Download-Tools/mkvtoolnix
1
0
Fork 0
mirror of https://gitlab.com/mbunkus/mkvtoolnix.git synced 2024-12-24 20:01:53 +00:00
Code Issues Actions Packages Projects Releases Wiki Activity
fb5fd4661a
mkvtoolnix/mkvextract.cpp

1363 lines
44 KiB
C++
Raw Blame History

/*
mkvextract -- extract tracks from Matroska files into other files
mkvextract.cpp
Written by Moritz Bunkus <moritz@bunkus.org>
Distributed under the GPL
see the file COPYING for details
or visit http://www.gnu.org/copyleft/gpl.html
*/
/*!
\file
\version $Id$
\brief extracts tracks from Matroska files into other files
\author Moritz Bunkus <moritz@bunkus.org>
*/
#include <errno.h>
#include <ctype.h>
#include <stdarg.h>
#include <stdlib.h>
#include <stdio.h>
#include <string.h>
#include <time.h>
#if defined(COMP_MSC)
#include <assert.h>
#else
#include <unistd.h>
#endif
#include <iostream>
#include <string>
#include <vector>
#include <algorithm>
#include <ogg/ogg.h>
#include <vorbis/codec.h>
extern "C" {
#include "avilib/avilib.h"
}
#include "EbmlHead.h"
#include "EbmlSubHead.h"
#include "EbmlStream.h"
#include "EbmlVoid.h"
#include "FileKax.h"
#include "KaxAttachements.h"
#include "KaxBlock.h"
#include "KaxBlockData.h"
#include "KaxChapters.h"
#include "KaxCluster.h"
#include "KaxClusterData.h"
#include "KaxCues.h"
#include "KaxCuesData.h"
#include "KaxInfo.h"
#include "KaxInfoData.h"
#include "KaxSeekHead.h"
#include "KaxSegment.h"
#include "KaxTags.h"
#include "KaxTracks.h"
#include "KaxTrackEntryData.h"
#include "KaxTrackAudio.h"
#include "KaxTrackVideo.h"
#include "common.h"
#include "matroska.h"
#include "mm_io.h"
using namespace LIBMATROSKA_NAMESPACE;
using namespace std;
#define NAME "mkvextract"
class ssa_line_c {
public:
char *line;
int num;
bool operator < (const ssa_line_c &cmp) const;
};
bool ssa_line_c::operator < (const ssa_line_c &cmp) const {
return num < cmp.num;
}
typedef struct {
char *out_name;
mm_io_c *mm_io;
avi_t *avi;
ogg_stream_state osstate;
int64_t tid;
bool in_use;
char track_type;
int type;
char *codec_id;
void *private_data;
int private_size;
float a_sfreq;
int a_channels, a_bps;
float v_fps;
int v_width, v_height;
int srt_num;
int conv_handle;
vector<ssa_line_c> ssa_lines;
wave_header wh;
int64_t bytes_written;
unsigned char *buffered_data;
int buffered_size;
int64_t packetno, last_end;
int header_sizes[3];
unsigned char *headers[3];
int aac_id, aac_profile, aac_srate_idx;
} mkv_track_t;
vector<mkv_track_t> tracks;
mkv_track_t *find_track(int tid) {
int i;
for (i = 0; i < tracks.size(); i++)
if (tracks[i].tid == tid)
return &tracks[i];
return NULL;
}
char typenames[14][20] = {"unknown", "Ogg" "AVI", "WAV", "SRT", "MP3", "AC3",
"chapter", "MicroDVD", "VobSub", "Matroska", "DTS",
"AAC", "SSA/ASS"};
void usage() {
fprintf(stdout,
"Usage: mkvextract -i inname [TID1:outname1 [TID2:outname2 ...]]\n\n"
" -i inname Use 'inname' as the source.\n"
" -c charset Convert text subtitles to this charset.\n"
" TID:outname Write track with the ID TID to 'outname'.\n"
"\n"
" -v, --verbose Increase verbosity.\n"
" -h, --help Show this help.\n"
" -V, --version Show version information.\n");
}
void parse_args(int argc, char **argv, char *&file_name) {
int i, conv_handle;
char *colon, *copy;
int64_t tid;
mkv_track_t track;
file_name = NULL;
verbose = 0;
// Find options that directly end the program.
for (i = 1; i < argc; i++)
if (!strcmp(argv[i], "-V") || !strcmp(argv[i], "--version")) {
fprintf(stdout, VERSIONINFO "\n");
exit(0);
} else if (!strcmp(argv[i], "-h") || !strcmp(argv[i], "-?") ||
!strcmp(argv[i], "--help")) {
usage();
exit(0);
}
conv_handle = 0;
// Now process all the other options.
for (i = 1; i < argc; i++)
if (!strcmp(argv[i], "-v") || !strcmp(argv[i], "--verbose"))
verbose++;
else if (!strcmp(argv[i], "-i")) {
if ((i + 1) >= argc) {
fprintf(stderr, "Error: -i lacks a file name.\n");
exit(1);
} else if (file_name != NULL) {
fprintf(stderr, "Error: Only one input file is allowed.\n");
exit(1);
}
file_name = argv[i + 1];
i++;
} else if (!strcmp(argv[i], "-c")) {
if ((i + 1) >= argc) {
fprintf(stderr, "Error: -c lacks a charset.\n");
exit(1);
}
conv_handle = utf8_init(argv[i + 1]);
i++;
} else {
copy = safestrdup(argv[i]);
colon = strchr(copy, ':');
if (colon == NULL) {
fprintf(stderr, "Error: Missing track ID in argument '%s'.\n",
argv[i]);
exit(1);
}
*colon = 0;
if (!parse_int(copy, tid) || (tid < 0)) {
fprintf(stderr, "Error: Invalid track ID in argument '%s'.\n",
argv[i]);
exit(1);
}
colon++;
if (*colon == 0) {
fprintf(stderr, "Error: Missing output filename in argument '%s'.\n",
argv[i]);
exit(1);
}
memset(&track, 0, sizeof(mkv_track_t));
track.tid = tid;
track.out_name = safestrdup(colon);
track.conv_handle = conv_handle;
tracks.push_back(track);
safefree(copy);
}
if (file_name == NULL) {
fprintf(stderr, "Error: No input file given.\n\n");
usage();
exit(0);
}
if (tracks.size() == 0) {
fprintf(stdout, "Nothing to do.\n\n");
usage();
exit(0);
}
}
#define ARGS_BUFFER_LEN (200 * 1024) // Ok let's be ridiculous here :)
static char args_buffer[ARGS_BUFFER_LEN];
void show_element(EbmlElement *l, int level, const char *fmt, ...) {
va_list ap;
char level_buffer[10];
if (level > 9)
die("mkvextract.cpp/show_element(): level > 9: %d", level);
if (verbose == 0)
return;
va_start(ap, fmt);
args_buffer[ARGS_BUFFER_LEN - 1] = 0;
vsnprintf(args_buffer, ARGS_BUFFER_LEN - 1, fmt, ap);
va_end(ap);
memset(&level_buffer[1], ' ', 9);
level_buffer[0] = '|';
level_buffer[level] = 0;
fprintf(stdout, "(%s) %s+ %s", NAME, level_buffer, args_buffer);
if (l != NULL)
fprintf(stdout, " at %llu", l->GetElementPosition());
fprintf(stdout, "\n");
}
void show_error(const char *fmt, ...) {
va_list ap;
va_start(ap, fmt);
args_buffer[ARGS_BUFFER_LEN - 1] = 0;
vsnprintf(args_buffer, ARGS_BUFFER_LEN - 1, fmt, ap);
va_end(ap);
fprintf(stdout, "(%s) %s\n", NAME, args_buffer);
}
void flush_ogg_pages(mkv_track_t &track) {
ogg_page page;
while (ogg_stream_flush(&track.osstate, &page)) {
track.mm_io->write(page.header, page.header_len);
track.mm_io->write(page.body, page.body_len);
}
}
void write_ogg_pages(mkv_track_t &track) {
ogg_page page;
while (ogg_stream_pageout(&track.osstate, &page)) {
track.mm_io->write(page.header, page.header_len);
track.mm_io->write(page.body, page.body_len);
}
}
void create_output_files() {
int i, k, offset;
bool something_to_do, is_ok;
unsigned char *c;
ogg_packet op;
something_to_do = false;
for (i = 0; i < tracks.size(); i++) {
if (tracks[i].in_use) {
tracks[i].in_use = false;
if (tracks[i].codec_id == NULL) {
fprintf(stdout, "Warning: Track ID %lld is missing the CodecID. "
"Skipping track.\n", tracks[i].tid);
continue;
}
// Video tracks:
if (tracks[i].track_type == 'v') {
tracks[i].type = TYPEAVI;
if ((tracks[i].v_width == 0) || (tracks[i].v_height == 0) ||
(tracks[i].v_fps == 0.0) || (tracks[i].private_data == NULL) ||
(tracks[i].private_size < sizeof(alBITMAPINFOHEADER))) {
fprintf(stdout, "Warning: Track ID %lld is missing some critical "
"information. Skipping track.\n", tracks[i].tid);
continue;
}
if (strcmp(tracks[i].codec_id, MKV_V_MSCOMP)) {
fprintf(stdout, "Warning: Extraction of video tracks with a CodecId "
"other than " MKV_V_MSCOMP " is not supported at the "
"moment. Skipping track %lld.\n", tracks[i].tid);
continue;
}
} else if (tracks[i].track_type == 'a') {
if ((tracks[i].a_sfreq == 0.0) || (tracks[i].a_channels == 0)) {
fprintf(stdout, "Warning: Track ID %lld is missing some critical "
"information. Skipping track.\n", tracks[i].tid);
continue;
}
if (!strcmp(tracks[i].codec_id, MKV_A_VORBIS)) {
tracks[i].type = TYPEOGM; // Yeah, I know, I know...
if (tracks[i].private_data == NULL) {
fprintf(stdout, "Warning: Track ID %lld is missing some critical "
"information. Skipping track.\n", tracks[i].tid);
continue;
}
c = (unsigned char *)tracks[i].private_data;
if (c[0] != 2) {
fprintf(stderr, "Error: Vorbis track ID %lld does not contain "
"valid headers. Skipping track.\n", tracks[i].tid);
continue;
}
is_ok = true;
offset = 1;
for (k = 0; k < 2; k++) {
int length = 0;
while ((c[offset] == (unsigned char)255) &&
(length < tracks[i].private_size)) {
length += 255;
offset++;
}
if (offset >= (tracks[i].private_size - 1)) {
fprintf(stderr, "Error: Vorbis track ID %lld does not contain "
"valid headers. Skipping track.\n", tracks[i].tid);
is_ok = false;
break;
}
length += c[offset];
offset++;
tracks[i].header_sizes[k] = length;
}
if (!is_ok)
continue;
tracks[i].headers[0] = &c[offset];
tracks[i].headers[1] = &c[offset + tracks[i].header_sizes[0]];
tracks[i].headers[2] = &c[offset + tracks[i].header_sizes[0] +
tracks[i].header_sizes[1]];
tracks[i].header_sizes[2] = tracks[i].private_size -
offset - tracks[i].header_sizes[0] - tracks[i].header_sizes[1];
} else if (!strcmp(tracks[i].codec_id, MKV_A_MP3))
tracks[i].type = TYPEMP3;
else if (!strcmp(tracks[i].codec_id, MKV_A_AC3))
tracks[i].type = TYPEAC3;
else if (!strcmp(tracks[i].codec_id, MKV_A_PCM)) {
tracks[i].type = TYPEWAV; // Yeah, I know, I know...
if (tracks[i].a_bps == 0) {
fprintf(stdout, "Warning: Track ID %lld is missing some critical "
"information. Skipping track.\n", tracks[i].tid);
continue;
}
} else if (!strncmp(tracks[i].codec_id, "A_AAC", 5)) {
if (strlen(tracks[i].codec_id) < strlen("A_AAC/MPEG4/LC")) {
fprintf(stdout, "Warning: Track ID %lld has an unknown AAC "
"type. Skipping track.\n", tracks[i].tid);
continue;
}
// A_AAC/MPEG4/MAIN
// 0123456789012345
if (tracks[i].codec_id[10] == '4')
tracks[i].aac_id = 0;
else if (tracks[i].codec_id[10] == '2')
tracks[i].aac_id = 1;
else {
fprintf(stdout, "Warning: Track ID %lld has an unknown AAC "
"type. Skipping track.\n", tracks[i].tid);
continue;
}
if (!strcmp(&tracks[i].codec_id[12], "MAIN"))
tracks[i].aac_profile = 0;
else if (!strcmp(&tracks[i].codec_id[12], "LC"))
tracks[i].aac_profile = 1;
else if (!strcmp(&tracks[i].codec_id[12], "SSR"))
tracks[i].aac_profile = 2;
else if (!strcmp(&tracks[i].codec_id[12], "LTP"))
tracks[i].aac_profile = 3;
else {
fprintf(stdout, "Warning: Track ID %lld has an unknown AAC "
"type. Skipping track.\n", tracks[i].tid);
continue;
}
if (92017 <= tracks[i].a_sfreq)
tracks[i].aac_srate_idx = 0;
else if (75132 <= tracks[i].a_sfreq)
tracks[i].aac_srate_idx = 1;
else if (55426 <= tracks[i].a_sfreq)
tracks[i].aac_srate_idx = 2;
else if (46009 <= tracks[i].a_sfreq)
tracks[i].aac_srate_idx = 3;
else if (37566 <= tracks[i].a_sfreq)
tracks[i].aac_srate_idx = 4;
else if (27713 <= tracks[i].a_sfreq)
tracks[i].aac_srate_idx = 5;
else if (23004 <= tracks[i].a_sfreq)
tracks[i].aac_srate_idx = 6;
else if (18783 <= tracks[i].a_sfreq)
tracks[i].aac_srate_idx = 7;
else if (13856 <= tracks[i].a_sfreq)
tracks[i].aac_srate_idx = 8;
else if (11502 <= tracks[i].a_sfreq)
tracks[i].aac_srate_idx = 9;
else if (9391 <= tracks[i].a_sfreq)
tracks[i].aac_srate_idx = 10;
else
tracks[i].aac_srate_idx = 11;
tracks[i].type = TYPEAAC;
} else if (!strcmp(tracks[i].codec_id, MKV_A_DTS)) {
fprintf(stdout, "Warning: Extraction of DTS is not supported - yet. "
"I promise I'll implement it. Really Soon Now (tm)! "
"Skipping track.\n");
continue;
} else {
fprintf(stdout, "Warning: Unsupported CodecID '%s' for ID %lld. "
"Skipping track.\n", tracks[i].codec_id, tracks[i].tid);
continue;
}
} else if (tracks[i].track_type == 's') {
if (!strcmp(tracks[i].codec_id, MKV_S_TEXTUTF8))
tracks[i].type = TYPESRT;
else if (!strcmp(tracks[i].codec_id, MKV_S_TEXTSSA) ||
!strcmp(tracks[i].codec_id, MKV_S_TEXTASS))
tracks[i].type = TYPESSA;
else {
fprintf(stdout, "Warning: Unsupported CodecID '%s' for ID %lld. "
"Skipping track.\n", tracks[i].codec_id, tracks[i].tid);
continue;
}
} else {
fprintf(stdout, "Warning: Unknown track type for ID %lld. Skipping "
"track.\n", tracks[i].tid);
continue;
}
tracks[i].in_use = true;
something_to_do = true;
} else
fprintf(stdout, "Warning: There is no track with the ID '%lld' in the "
"input file.\n", tracks[i].tid);
}
if (!something_to_do) {
fprintf(stdout, "Nothing to do. Exiting.\n");
exit(0);
}
// Now finally create some files. Hell, I *WANT* to create something now.
// RIGHT NOW! Or I'll go insane...
for (i = 0; i < tracks.size(); i++) {
if (tracks[i].in_use) {
if (tracks[i].type == TYPEAVI) {
alBITMAPINFOHEADER *bih;
char ccodec[5];
tracks[i].avi = AVI_open_output_file(tracks[i].out_name);
if (tracks[i].avi == NULL) {
fprintf(stderr, "Error: Could not create '%s'. Reason: %s.\n",
tracks[i].out_name, AVI_strerror());
exit(1);
}
bih = (alBITMAPINFOHEADER *)tracks[i].private_data;
memcpy(ccodec, &bih->bi_compression, 4);
ccodec[4] = 0;
AVI_set_video(tracks[i].avi, tracks[i].v_width, tracks[i].v_height,
tracks[i].v_fps, ccodec);
fprintf(stdout, "Extracting track ID %lld to an AVI file '%s'.\n",
tracks[i].tid, tracks[i].out_name);
} else {
try {
tracks[i].mm_io = new mm_io_c(tracks[i].out_name, MODE_CREATE);
} catch (exception &ex) {
fprintf(stderr, "Error: Could not create '%s'. Reason: %d (%s). "
"Aborting.\n", tracks[i].out_name, errno, strerror(errno));
exit(1);
}
fprintf(stdout, "Extracting track ID %lld to a %s file '%s'.\n",
tracks[i].tid, typenames[tracks[i].type - 1],
tracks[i].out_name);
if (tracks[i].type == TYPEOGM) {
ogg_stream_init(&tracks[i].osstate, rand());
// Handle the three header packets: Headers, comments, codec
// setup data.
op.b_o_s = 1;
op.e_o_s = 0;
op.packetno = 0;
op.packet = tracks[i].headers[0];
op.bytes = tracks[i].header_sizes[0];
op.granulepos = 0;
ogg_stream_packetin(&tracks[i].osstate, &op);
flush_ogg_pages(tracks[i]);
op.b_o_s = 0;
op.packetno = 1;
op.packet = tracks[i].headers[1];
op.bytes = tracks[i].header_sizes[1];
ogg_stream_packetin(&tracks[i].osstate, &op);
op.packetno = 2;
op.packet = tracks[i].headers[2];
op.bytes = tracks[i].header_sizes[2];
ogg_stream_packetin(&tracks[i].osstate, &op);
flush_ogg_pages(tracks[i]);
tracks[i].packetno = 3;
} else if (tracks[i].type == TYPEWAV) {
wave_header *wh = &tracks[i].wh;
// Write the WAV header.
memcpy(&wh->riff.id, "RIFF", 4);
memcpy(&wh->riff.wave_id, "WAVE", 4);
memcpy(&wh->format.id, "fmt ", 4);
put_uint32(&wh->format.len, 16);
put_uint16(&wh->common.wFormatTag, 1);
put_uint16(&wh->common.wChannels, tracks[i].a_channels);
put_uint32(&wh->common.dwSamplesPerSec, (int)tracks[i].a_sfreq);
put_uint32(&wh->common.dwAvgBytesPerSec,
tracks[i].a_channels * (int)tracks[i].a_sfreq *
tracks[i].a_bps / 8);
put_uint16(&wh->common.wBlockAlign, 4);
put_uint16(&wh->common.wBitsPerSample, tracks[i].a_bps);
memcpy(&wh->data.id, "data", 4);
tracks[i].mm_io->write(wh, sizeof(wave_header));
} else if (tracks[i].type == TYPESRT)
tracks[i].srt_num = 1;
else if (tracks[i].type == TYPESSA) {
char *s;
s = (char *)safemalloc(tracks[i].private_size + 1);
memcpy(s, tracks[i].private_data, tracks[i].private_size);
s[tracks[i].private_size] = 0;
tracks[i].mm_io->puts_unl(s);
tracks[i].mm_io->puts_unl("\n[Events]\nFormat: Marked, Start, End, "
"Style, Name, MarginL, MarginR, MarginV, "
"Effect, Text\n");
safefree(s);
}
}
}
}
}
void handle_data(KaxBlock *block, int64_t block_duration, bool has_ref) {
mkv_track_t *track;
int i, len, num;
int64_t start, end;
char *s, buffer[200], *s2, adts[56 / 8];
vector<string> fields;
string line, comma = ",";
ssa_line_c ssa_line;
track = find_track(block->TrackNum());
if ((track == NULL) || !track->in_use){
delete block;
return;
}
start = block->GlobalTimecode() / 1000000; // in ms
end = start + block_duration;
for (i = 0; i < block->NumberFrames(); i++) {
DataBuffer &data = block->GetBuffer(i);
switch (track->type) {
case TYPEAVI:
AVI_write_frame(track->avi, (char *)data.Buffer(), data.Size(),
has_ref ? 0 : 1);
break;
case TYPEOGM:
if (track->buffered_data != NULL) {
ogg_packet op;
op.b_o_s = 0;
op.e_o_s = 0;
op.packetno = track->packetno;
op.packet = track->buffered_data;
op.bytes = track->buffered_size;
op.granulepos = start * (int64_t)track->a_sfreq / 1000;
ogg_stream_packetin(&track->osstate, &op);
safefree(track->buffered_data);
write_ogg_pages(*track);
track->packetno++;
}
track->buffered_data = (unsigned char *)safememdup(data.Buffer(),
data.Size());
track->buffered_size = data.Size();
track->last_end = end;
break;
case TYPESRT:
// Do the charset conversion.
s = (char *)safemalloc(data.Size() + 1);
memcpy(s, data.Buffer(), data.Size());
s[data.Size()] = 0;
s2 = from_utf8(tracks[i].conv_handle, s);
safefree(s);
len = strlen(s2);
s = (char *)safemalloc(len + 3);
strcpy(s, s2);
safefree(s2);
s[len] = '\n';
s[len + 1] = '\n';
s[len + 2] = 0;
// Print the entry's number.
sprintf(buffer, "%d\n", tracks[i].srt_num);
tracks[i].srt_num++;
tracks[i].mm_io->write(buffer, strlen(buffer));
// Print the timestamps.
sprintf(buffer, "%02lld:%02lld:%02lld,%03lld --> %02lld:%02lld:%02lld,"
"%03lld\n",
start / 1000 / 60 / 60, (start / 1000 / 60) % 60,
(start / 1000) % 60, start % 1000,
end / 1000 / 60 / 60, (end / 1000 / 60) % 60,
(end / 1000) % 60, end % 1000);
tracks[i].mm_io->write(buffer, strlen(buffer));
// Print the text itself.
tracks[i].mm_io->puts_unl(s);
safefree(s);
break;
case TYPESSA:
s = (char *)safemalloc(data.Size() + 1);
memcpy(s, data.Buffer(), data.Size());
s[data.Size()] = 0;
// Split the line into the fields.
// Specs say that the following fields are to put into the block:
// 0: ReadOrder, 1: Layer, 2: Style, 3: Name, 4: MarginL, 5: MarginR,
// 6: MarginV, 7: Effect, 8: Text
fields = split(s, ",", 9);
safefree(s);
if (fields.size() != 9) {
fprintf(stdout, "Warning: Invalid format for a SSA line ('%s'). "
"Ignoring this entry.\n", s);
continue;
}
// Convert the ReadOrder entry so that we can re-order the entries
// later.
if (!parse_int(fields[0].c_str(), num)) {
fprintf(stdout, "Warning: Invalid format for a SSA line ('%s'). "
"Ignoring this entry.\n", s);
continue;
}
// Reconstruct the 'original' line. It'll look like this for SSA:
// Marked, Start, End, Style, Name, MarginL, MarginR, MarginV, Effect,
// Text
line = string("Dialogue: Marked=0,");
// Append the start and end time.
sprintf(buffer, "%lld:%02lld:%02lld.%02lld",
start / 1000 / 60 / 60, (start / 1000 / 60) % 60,
(start / 1000) % 60, (start % 1000) / 10);
line += string(buffer) + comma;
sprintf(buffer, "%lld:%02lld:%02lld.%02lld",
end / 1000 / 60 / 60, (end / 1000 / 60) % 60,
(end / 1000) % 60, (end % 1000) / 10);
line += string(buffer) + comma;
// Append the other fields.
line += fields[2] + comma + // Style
fields[3] + comma + // Name
fields[4] + comma + // MarginL
fields[5] + comma + // MarginR
fields[6] + comma + // MarginV
fields[7] + comma; // Effect
// Do the charset conversion.
s = from_utf8(tracks[i].conv_handle, fields[8].c_str());
line += string(s) + "\n";
safefree(s);
// Now store that entry.
ssa_line.num = num;
ssa_line.line = safestrdup(line.c_str());
tracks[i].ssa_lines.push_back(ssa_line);
break;
case TYPEAAC:
// Recreate the ADTS headers. What a fun. Like runing headlong into
// a solid wall. But less painful. Well such is life, you know.
// But then again I've just seen a beautiful girl walking by my
// window, and suddenly the world is a bright place. Everything's
// a matter of perspective. And if I didn't enjoy writing even this
// code then I wouldn't do it at all. So let's get to it!
memset(adts, 0, 56 / 8);
// sync word, 12 bits
adts[0] = 0xff;
adts[1] = 0xf0;
// ID, 1 bit
adts[1] |= track->aac_id << 3;
// layer: 2 bits = 00
// protection absent: 1 bit = 1 (ASSUMPTION!)
adts[1] |= 1;
// profile, 2 bits
adts[2] = track->aac_profile << 6;
// sampling frequency index, 4 bits
adts[2] |= track->aac_srate_idx << 2;
// private, 1 bit = 0 (ASSUMPTION!)
// channels, 3 bits
adts[2] |= (track->a_channels & 4) >> 2;
adts[3] |= (track->a_channels & 3) << 6;
// original/copy, 1 bit = 0(ASSUMPTION!)
// home, 1 bit = 0 (ASSUMPTION!)
// copyright id bit, 1 bit = 0 (ASSUMPTION!)
// copyright id start, 1 bit = 0 (ASSUMPTION!)
// frame length, 13 bits
len = data.Size() + 7;
adts[3] |= len >> 11;
adts[4] = (len >> 3) & 0xff;
adts[5] = (len & 7) << 5;
// adts buffer fullness, 11 bits, 0x7ff = VBR (ASSUMPTION!)
adts[5] |= 0x1f;
adts[6] = 0xfc;
// number of raw frames, 2 bits, 0 (meaning 1 frame) (ASSUMPTION!)
// Write the ADTS header and the data itself.
track->mm_io->write(adts, 56 / 8);
track->mm_io->write(data.Buffer(), data.Size());
break;
default:
track->mm_io->write(data.Buffer(), data.Size());
track->bytes_written += data.Size();
}
}
delete block;
}
void close_files() {
int i, k;
ogg_packet op;
for (i = 0; i < tracks.size(); i++) {
if (tracks[i].in_use) {
switch (tracks[i].type) {
case TYPEAVI:
AVI_close(tracks[i].avi);
break;
case TYPEOGM:
// Set the "end of stream" marker on the last packet, handle it
// and flush all remaining Ogg pages.
op.b_o_s = 0;
op.e_o_s = 1;
op.packetno = tracks[i].packetno;
op.packet = tracks[i].buffered_data;
op.bytes = tracks[i].buffered_size;
op.granulepos = tracks[i].last_end * (int64_t)tracks[i].a_sfreq /
1000;
ogg_stream_packetin(&tracks[i].osstate, &op);
safefree(tracks[i].buffered_data);
flush_ogg_pages(tracks[i]);
delete tracks[i].mm_io;
break;
case TYPEWAV:
// Fix the header with the real number of bytes written.
tracks[i].mm_io->setFilePointer(0);
tracks[i].wh.riff.len = tracks[i].bytes_written + 36;
tracks[i].wh.data.len = tracks[i].bytes_written;
tracks[i].mm_io->write(&tracks[i].wh, sizeof(wave_header));
delete tracks[i].mm_io;
break;
case TYPESSA:
// Sort the SSA lines according to their ReadOrder number and
// write them.
sort(tracks[i].ssa_lines.begin(), tracks[i].ssa_lines.end());
for (k = 0; k < tracks[i].ssa_lines.size(); k++) {
tracks[i].mm_io->puts_unl(tracks[i].ssa_lines[k].line);
safefree(tracks[i].ssa_lines[k].line);
}
delete tracks[i].mm_io;
break;
default:
delete tracks[i].mm_io;
}
}
}
}
bool process_file(const char *file_name) {
int upper_lvl_el;
// Elements for different levels
EbmlElement *l0 = NULL, *l1 = NULL, *l2 = NULL, *l3 = NULL, *l4 = NULL;
EbmlStream *es;
KaxCluster *cluster;
KaxBlock *block;
mkv_track_t *track;
uint64_t cluster_tc, tc_scale = TIMECODE_SCALE, file_size;
char mkv_track_type;
bool ms_compat, delete_element, has_reference;
int64_t block_duration;
mm_io_c *in;
// open input file
try {
in = new mm_io_c(file_name, MODE_READ);
} catch (std::exception &ex) {
show_error("Error: Couldn't open input file %s (%s).", file_name,
strerror(errno));
return false;
}
in->setFilePointer(0, seek_end);
file_size = in->getFilePointer();
in->setFilePointer(0, seek_beginning);
try {
es = new EbmlStream(*in);
// Find the EbmlHead element. Must be the first one.
l0 = es->FindNextID(EbmlHead::ClassInfos, 0xFFFFFFFFL);
if (l0 == NULL) {
show_error("Error: No EBML head found.");
delete es;
return false;
}
// Don't verify its data for now.
l0->SkipData(*es, l0->Generic().Context);
delete l0;
while (1) {
// Next element must be a segment
l0 = es->FindNextID(KaxSegment::ClassInfos, 0xFFFFFFFFL);
if (l0 == NULL) {
show_error("No segment/level 0 element found.");
return false;
}
if (EbmlId(*l0) == KaxSegment::ClassInfos.GlobalId) {
show_element(l0, 0, "Segment");
break;
}
show_element(l0, 0, "Next level 0 element is not a segment but %s",
typeid(*l0).name());
l0->SkipData(*es, l0->Generic().Context);
delete l0;
}
upper_lvl_el = 0;
// We've got our segment, so let's find the tracks
l1 = es->FindNextElement(l0->Generic().Context, upper_lvl_el, 0xFFFFFFFFL,
true, 1);
while (l1 != NULL) {
if (upper_lvl_el != 0)
break;
if (EbmlId(*l1) == KaxInfo::ClassInfos.GlobalId) {
// General info about this Matroska file
show_element(l1, 1, "Segment information");
l2 = es->FindNextElement(l1->Generic().Context, upper_lvl_el,
0xFFFFFFFFL, true, 1);
while (l2 != NULL) {
if (upper_lvl_el != 0)
break;
if (EbmlId(*l2) == KaxTimecodeScale::ClassInfos.GlobalId) {
KaxTimecodeScale &ktc_scale = *static_cast<KaxTimecodeScale *>(l2);
ktc_scale.ReadData(es->I_O());
tc_scale = uint64(ktc_scale);
show_element(l2, 2, "Timecode scale: %llu", tc_scale);
}
l2->SkipData(*es, l2->Generic().Context);
delete l2;
l2 = es->FindNextElement(l1->Generic().Context, upper_lvl_el,
0xFFFFFFFFL, true, 1);
}
} else if (EbmlId(*l1) == KaxTracks::ClassInfos.GlobalId) {
// Yep, we've found our KaxTracks element. Now find all tracks
// contained in this segment.
show_element(l1, 1, "Segment tracks");
l2 = es->FindNextElement(l1->Generic().Context, upper_lvl_el,
0xFFFFFFFFL, true, 1);
while (l2 != NULL) {
if (upper_lvl_el != 0)
break;
if (EbmlId(*l2) == KaxTrackEntry::ClassInfos.GlobalId) {
// We actually found a track entry :) We're happy now.
show_element(l2, 2, "A track");
track = NULL;
mkv_track_type = '?';
ms_compat = false;
l3 = es->FindNextElement(l2->Generic().Context, upper_lvl_el,
0xFFFFFFFFL, true, 1);
while (l3 != NULL) {
if (upper_lvl_el != 0)
break;
// Now evaluate the data belonging to this track
if (EbmlId(*l3) == KaxTrackNumber::ClassInfos.GlobalId) {
char *msg;
KaxTrackNumber &tnum = *static_cast<KaxTrackNumber *>(l3);
tnum.ReadData(es->I_O());
if ((track = find_track(uint32(tnum))) == NULL)
msg = "extraction not requested";
else {
msg = "will extract this track";
track->in_use = true;
}
show_element(l3, 3, "Track number: %u (%s)", uint32(tnum),
msg);
} else if (EbmlId(*l3) == KaxTrackType::ClassInfos.GlobalId) {
KaxTrackType &ttype = *static_cast<KaxTrackType *>(l3);
ttype.ReadData(es->I_O());
switch (uint8(ttype)) {
case track_audio:
mkv_track_type = 'a';
break;
case track_video:
mkv_track_type = 'v';
break;
case track_subtitle:
mkv_track_type = 's';
break;
default:
mkv_track_type = '?';
break;
}
show_element(l3, 3, "Track type: %s",
mkv_track_type == 'a' ? "audio" :
mkv_track_type == 'v' ? "video" :
mkv_track_type == 's' ? "subtitles" :
"unknown");
if (track != NULL)
track->track_type = mkv_track_type;
} else if (EbmlId(*l3) == KaxTrackAudio::ClassInfos.GlobalId) {
show_element(l3, 3, "Audio track");
l4 = es->FindNextElement(l3->Generic().Context, upper_lvl_el,
0xFFFFFFFFL, true, 1);
while (l4 != NULL) {
if (upper_lvl_el != 0)
break;
if (EbmlId(*l4) ==
KaxAudioSamplingFreq::ClassInfos.GlobalId) {
KaxAudioSamplingFreq &freq =
*static_cast<KaxAudioSamplingFreq*>(l4);
freq.ReadData(es->I_O());
show_element(l4, 4, "Sampling frequency: %f",
float(freq));
if (track != NULL)
track->a_sfreq = float(freq);
} else if (EbmlId(*l4) ==
KaxAudioChannels::ClassInfos.GlobalId) {
KaxAudioChannels &channels =
*static_cast<KaxAudioChannels*>(l4);
channels.ReadData(es->I_O());
show_element(l4, 4, "Channels: %u", uint8(channels));
if (track != NULL)
track->a_channels = uint8(channels);
} else if (EbmlId(*l4) ==
KaxAudioBitDepth::ClassInfos.GlobalId) {
KaxAudioBitDepth &bps =
*static_cast<KaxAudioBitDepth*>(l4);
bps.ReadData(es->I_O());
show_element(l4, 4, "Bit depth: %u", uint8(bps));
if (track != NULL)
track->a_bps = uint8(bps);
}
l4->SkipData(*es,
l4->Generic().Context);
delete l4;
l4 = es->FindNextElement(l3->Generic().Context,
upper_lvl_el, 0xFFFFFFFFL, true, 1);
} // while (l4 != NULL)
} else if (EbmlId(*l3) == KaxTrackVideo::ClassInfos.GlobalId) {
show_element(l3, 3, "Video track");
l4 = es->FindNextElement(l3->Generic().Context, upper_lvl_el,
0xFFFFFFFFL, true, 1);
while (l4 != NULL) {
if (upper_lvl_el != 0)
break;
if (EbmlId(*l4) == KaxVideoPixelWidth::ClassInfos.GlobalId) {
KaxVideoPixelWidth &width =
*static_cast<KaxVideoPixelWidth *>(l4);
width.ReadData(es->I_O());
show_element(l4, 4, "Pixel width: %u", uint16(width));
if (track != NULL)
track->v_width = uint16(width);
} else if (EbmlId(*l4) ==
KaxVideoPixelHeight::ClassInfos.GlobalId) {
KaxVideoPixelHeight &height =
*static_cast<KaxVideoPixelHeight *>(l4);
height.ReadData(es->I_O());
show_element(l4, 4, "Pixel height: %u", uint16(height));
if (track != NULL)
track->v_height = uint16(height);
} else if (EbmlId(*l4) ==
KaxVideoFrameRate::ClassInfos.GlobalId) {
KaxVideoFrameRate &framerate =
*static_cast<KaxVideoFrameRate *>(l4);
framerate.ReadData(es->I_O());
show_element(l4, 4, "Frame rate: %f", float(framerate));
if (track != NULL)
track->v_fps = float(framerate);
}
l4->SkipData(*es, l4->Generic().Context);
delete l4;
l4 = es->FindNextElement(l3->Generic().Context,
upper_lvl_el, 0xFFFFFFFFL, true, 1);
} // while (l4 != NULL)
} else if (EbmlId(*l3) == KaxCodecID::ClassInfos.GlobalId) {
KaxCodecID &codec_id = *static_cast<KaxCodecID*>(l3);
codec_id.ReadData(es->I_O());
show_element(l3, 3, "Codec ID: %s", string(codec_id).c_str());
if ((!strcmp(string(codec_id).c_str(), MKV_V_MSCOMP) &&
(mkv_track_type == 'v')) ||
(!strcmp(string(codec_id).c_str(), MKV_A_ACM) &&
(mkv_track_type == 'a')))
ms_compat = true;
if (track != NULL)
track->codec_id = safestrdup(string(codec_id).c_str());
} else if (EbmlId(*l3) == KaxCodecPrivate::ClassInfos.GlobalId) {
char pbuffer[100];
KaxCodecPrivate &c_priv = *static_cast<KaxCodecPrivate*>(l3);
c_priv.ReadData(es->I_O());
if (ms_compat && (mkv_track_type == 'v') &&
(c_priv.GetSize() >= sizeof(alBITMAPINFOHEADER))) {
alBITMAPINFOHEADER *bih =
(alBITMAPINFOHEADER *)&binary(c_priv);
unsigned char *fcc = (unsigned char *)&bih->bi_compression;
sprintf(pbuffer, " (FourCC: %c%c%c%c, 0x%08x)",
fcc[0], fcc[1], fcc[2], fcc[3],
get_uint32(&bih->bi_compression));
} else if (ms_compat && (mkv_track_type == 'a') &&
(c_priv.GetSize() >= sizeof(alWAVEFORMATEX))) {
alWAVEFORMATEX *wfe = (alWAVEFORMATEX *)&binary(c_priv);
sprintf(pbuffer, " (format tag: 0x%04x)",
get_uint16(&wfe->w_format_tag));
} else
pbuffer[0] = 0;
show_element(l3, 3, "CodecPrivate, length %llu%s",
c_priv.GetSize(), pbuffer);
if (track != NULL) {
track->private_data = safememdup(&binary(c_priv),
c_priv.GetSize());
track->private_size = c_priv.GetSize();
}
} else if (EbmlId(*l3) ==
KaxTrackDefaultDuration::ClassInfos.GlobalId) {
KaxTrackDefaultDuration &def_duration =
*static_cast<KaxTrackDefaultDuration*>(l3);
def_duration.ReadData(es->I_O());
show_element(l3, 3, "Default duration: %.3fms (%.3f fps for "
"a video track)",
(float)uint64(def_duration) / 1000000.0,
1000000000.0 / (float)uint64(def_duration));
if (track != NULL)
track->v_fps = 1000000000.0 / (float)uint64(def_duration);
}
if (upper_lvl_el > 0) { // we're coming from l4
upper_lvl_el--;
delete l3;
l3 = l4;
if (upper_lvl_el > 0)
break;
} else {
l3->SkipData(*es,
l3->Generic().Context);
delete l3;
l3 = es->FindNextElement(l2->Generic().Context, upper_lvl_el,
0xFFFFFFFFL, true, 1);
}
} // while (l3 != NULL)
}
if (upper_lvl_el > 0) { // we're coming from l3
upper_lvl_el--;
delete l2;
l2 = l3;
if (upper_lvl_el > 0)
break;
} else {
l2->SkipData(*es,
l2->Generic().Context);
delete l2;
l2 = es->FindNextElement(l1->Generic().Context, upper_lvl_el,
0xFFFFFFFFL, true, 1);
}
} // while (l2 != NULL)
// Headers have been parsed completely. Now create the output files
// and stuff.
create_output_files();
} else if (EbmlId(*l1) == KaxCluster::ClassInfos.GlobalId) {
show_element(l1, 1, "Cluster");
cluster = (KaxCluster *)l1;
if (verbose == 0) {
fprintf(stdout, "Progress: %d%%\r", (int)(in->getFilePointer() *
100 / file_size));
fflush(stdout);
}
l2 = es->FindNextElement(l1->Generic().Context, upper_lvl_el,
0xFFFFFFFFL, true, 1);
while (l2 != NULL) {
if (upper_lvl_el != 0)
break;
if (EbmlId(*l2) == KaxClusterTimecode::ClassInfos.GlobalId) {
KaxClusterTimecode &ctc = *static_cast<KaxClusterTimecode *>(l2);
ctc.ReadData(es->I_O());
cluster_tc = uint64(ctc);
show_element(l2, 2, "Cluster timecode: %.3fs",
(float)cluster_tc * (float)tc_scale / 1000000000.0);
cluster->InitTimecode(cluster_tc, tc_scale);
} else if (EbmlId(*l2) == KaxBlockGroup::ClassInfos.GlobalId) {
show_element(l2, 2, "Block group");
block_duration = 0;
has_reference = false;
l3 = es->FindNextElement(l2->Generic().Context, upper_lvl_el,
0xFFFFFFFFL, false, 1);
while (l3 != NULL) {
if (upper_lvl_el > 0)
break;
delete_element = true;
if (EbmlId(*l3) == KaxBlock::ClassInfos.GlobalId) {
block = (KaxBlock *)l3;
block->ReadData(es->I_O());
block->SetParent(*cluster);
show_element(l3, 3, "Block (track number %u, %d frame(s), "
"timecode %.3fs)", block->TrackNum(),
block->NumberFrames(),
(float)block->GlobalTimecode() / 1000000000.0);
delete_element = false;
} else if (EbmlId(*l3) ==
KaxBlockDuration::ClassInfos.GlobalId) {
KaxBlockDuration &duration =
*static_cast<KaxBlockDuration *>(l3);
duration.ReadData(es->I_O());
show_element(l3, 3, "Block duration: %.3fms",
((float)uint64(duration)) * tc_scale / 1000000.0);
block_duration = uint64(duration) * tc_scale / 1000000;
} else if (EbmlId(*l3) ==
KaxReferenceBlock::ClassInfos.GlobalId) {
KaxReferenceBlock &reference =
*static_cast<KaxReferenceBlock *>(l3);
reference.ReadData(es->I_O());
show_element(l3, 3, "Reference block: %.3fms",
((float)int64(reference)) * tc_scale / 1000000.0);
has_reference = true;
}
l3->SkipData(*es, l3->Generic().Context);
if (delete_element)
delete l3;
l3 = es->FindNextElement(l2->Generic().Context, upper_lvl_el,
0xFFFFFFFFL, true, 1);
} // while (l3 != NULL)
// Now write the stuff to the file. Or not. Or get something to
// drink. Or read a good book. Dunno, your choice, really.
handle_data(block, block_duration, has_reference);
}
if (upper_lvl_el > 0) { // we're coming from l3
upper_lvl_el--;
delete l2;
l2 = l3;
if (upper_lvl_el > 0)
break;
} else {
l2->SkipData(*es,
l2->Generic().Context);
delete l2;
l2 = es->FindNextElement(l1->Generic().Context, upper_lvl_el,
0xFFFFFFFFL, true, 1);
}
} // while (l2 != NULL)
}
if (upper_lvl_el > 0) { // we're coming from l2
upper_lvl_el--;
delete l1;
l1 = l2;
if (upper_lvl_el > 0)
break;
} else {
l1->SkipData(*es, l1->Generic().Context);
delete l1;
l1 = es->FindNextElement(l0->Generic().Context, upper_lvl_el,
0xFFFFFFFFL, true, 1);
}
} // while (l1 != NULL)
delete l0;
delete es;
delete in;
// Now just close the files and go to sleep. Mummy will sing you a
// lullaby. Just close your eyes, listen to her sweet voice, singing,
// singing, fading... fad... ing...
close_files();
return true;
} catch (exception &ex) {
show_error("Caught exception: %s", ex.what());
delete in;
return false;
}
}
int main(int argc, char **argv) {
char *input_file;
#if defined(SYS_UNIX)
nice(2);
#endif
utf8_init(NULL);
parse_args(argc, argv, input_file);
process_file(input_file);
if (verbose == 0)
fprintf(stdout, "Progress: 100%%\n");
utf8_done();
return 0;
}
Reference in New Issue View Git Blame Copy Permalink