ogginfo2.c 21 KB
Newer Older
1 2 3 4 5 6 7 8 9 10 11 12 13
/* Ogginfo
 *
 * A tool to describe ogg file contents and metadata.
 *
 * Copyright 2002 Michael Smith <msmith@layrinth.net.au>
 * Licensed under the GNU GPL, distributed with this program.
 */


#include <stdio.h>
#include <stdlib.h>
#include <errno.h>
#include <string.h>
14 15
#include <stdarg.h>
#include <getopt.h>
16 17 18 19 20 21 22 23 24 25

#include <ogg/ogg.h>
#include <vorbis/codec.h>

#include <locale.h>
#include "utf8.h"
#include "i18n.h"

#define CHUNK 4500

26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42
/* Different implementations have different format strings for 64 bit ints. */
#ifdef _WIN32
#define INT64FORMAT "%I64d"
#else
#define INT64FORMAT "%Ld"
#endif

/* TODO:
 *
 * - detect decreasing granulepos
 * - detect violations of muxing constraints
 * - better EOS detection (when EOS not explicitly set)
 * - detect granulepos 'gaps' (possibly vorbis-specific).
 * - check for serial number == (unsigned)-1 (will break some tools?)
 * - more options (e.g. less or more verbose)
 */

43 44 45 46 47 48
typedef struct _stream_processor {
    void (*process_page)(struct _stream_processor *, ogg_page *);
    void (*process_end)(struct _stream_processor *);
    int isillegal;
    int shownillegal;
    int isnew;
49 50
    long seqno;
    int lostseq;
51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80

    int start;
    int end;

    int num;
    char *type;

    ogg_uint32_t serial; /* must be 32 bit unsigned */
    ogg_stream_state os;
    void *data;
} stream_processor;

typedef struct {
    stream_processor *streams;
    int allocated;
    int used;

    int in_headers;
} stream_set;

typedef struct {
    vorbis_info vi;
    vorbis_comment vc;

    long bytes;
    ogg_int64_t lastgranulepos;

    int doneheaders;
} misc_vorbis_info;

81 82 83 84
static int printinfo = 1;
static int printwarn = 1;
static int verbose = 1;

85 86 87 88 89 90 91 92 93 94
static stream_set *create_stream_set(void) {
    stream_set *set = calloc(1, sizeof(stream_set));

    set->streams = calloc(5, sizeof(stream_processor));
    set->allocated = 5;
    set->used = 0;

    return set;
}

95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127
static void info(char *format, ...) 
{
    va_list ap;

    if(!printinfo)
        return;

    va_start(ap, format);
    vfprintf(stdout, format, ap);
    va_end(ap);
}

static void warn(char *format, ...) 
{
    va_list ap;

    if(!printwarn)
        return;

    va_start(ap, format);
    vfprintf(stdout, format, ap);
    va_end(ap);
}

static void error(char *format, ...) 
{
    va_list ap;

    va_start(ap, format);
    vfprintf(stdout, format, ap);
    va_end(ap);
}

128 129 130
static void vorbis_process(stream_processor *stream, ogg_page *page )
{
    ogg_packet packet;
131
    misc_vorbis_info *inf = stream->data;
132 133 134 135 136
    int i, header=0;

    ogg_stream_pagein(&stream->os, page);

    while(ogg_stream_packetout(&stream->os, &packet) > 0) {
137 138 139
        if(inf->doneheaders < 3) {
            if(vorbis_synthesis_headerin(&inf->vi, &inf->vc, &packet) < 0) {
                warn(_("Warning: Could not decode vorbis header "
140 141 142 143
                       "packet - invalid vorbis stream (%d)\n"), stream->num);
                continue;
            }
            header = 1;
144 145
            inf->doneheaders++;
            if(inf->doneheaders == 3) {
146 147 148 149 150
                if(ogg_page_granulepos(page) != 0 || ogg_page_continued(page))
                    warn(_("Warning: Vorbis stream %d does not have headers "
                           "correctly framed. Terminal header page contains "
                           "additional packets or has non-zero granulepos\n"),
                            stream->num);
151 152 153 154 155 156 157 158 159 160 161
                info(_("Vorbis headers parsed for stream %d, "
                       "information follows...\n"), stream->num);

                info(_("Version: %d\n"), inf->vi.version);
                info(_("Vendor: %s\n"), inf->vc.vendor);
                info(_("Channels: %d\n"), inf->vi.channels);
                info(_("Rate: %ld\n\n"), inf->vi.rate);

                if(inf->vi.bitrate_nominal >= 0)
                    info(_("Nominal bitrate: %f kb/s\n"), 
                            (double)inf->vi.bitrate_nominal / 1000.0);
162
                else
163
                    info(_("Nominal bitrate not set\n"));
164

165 166 167
                if(inf->vi.bitrate_upper >= 0)
                    info(_("Upper bitrate: %f kb/s\n"), 
                            (double)inf->vi.bitrate_upper / 1000.0);
168
                else
169
                    info(_("Upper bitrate not set\n"));
170

171 172 173
                if(inf->vi.bitrate_lower >= 0)
                    info(_("Lower bitrate: %f kb/s\n"), 
                            (double)inf->vi.bitrate_lower / 1000.0);
174
                else
175
                    info(_("Lower bitrate not set\n"));
176

177 178
                if(inf->vc.comments > 0)
                    info(_("User comments section follows...\n"));
179

180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320
                for(i=0; i < inf->vc.comments; i++) {
                    char *sep = strchr(inf->vc.user_comments[i], '=');
                    char *decoded;
                    int j;
                    int broken = 0;
                    unsigned char *val;
                    int bytes;
                    int remaining;

                    if(sep == NULL) {
                        warn(_("Warning: Comment %d in stream %d is invalidly "
                               "formatted, does not contain '=': \"%s\"\n"), 
                               i, stream->num, inf->vc.user_comments[i]);
                        continue;
                    }

                    for(j=0; j < sep-inf->vc.user_comments[i]; j++) {
                        if(inf->vc.user_comments[i][j] < 0x20 ||
                           inf->vc.user_comments[i][j] > 0x7D) {
                            warn(_("Warning: Invalid comment fieldname in "
                                   "comment %d (stream %d): \"%s\"\n"),
                                    i, stream->num, inf->vc.user_comments[i]);
                            broken = 1;
                            break;
                        }
                    }

                    if(broken)
                        continue;

                    val = inf->vc.user_comments[i];

                    j = sep-inf->vc.user_comments[i]+1;
                    while(j < inf->vc.comment_lengths[i])
                    {
                        remaining = inf->vc.comment_lengths[i] - j;
                        if((val[j] & 0x80) == 0)
                            bytes = 1;
                        else if((val[j] & 0x40) == 0)
                            bytes = 2;
                        else if((val[j] & 0x20) == 0)
                            bytes = 3;
                        else if((val[j] & 0x10) == 0)
                            bytes = 4;
                        else if((val[j] & 0x08) == 0)
                            bytes = 5;
                        else if((val[j] & 0x04) == 0)
                            bytes = 6;
                        else {
                            warn(_("Warning: Illegal UTF-8 sequence in comment "
                                   "%d (stream %d): length marker wrong\n"),
                                    i, stream->num);
                            broken = 1;
                            break;
                        }

                        if(bytes > remaining) {
                            warn(_("Warning: Illegal UTF-8 sequence in comment "
                                   "%d (stream %d): too few bytes\n"),
                                    i, stream->num);
                            broken = 1;
                            break;
                        }

                        switch(bytes) {
                            case 1:
                                /* No more checks needed */
                                break;
                            case 2:
                                if((val[j+1] & 0xC0) != 0x80)
                                    broken = 1;
                                if((val[j] & 0xFE) == 0xC0)
                                    broken = 1;
                                break;
                            case 3:
                                if(!((val[j] == 0xE0 && val[j+1] >= 0xA0 && 
                                        val[j+1] <= 0xBF && 
                                        (val[j+2] & 0xC0) == 0x80) ||
                                   (val[j] >= 0xE1 && val[j] <= 0xEC &&
                                        (val[j+1] & 0xC0) == 0x80 &&
                                        (val[j+2] & 0xC0) == 0x80) ||
                                   (val[j] == 0xED && val[j+1] >= 0x80 &&
                                        val[j+1] <= 0x9F &&
                                        (val[j+2] & 0xC0) == 0x80) ||
                                   (val[j] >= 0xEE && val[j] <= 0xEF &&
                                        (val[j+1] & 0xC0) == 0x80 &&
                                        (val[j+2] & 0xC0) == 0x80)))
                                    broken = 1;
                                if(val[j] == 0xE0 && (val[j+1] & 0xE0) == 0x80)
                                    broken = 1;
                                break;
                            case 4:
                                if(!((val[j] == 0xF0 && val[j+1] >= 0x90 &&
                                        val[j+1] <= 0xBF &&
                                        (val[j+2] & 0xC0) == 0x80 &&
                                        (val[j+3] & 0xC0) == 0x80) ||
                                     (val[j] >= 0xF1 && val[j] <= 0xF3 &&
                                        (val[j+1] & 0xC0) == 0x80 &&
                                        (val[j+2] & 0xC0) == 0x80 &&
                                        (val[j+3] & 0xC0) == 0x80) ||
                                     (val[j] == 0xF4 && val[j+1] >= 0x80 &&
                                        val[j+1] <= 0x8F &&
                                        (val[j+2] & 0xC0) == 0x80 &&
                                        (val[j+3] & 0xC0) == 0x80)))
                                    broken = 1;
                                if(val[j] == 0xF0 && (val[j+1] & 0xF0) == 0x80)
                                    broken = 1;
                                break;
                            /* 5 and 6 aren't actually allowed at this point*/  
                            case 5:
                                broken = 1;
                                break;
                            case 6:
                                broken = 1;
                                break;
                        }

                        if(broken) {
                            warn(_("Warning: Illegal UTF-8 sequence in comment "
                                   "%d (stream %d): invalid sequence\n"),
                                    i, stream->num);
                            broken = 1;
                            break;
                        }

                        j += bytes;
                    }

                    if(!broken) {
                        /* A hack around brokenness in the utf8 decoder */
                        if(strlen(sep+1) == 0)
                            decoded = sep+1;
                        else if(utf8_decode(sep+1, &decoded) < 0) {
                            warn(_("Warning: Failure in utf8 decoder. This "
                                   "should be impossible\n"));
                            continue;
                        }
                        *sep = 0;
                        info("\t%s=%s\n", inf->vc.user_comments[i], decoded);
                    }
                }
321 322 323 324 325
            }
        }
    }

    if(!header) {
326 327 328 329 330 331 332 333 334
        ogg_int64_t gp = ogg_page_granulepos(page);
        if(gp > 0) {
            if(gp < inf->lastgranulepos)
                warn(_("Warning: granulepos in stream %d decreases from " 
                        INT64FORMAT " to " INT64FORMAT "\n"), stream->num,
                        inf->lastgranulepos, gp);
            inf->lastgranulepos = gp;
        }
        inf->bytes += page->header_len + page->body_len;
335 336 337 338 339
    }
}

static void vorbis_end(stream_processor *stream) 
{
340
    misc_vorbis_info *inf = stream->data;
341 342 343
    long minutes, seconds;
    double bitrate, time;

344
    time = (double)inf->lastgranulepos / inf->vi.rate;
345 346
    minutes = (long)time / 60;
    seconds = (long)time - minutes*60;
347
    bitrate = inf->bytes*8 / time / 1000.0;
348

349 350 351 352 353
    info(_("Vorbis stream %d:\n"
           "\tTotal data length: %ld bytes\n"
           "\tPlayback length: %ldm:%02lds\n"
           "\tAverage bitrate: %f kbps\n"), 
            stream->num,inf->bytes, minutes, seconds, bitrate);
354

355 356
    vorbis_comment_clear(&inf->vc);
    vorbis_info_clear(&inf->vi);
357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377

    free(stream->data);
}

static void process_other(stream_processor *stream, ogg_page *page )
{
    ogg_packet packet;

    ogg_stream_pagein(&stream->os, page);

    while(ogg_stream_packetout(&stream->os, &packet) > 0) {
        /* Should we do anything here? Currently, we don't */
    }
}
        

static void free_stream_set(stream_set *set)
{
    int i;
    for(i=0; i < set->used; i++) {
        if(!set->streams[i].end) {
378
            warn(_("Warning: EOS not set on stream %d\n"), 
379
                    set->streams[i].num);
380
            set->streams[i].process_end(&set->streams[i]);
381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400
        }
        ogg_stream_clear(&set->streams[i].os);
    }

    free(set->streams);
    free(set);
}

static int streams_open(stream_set *set)
{
    int i;
    int res=0;
    for(i=0; i < set->used; i++) {
        if(!set->streams[i].end)
            res++;
    }

    return res;
}

401
static void other_start(stream_processor *stream, char *type)
402
{
403 404 405 406
    if(type)
        stream->type = type;
    else
        stream->type = "unknown";
407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488
    stream->process_page = process_other;
    stream->process_end = NULL;
}

static void vorbis_start(stream_processor *stream)
{
    misc_vorbis_info *info;

    stream->type = "vorbis";
    stream->process_page = vorbis_process;
    stream->process_end = vorbis_end;

    stream->data = calloc(1, sizeof(misc_vorbis_info));

    info = stream->data;

    vorbis_comment_init(&info->vc);
    vorbis_info_init(&info->vi);

}

static stream_processor *find_stream_processor(stream_set *set, ogg_page *page)
{
    ogg_uint32_t serial = ogg_page_serialno(page);
    int i, found = 0;
    int invalid = 0;
    stream_processor *stream;

    for(i=0; i < set->used; i++) {
        if(serial == set->streams[i].serial) {
            /* We have a match! */
            found = 1;
            stream = &(set->streams[i]);

            set->in_headers = 0;
            /* if we have detected EOS, then this can't occur here. */
            if(stream->end) {
                stream->isillegal = 1;
                return stream;
            }

            stream->isnew = 0;
            stream->start = ogg_page_bos(page);
            stream->end = ogg_page_eos(page);
            stream->serial = serial;
            return stream;
        }
    }

    /* If there are streams open, and we've reached the end of the
     * headers, then we can't be starting a new stream.
     * XXX: might this sometimes catch ok streams if EOS flag is missing,
     * but the stream is otherwise ok?
     */
    if(streams_open(set) && !set->in_headers)
        invalid = 1;

    set->in_headers = 1;

    if(set->allocated < set->used)
        stream = &set->streams[set->used];
    else {
        set->allocated += 5;
        set->streams = realloc(set->streams, sizeof(stream_processor)*
                set->allocated);
        stream = &set->streams[set->used];
    }
    set->used++;
    stream->num = set->used; /* We count from 1 */

    stream->isnew = 1;
    stream->isillegal = invalid;

    {
        int res;
        ogg_packet packet;

        /* We end up processing the header page twice, but that's ok. */
        ogg_stream_init(&stream->os, serial);
        ogg_stream_pagein(&stream->os, page);
        res = ogg_stream_packetout(&stream->os, &packet);
        if(res <= 0) {
489
            warn(_("Warning: Invalid header page, no packet found\n"));
490 491 492 493 494
            return NULL;
        }

        if(packet.bytes >= 7 && memcmp(packet.packet, "\001vorbis", 7)==0)
            vorbis_start(stream);
495 496 497 498
        else if(packet.bytes >= 8 && memcmp(packet.packet, "OggMIDI\0", 8)==0) 
            other_start(stream, "MIDI");
        else
            other_start(stream, NULL);
499 500 501

        res = ogg_stream_packetout(&stream->os, &packet);
        if(res > 0) {
502
            warn(_("Warning: Invalid header page in stream %d, "
503 504 505 506 507 508 509 510 511 512 513 514 515 516 517
                              "contains multiple packets\n"), stream->num);
        }

        /* re-init, ready for processing */
        ogg_stream_clear(&stream->os);
        ogg_stream_init(&stream->os, serial);
   }

   stream->start = ogg_page_bos(page);
   stream->end = ogg_page_eos(page);
   stream->serial = serial;

   return stream;
}

518 519
static int get_next_page(FILE *f, ogg_sync_state *sync, ogg_page *page, 
        ogg_int64_t *written)
520 521 522 523 524 525 526
{
    int ret;
    char *buffer;
    int bytes;

    while((ret = ogg_sync_pageout(sync, page)) <= 0) {
        if(ret < 0)
527 528
            warn(_("Warning: Hole in data found at approximate offset "
                   INT64FORMAT " bytes. Corrupted ogg.\n"), *written);
529 530 531

        buffer = ogg_sync_buffer(sync, CHUNK);
        bytes = fread(buffer, 1, CHUNK, f);
532 533
        if(bytes <= 0) {
            ogg_sync_wrote(sync, 0);
534
            return 0;
535 536 537
        }
        ogg_sync_wrote(sync, bytes);
        *written += bytes;
538 539 540 541 542 543 544 545 546 547
    }

    return 1;
}

static void process_file(char *filename) {
    FILE *file = fopen(filename, "rb");
    ogg_sync_state sync;
    ogg_page page;
    stream_set *processors = create_stream_set();
548 549
    int gotpage = 0;
    ogg_int64_t written = 0;
550 551

    if(!file) {
552
        error(_("Error opening input file \"%s\": %s\n"), filename,
553 554 555 556
                    strerror(errno));
        return;
    }

557
    printf(_("Processing file \"%s\"...\n\n"), filename);
558 559 560

    ogg_sync_init(&sync);

561 562
    while(get_next_page(file, &sync, &page, &written)) {
        gotpage = 1;
563 564 565
        stream_processor *p = find_stream_processor(processors, &page);

        if(!p) {
566
            error(_("Could not find a processor for stream, bailing\n"));
567 568 569 570
            return;
        }

        if(p->isillegal && !p->shownillegal) {
571 572
            warn(_("Warning: illegally placed page(s) for logical stream %d\n"
                   "This indicates a corrupt ogg file.\n"), p->num);
573 574 575 576 577
            p->shownillegal = 1;
            continue;
        }

        if(p->isnew) {
578
            info(_("New logical stream (#%d, serial: %08x): type %s\n"), 
579 580
                    p->num, p->serial, p->type);
            if(!p->start)
581
                warn(_("Warning: stream start flag not set on stream %d\n"),
582 583 584
                        p->num);
        }
        else if(p->start)
585
            warn(_("Warning: stream start flag found in mid-stream "
586 587
                      "on stream %d\n"), p->num);

588 589 590 591 592 593 594 595 596 597 598
        if(p->seqno++ != ogg_page_pageno(&page)) {
            if(!p->lostseq) 
                warn(_("Warning: sequence number gap in stream %d. Got page "
                       "%ld when expecting page %ld. Indicates missing data.\n"
                       ), p->num, ogg_page_pageno(&page), p->seqno - 1);
            p->seqno = ogg_page_pageno(&page);
            p->lostseq = 1;
        }
        else
            p->lostseq = 0;

599 600 601 602 603 604
        if(!p->isillegal) {
            p->process_page(p, &page);

            if(p->end) {
                if(p->process_end)
                    p->process_end(p);
605
                info(_("Logical stream %d ended\n"), p->num);
606 607 608 609 610
                p->isillegal = 1;
            }
        }
    }

611 612 613 614
    if(!gotpage)
        error(_("Error: No ogg data found in file \"%s\".\n"
                "Input probably not ogg.\n"), filename);

615 616 617 618 619 620 621
    free_stream_set(processors);

    ogg_sync_clear(&sync);

    fclose(file);
}

622 623 624 625 626 627 628 629 630 631 632 633 634
static void usage(void) {
    printf(_("ogginfo 1.0\n"
             "(c) 2002 Michael Smith <msmith@labyrinth.net.au>\n"
             "\n"
             "Usage: ogginfo [flags] files1.ogg [file2.ogg ... fileN.ogg]\n"
             "Flags supported:\n"
             "\t-h Show this help message\n"
             "\t-q Make less verbose. Once will remove detailed informative\n"
             "\t   messages, two will remove warnings\n"
             "\t-v Make more verbose. This may enable more detailed checks\n"
             "\t   for some stream types.\n\n"));
}

635
int main(int argc, char **argv) {
636
    int f, ret;
637 638 639 640 641 642 643

    setlocale(LC_ALL, "");
    bindtextdomain(PACKAGE, LOCALEDIR);
    textdomain(PACKAGE);

    if(argc < 2) {
        fprintf(stderr, 
644
                _("Usage: ogginfo [flags] file1.ogg [file2.ogg ... fileN.ogg]\n"
645 646
                  "\n"
                  "Ogginfo is a tool for printing information about ogg files\n"
647 648
                  "and for diagnosing problems with them.\n"
                  "Full help shown with \"ogginfo -h\".\n"));
649 650 651
        exit(1);
    }

652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671
    while((ret = getopt(argc, argv, "hvq")) >= 0) {
        switch(ret) {
            case 'h':
                usage();
                return 0;
            case 'v':
                verbose++;
                break;
            case 'q':
                verbose--;
                break;
        }
    }

    if(verbose < 1)
        printinfo = 0;
    if(verbose < 0) 
        printwarn = 0;

    if(optind >= argc) {
672 673
        fprintf(stderr, 
                _("No input files specified. \"ogginfo -h\" for help\n"));
674 675 676 677
        return 1;
    }

    for(f=optind; f < argc; f++) {
678 679 680 681 682 683
        process_file(argv[f]);
    }

    return 0;
}