.size());
} break;
#ifdef EXV_HAVE_XMP_TOOLKIT
case kpsXMP: {
std::string xmp;
if (Exiv2::XmpParser::encode(xmp, xmpData())) {
throw Exiv2::Error(Exiv2::ErrorCode::kerErrorMessage, "Failed to serialize XMP data");
}
out << xmp;
} break;
#endif
case kpsBasic: // drop
case kpsRecursive: {
openOrThrow();
IoCloser closer(*io_);
uint64_t address = 0;
const auto file_end = io_->size();
while (address < file_end) {
io_->seek(address, BasicIo::beg);
address = boxHandler(out, option, file_end, depth);
}
} break;
}
}
void BmffImage::writeMetadata() {
// bmff files are read-only
throw(Error(ErrorCode::kerWritingImageFormatUnsupported, "BMFF"));
} // BmffImage::writeMetadata
// *************************************************************************
// free functions
Image::UniquePtr newBmffInstance(BasicIo::UniquePtr io, bool create) {
auto image = std::make_unique<BmffImage>(std::move(io), create);
if (!image->good()) {
return nullptr;
}
return image;
}
bool isBmffType(BasicIo& iIo, bool advance) {
if (!enabled) {
return false;
}
const int32_t len = 12;
byte buf[len];
iIo.read(buf, len);
if (iIo.error() || iIo.eof()) {
return false;
}
// bmff should start with "ftyp"
bool const is_ftyp = (buf[4] == 'f' && buf[5] == 't' && buf[6] == 'y' && buf[7] == 'p');
// jxl files have a special start indicator of "JXL "
bool const is_jxl = (buf[4] == 'J' && buf[5] == 'X' && buf[6] == 'L' && buf[7] == ' ');
bool matched = is_jxl || is_ftyp;
if (!advance || !matched) {
iIo.seek(0, BasicIo::beg);
}
return matched;
}
} // namespace Exiv2
#else // ifdef EXV_ENABLE_BMFF
namespace Exiv2 {
EXIV2API bool enableBMFF(bool) {
return false;
}
} // namespace Exiv2
#endif