int bmp_isfile(const uint8_t *data, size_t input_len, size_t *lengthptr)
{
if (input_len < (BMP_HEADER_SIZE + DIB_HEADER_SIZE) || !IS_BMP_MAGIC(data))
return 0;
const struct bmp_header *header = (const struct bmp_header *)data;
size_t filesize = le32toh(header->filesize);
uint16_t reserved1 = le16toh(header->reserved1);
uint16_t reserved2 = le16toh(header->reserved2);
size_t dataoffset = le32toh(header->dataoffset);
size_t headersize = le32toh(header->headersize);
int32_t width = le32toh(header->width);
int32_t height = le32toh(header->height);
uint16_t planes = le16toh(header->planes);
uint16_t bpp = le16toh(header->bpp);
uint32_t compression = le32toh(header->compression);
size_t datasize = le32toh(header->datasize);
size_t palettecolors = le32toh(header->palettecolors);
size_t colortblsize = 4 * palettecolors;
// integer overflow?
if (SIZE_MAX / 4 < palettecolors)
return 0;
// sanity of all sizes and offsets
if (filesize < (BMP_HEADER_SIZE + DIB_HEADER_SIZE) || dataoffset >= filesize || datasize >= filesize ||
headersize >= filesize || headersize < DIB_HEADER_SIZE || colortblsize >= filesize ||
dataoffset < (BMP_HEADER_SIZE + headersize + colortblsize) ||
filesize - datasize < dataoffset)
return 0;
// legal range of values?
if (width <= 0 || height == 0 || planes != 1 || bpp == 0 ||
reserved1 != 0 || reserved2 != 0 || compression > 6)
return 0;
if (filesize > input_len)
return 0;
if (lengthptr)
*lengthptr = filesize;
return 1;
}
int do_extract(const uint8_t *filedata, size_t filesize, const struct extract_options *options, size_t *numfilesptr, size_t *sumsizeptr)
{
const uint8_t *ptr = NULL, *end = NULL;
enum fileformat format = NONE;
size_t sumsize = 0;
size_t length = 0;
int success = 1;
int formats = options->formats;
char *outfilename = NULL;
size_t numfiles = 0;
const char *filename = basename(options->filepath);
// max. ext length is 16 characters
size_t namelen = strlen(options->outdir) + strlen(filename) + 37;
struct mpg123_info mpg123;
struct ogg_info ogg;
struct file_info info = {0, 0};
size_t count = 0; // e.g. for tracks count in midi
const uint8_t *audio_start = NULL;
size_t input_len = 0;
outfilename = malloc(namelen);
if (outfilename == NULL)
{
perror(options->filepath);
goto error;
}
if (!options->quiet)
{
double slice_size = 0;
const char *slice_unit = format_size(filesize, &slice_size);
printf("Extracting 0x%08"PRIx64" ... 0x%08"PRIx64" (%g %s) from %s\n",
options->offset,
options->offset + filesize,
slice_size, slice_unit,
options->filepath);
}
#define WRITE_FILE(data, length, ext) \
if (write_file((data), length, options, filename, (size_t)((data) - filedata), (ext), outfilename, namelen)) \
{ \
++ numfiles; \
sumsize += length; \
}
ptr = filedata;
end = filedata + filesize;
for (input_len = filesize; input_len >= 4; input_len = (size_t)(end - ptr))
{
uint32_t magic = MAGIC(ptr);
if (formats & OGG && magic == OGG_MAGIC && ogg_ispage(ptr, input_len, &ogg))
{
uint32_t pageno = ogg.pageno;
audio_start = ptr;
for (;;)
{
ptr += ogg.length;
if (!ogg_ispage(ptr, (size_t)(end - ptr), &ogg) || ogg.pageno <= pageno)
break;
pageno = ogg.pageno;
}
WRITE_FILE(audio_start, ptr - audio_start, "ogg");
continue;
}
if (formats & RIFF && magic == RIFF_MAGIC && riff_isfile(ptr, input_len, &info))
{
WRITE_FILE(ptr, info.length, info.ext);
ptr += info.length;
continue;
}
if (formats & AIFF && magic == FORM_MAGIC && aiff_isfile(ptr, input_len, &info))
{
WRITE_FILE(ptr, info.length, info.ext);
ptr += info.length;
continue;
}
if (formats & MIDI && magic == MIDI_MAGIC && midi_isheader(ptr, input_len, &length, &count))
{
audio_start = ptr;
do {
ptr += length;
} while (count-- > 0 && midi_istrack(ptr, (size_t)(end - ptr), &length));
if (count != 0 && !(options->quiet))
{
fprintf(stderr, "warning: midi file misses %"PRIzu" tracks\n", count);
}
WRITE_FILE(audio_start, ptr - audio_start, "mid");
continue;
}
format = NONE;
if (formats & ID3v2 && IS_ID3v2_MAGIC(ptr) && id3v2_istag(ptr, input_len, 0, &length))
{
format = ID3v2;
}
if (formats & MPG123 && IS_MPG123_MAGIC(ptr))
{
format = MPG123;
length = 0;
}
if (format & (ID3v2 | MPG123) && mpg123_isframe(ptr + length, input_len - length, &mpg123))
{
uint8_t version = mpg123.version;
uint8_t layer = mpg123.layer;
audio_start = ptr;
ptr += length;
do {
ptr += mpg123.frame_size;
} while (mpg123_isframe(ptr, (size_t)(end - ptr), &mpg123)
&& mpg123.version == version
&& mpg123.layer == layer);
if (id3v1_istag(ptr, (size_t)(end - ptr), &length))
{
ptr += length;
}
if (formats & ID3v2 && id3v2_istag(ptr, (size_t)(end - ptr), 1, &length))
{
ptr += length;
}
WRITE_FILE(audio_start, ptr - audio_start,
layer == 1 ? "mp1" :
layer == 2 ? "mp2" :
layer == 3 ? "mp3" :
"mpg");
continue;
}
if (formats & IT && magic == IT_MAGIC && it_isfile(ptr, input_len, &length))
{
WRITE_FILE(ptr, length, "it");
ptr += length;
continue;
}
if (formats & XM && magic == XM_MAGIC && xm_isfile(ptr, input_len, &length))
{
WRITE_FILE(ptr, length, "xm");
ptr += length;
continue;
}
if (formats & ASF && magic == ASF_MAGIC && asf_isfile(ptr, input_len, &length))
{
WRITE_FILE(ptr, length, "asf");
ptr += length;
continue;
}
if (formats & AU && magic == AU_MAGIC && au_isfile(ptr, input_len, &length))
{
WRITE_FILE(ptr, length, "au");
ptr += length;
continue;
}
if (formats & PNG && magic == PNG_MAGIC && png_isfile(ptr, input_len, &length))
{
WRITE_FILE(ptr, length, "png");
ptr += length;
continue;
}
if (formats & GIF && magic == GIF_MAGIC && gif_isfile(ptr, input_len, &length))
{
WRITE_FILE(ptr, length, "gif");
ptr += length;
continue;
}
if (formats & (MPEG1 | MPEGPS | MPEGVS) && IS_MPEG_MAGIC(magic) && mpeg_isfile(ptr, input_len, formats, &length))
{
WRITE_FILE(ptr, length, "mpg");
ptr += length;
continue;
}
if (formats & MPEGTS && IS_MPEG_TS_MAGIC(ptr) && mpeg_isfile(ptr, input_len, formats, &length))
{
WRITE_FILE(ptr, length, "mpg");
ptr += length;
continue;
}
if (formats & JPEG && IS_JPG_MAGIC(magic) && jpg_isfile(ptr, input_len, &length))
{
WRITE_FILE(ptr, length, "jpg");
ptr += length;
continue;
}
if (formats & BINK && IS_BINK_MAGIC(magic) && bink_isfile(ptr, input_len, &length))
{
WRITE_FILE(ptr, length, "bik");
ptr += length;
continue;
}
if (formats & BMP && IS_BMP_MAGIC(ptr) && bmp_isfile(ptr, input_len, &length))
{
WRITE_FILE(ptr, length, "bmp");
ptr += length;
continue;
}
if (formats & SMK && IS_SMK_MAGIC(magic) && smk_isfile(ptr, input_len, &length))
{
WRITE_FILE(ptr, length, "smk");
ptr += length;
continue;
}
if (formats & MP4 && input_len > MP4_HEADER_SIZE &&
MAGIC(ptr + MP4_MAGIC_OFFSET) == MP4_MAGIC &&
mp4_isfile(ptr, input_len, &info))
{
WRITE_FILE(ptr, info.length, info.ext);
ptr += info.length;
continue;
}
if (formats & S3M && input_len > S3M_MAGIC_OFFSET + 4 &&
MAGIC(ptr + S3M_MAGIC_OFFSET) == S3M_MAGIC &&
s3m_isfile(ptr, input_len, &length))
{
WRITE_FILE(ptr, length, "s3m");
ptr += length;
continue;
}
if (formats & MOD && input_len > MOD_MAGIC_OFFSET + 4)
{
const uint8_t *modmagic = ptr + MOD_MAGIC_OFFSET;
if (IS_MOD_MAGIC(modmagic) && mod_isfile(ptr, input_len, &length))
{
WRITE_FILE(ptr, length, "mod");
ptr += length;
continue;
}
}
if (formats & UTF_32LE && utf32le_isfile(ptr, input_len, &info)) {
WRITE_FILE(ptr, info.length, info.ext);
ptr += info.length;
continue;
}
if (formats & UTF_32BE && utf32be_isfile(ptr, input_len, &info)) {
WRITE_FILE(ptr, info.length, info.ext);
ptr += info.length;
continue;
}
if (formats & UTF_16LE && utf16le_isfile(ptr, input_len, &info)) {
WRITE_FILE(ptr, info.length, info.ext);
ptr += info.length;
continue;
}
if (formats & UTF_16BE && utf16be_isfile(ptr, input_len, &info)) {
WRITE_FILE(ptr, info.length, info.ext);
ptr += info.length;
continue;
}
if (formats & UTF_8 && utf8_isfile(ptr, input_len, &info)) {
WRITE_FILE(ptr, info.length, info.ext);
ptr += info.length;
continue;
}
else if (formats & ASCII && ascii_isfile(ptr, input_len, &info)) {
WRITE_FILE(ptr, info.length, info.ext);
ptr += info.length;
continue;
}
++ ptr;
}
goto cleanup;
error:
success = 0;
cleanup:
if (outfilename)
free(outfilename);
if (numfilesptr)
*numfilesptr = numfiles;
if (sumsizeptr)
*sumsizeptr = sumsize;
return success;
}