/* libSoX Macintosh HCOM format. * These are really FSSD type files with Huffman compression, * in MacBinary format. * TODO: make the MacBinary format optional (so that .data files * are also acceptable). (How to do this on output?) * * September 25, 1991 * Copyright 1991 Guido van Rossum And Sundry Contributors * This source code is freely redistributable and may be used for * any purpose. This copyright notice must be maintained. * Guido van Rossum And Sundry Contributors are not responsible for * the consequences of using this software. * * April 28, 1998 - Chris Bagwell (cbagwell@sprynet.com) * * Rearranged some functions so that they are declared before they are * used, clearing up some compiler warnings. Because these functions * passed floats, it helped some dumb compilers pass stuff on the * stack correctly. * */ #include "sox_i.h" #include #include #include #include /* FIXME: eliminate these 2 functions */ static void put32_be(unsigned char **p, int32_t val) { *(*p)++ = (val >> 24) & 0xff; *(*p)++ = (val >> 16) & 0xff; *(*p)++ = (val >> 8) & 0xff; *(*p)++ = val & 0xff; } static void put16_be(unsigned char **p, int val) { *(*p)++ = (val >> 8) & 0xff; *(*p)++ = val & 0xff; } /* Dictionary entry for Huffman (de)compression */ typedef struct { long frequ; short dict_leftson; short dict_rightson; } dictent; typedef struct { /* Static data from the header */ dictent *dictionary; int32_t checksum; int deltacompression; /* Engine state */ long huffcount; long cksum; int dictentry; int nrbits; uint32_t current; short sample; /* Dictionary */ dictent *de; int32_t new_checksum; int nbits; int32_t curword; /* Private data used by writer */ unsigned char *data; /* Buffer allocated with lsx_malloc */ size_t size; /* Size of allocated buffer */ size_t pos; /* Where next byte goes */ } priv_t; static int dictvalid(int n, int size, int left, int right) { if (n > 0 && left < 0) return 1; return (unsigned)left < size && (unsigned)right < size; } static int startread(sox_format_t * ft) { priv_t *p = (priv_t *) ft->priv; int i; char buf[5]; uint32_t datasize, rsrcsize; uint32_t huffcount, checksum, compresstype, divisor; unsigned short dictsize; int rc; /* Skip first 65 bytes of header */ rc = lsx_skipbytes(ft, (size_t) 65); if (rc) return rc; /* Check the file type (bytes 65-68) */ if (lsx_reads(ft, buf, (size_t)4) == SOX_EOF || strncmp(buf, "FSSD", (size_t)4) != 0) { lsx_fail_errno(ft,SOX_EHDR,"Mac header type is not FSSD"); return (SOX_EOF); } /* Skip to byte 83 */ rc = lsx_skipbytes(ft, (size_t) 83-69); if (rc) return rc; /* Get essential numbers from the header */ lsx_readdw(ft, &datasize); /* bytes 83-86 */ lsx_readdw(ft, &rsrcsize); /* bytes 87-90 */ /* Skip the rest of the header (total 128 bytes) */ rc = lsx_skipbytes(ft, (size_t) 128-91); if (rc != 0) return rc; /* The data fork must contain a "HCOM" header */ if (lsx_reads(ft, buf, (size_t)4) == SOX_EOF || strncmp(buf, "HCOM", (size_t)4) != 0) { lsx_fail_errno(ft,SOX_EHDR,"Mac data fork is not HCOM"); return (SOX_EOF); } /* Then follow various parameters */ lsx_readdw(ft, &huffcount); lsx_readdw(ft, &checksum); lsx_readdw(ft, &compresstype); if (compresstype > 1) { lsx_fail_errno(ft,SOX_EHDR,"Bad compression type in HCOM header"); return (SOX_EOF); } lsx_readdw(ft, &divisor); if (divisor == 0 || divisor > 4) { lsx_fail_errno(ft,SOX_EHDR,"Bad sampling rate divisor in HCOM header"); return (SOX_EOF); } lsx_readw(ft, &dictsize); /* Translate to sox parameters */ ft->encoding.encoding = SOX_ENCODING_HCOM; ft->encoding.bits_per_sample = 8; ft->signal.rate = 22050 / divisor; ft->signal.channels = 1; ft->signal.length = huffcount; /* Allocate memory for the dictionary */ p->dictionary = lsx_malloc(511 * sizeof(dictent)); /* Read dictionary */ for(i = 0; i < dictsize; i++) { lsx_readsw(ft, &(p->dictionary[i].dict_leftson)); lsx_readsw(ft, &(p->dictionary[i].dict_rightson)); lsx_debug("%d %d", p->dictionary[i].dict_leftson, p->dictionary[i].dict_rightson); if (!dictvalid(i, dictsize, p->dictionary[i].dict_leftson, p->dictionary[i].dict_rightson)) { lsx_fail_errno(ft, SOX_EHDR, "Invalid dictionary"); return SOX_EOF; } } rc = lsx_skipbytes(ft, (size_t) 1); /* skip pad byte */ if (rc) return rc; /* Initialized the decompression engine */ p->checksum = checksum; p->deltacompression = compresstype; if (!p->deltacompression) lsx_debug("HCOM data using value compression"); p->huffcount = huffcount; p->cksum = 0; p->dictentry = 0; p->nrbits = -1; /* Special case to get first byte */ return (SOX_SUCCESS); } static size_t read_samples(sox_format_t * ft, sox_sample_t *buf, size_t len) { register priv_t *p = (priv_t *) ft->priv; int done = 0; unsigned char sample_rate; if (p->nrbits < 0) { /* The first byte is special */ if (p->huffcount == 0) return 0; /* Don't know if this can happen... */ if (lsx_readb(ft, &sample_rate) == SOX_EOF) { return (0); } p->sample = sample_rate; *buf++ = SOX_UNSIGNED_8BIT_TO_SAMPLE(p->sample,); p->huffcount--; p->nrbits = 0; done++; len--; if (len == 0) return done; } while (p->huffcount > 0) { if(p->nrbits == 0) { lsx_readdw(ft, &(p->current)); if (lsx_eof(ft)) { lsx_fail_errno(ft,SOX_EOF,"unexpected EOF in HCOM data"); return (0); } p->cksum += p->current; p->nrbits = 32; } if(p->current & 0x80000000) { p->dictentry = p->dictionary[p->dictentry].dict_rightson; } else { p->dictentry = p->dictionary[p->dictentry].dict_leftson; } p->current = p->current << 1; p->nrbits--; if(p->dictionary[p->dictentry].dict_leftson < 0) { short datum; datum = p->dictionary[p->dictentry].dict_rightson; if (!p->deltacompression) p->sample = 0; p->sample = (p->sample + datum) & 0xff; p->huffcount--; *buf++ = SOX_UNSIGNED_8BIT_TO_SAMPLE(p->sample,); p->dictentry = 0; done++; len--; if (len == 0) break; } } return done; } static int stopread(sox_format_t * ft) { register priv_t *p = (priv_t *) ft->priv; if (p->huffcount != 0) { lsx_fail_errno(ft,SOX_EFMT,"not all HCOM data read"); return (SOX_EOF); } if(p->cksum != p->checksum) { lsx_fail_errno(ft,SOX_EFMT,"checksum error in HCOM data"); return (SOX_EOF); } free(p->dictionary); p->dictionary = NULL; return (SOX_SUCCESS); } #define BUFINCR (10*BUFSIZ) static int startwrite(sox_format_t * ft) { priv_t * p = (priv_t *) ft->priv; p->size = BUFINCR; p->pos = 0; p->data = lsx_malloc(p->size); return SOX_SUCCESS; } static size_t write_samples(sox_format_t * ft, const sox_sample_t *buf, size_t len) { priv_t *p = (priv_t *) ft->priv; sox_sample_t datum; size_t i; if (len == 0) return 0; if (p->pos == INT32_MAX) return SOX_EOF; if (p->pos + len > INT32_MAX) { lsx_warn("maximum file size exceeded"); len = INT32_MAX - p->pos; } if (p->pos + len > p->size) { p->size = ((p->pos + len) / BUFINCR + 1) * BUFINCR; p->data = lsx_realloc(p->data, p->size); } for (i = 0; i < len; i++) { SOX_SAMPLE_LOCALS; datum = *buf++; p->data[p->pos++] = SOX_SAMPLE_TO_UNSIGNED_8BIT(datum, ft->clips); } return len; } static void makecodes(int e, int c, int s, int b, dictent newdict[511], long codes[256], long codesize[256]) { assert(b); /* Prevent stack overflow */ if (newdict[e].dict_leftson < 0) { codes[newdict[e].dict_rightson] = c; codesize[newdict[e].dict_rightson] = s; } else { makecodes(newdict[e].dict_leftson, c, s + 1, b << 1, newdict, codes, codesize); makecodes(newdict[e].dict_rightson, c + b, s + 1, b << 1, newdict, codes, codesize); } } static void putcode(sox_format_t * ft, long codes[256], long codesize[256], unsigned c, unsigned char **df) { priv_t *p = (priv_t *) ft->priv; long code, size; int i; code = codes[c]; size = codesize[c]; for(i = 0; i < size; i++) { p->curword <<= 1; if (code & 1) p->curword += 1; p->nbits++; if (p->nbits == 32) { put32_be(df, p->curword); p->new_checksum += p->curword; p->nbits = 0; p->curword = 0; } code >>= 1; } } static void compress(sox_format_t * ft, unsigned char **df, int32_t *dl) { priv_t *p = (priv_t *) ft->priv; int samplerate; unsigned char *datafork = *df; unsigned char *ddf, *dfp; short dictsize; int frequtable[256]; long codes[256], codesize[256]; dictent newdict[511]; int i, sample, j, k, d, l, frequcount; int64_t csize; sample = *datafork; memset(frequtable, 0, sizeof(frequtable)); memset(codes, 0, sizeof(codes)); memset(codesize, 0, sizeof(codesize)); memset(newdict, 0, sizeof(newdict)); for (i = 1; i < *dl; i++) { d = (datafork[i] - (sample & 0xff)) & 0xff; /* creates absolute entries LMS */ sample = datafork[i]; datafork[i] = d; assert(d >= 0 && d <= 255); /* check our table is accessed correctly */ frequtable[d]++; } p->de = newdict; for (i = 0; i < 256; i++) if (frequtable[i] != 0) { p->de->frequ = -frequtable[i]; p->de->dict_leftson = -1; p->de->dict_rightson = i; p->de++; } frequcount = p->de - newdict; for (i = 0; i < frequcount; i++) { for (j = i + 1; j < frequcount; j++) { if (newdict[i].frequ > newdict[j].frequ) { k = newdict[i].frequ; newdict[i].frequ = newdict[j].frequ; newdict[j].frequ = k; k = newdict[i].dict_leftson; newdict[i].dict_leftson = newdict[j].dict_leftson; newdict[j].dict_leftson = k; k = newdict[i].dict_rightson; newdict[i].dict_rightson = newdict[j].dict_rightson; newdict[j].dict_rightson = k; } } } while (frequcount > 1) { j = frequcount - 1; p->de->frequ = newdict[j - 1].frequ; p->de->dict_leftson = newdict[j - 1].dict_leftson; p->de->dict_rightson = newdict[j - 1].dict_rightson; l = newdict[j - 1].frequ + newdict[j].frequ; for (i = j - 2; i >= 0 && l < newdict[i].frequ; i--) newdict[i + 1] = newdict[i]; i = i + 1; newdict[i].frequ = l; newdict[i].dict_leftson = j; newdict[i].dict_rightson = p->de - newdict; p->de++; frequcount--; } dictsize = p->de - newdict; makecodes(0, 0, 0, 1, newdict, codes, codesize); csize = 0; for (i = 0; i < 256; i++) csize += frequtable[i] * codesize[i]; l = (((csize + 31) >> 5) << 2) + 24 + dictsize * 4; lsx_debug(" Original size: %6d bytes", *dl); lsx_debug("Compressed size: %6d bytes", l); datafork = lsx_malloc((size_t)l); ddf = datafork + 22; for(i = 0; i < dictsize; i++) { put16_be(&ddf, newdict[i].dict_leftson); put16_be(&ddf, newdict[i].dict_rightson); } *ddf++ = 0; *ddf++ = *(*df)++; p->new_checksum = 0; p->nbits = 0; p->curword = 0; for (i = 1; i < *dl; i++) putcode(ft, codes, codesize, *(*df)++, &ddf); if (p->nbits != 0) { codes[0] = 0; codesize[0] = 32 - p->nbits; putcode(ft, codes, codesize, 0, &ddf); } memcpy(datafork, "HCOM", (size_t)4); dfp = datafork + 4; put32_be(&dfp, *dl); put32_be(&dfp, p->new_checksum); put32_be(&dfp, 1); samplerate = 22050 / ft->signal.rate + .5; put32_be(&dfp, samplerate); put16_be(&dfp, dictsize); *df = datafork; /* reassign passed pointer to new datafork */ *dl = l; /* and its compressed length */ } /* End of hcom utility routines */ static int stopwrite(sox_format_t * ft) { priv_t *p = (priv_t *) ft->priv; unsigned char *compressed_data = p->data; int32_t compressed_len = p->pos; int rc = SOX_SUCCESS; /* Compress it all at once */ if (compressed_len) { compress(ft, &compressed_data, &compressed_len); free(p->data); } /* Write the header */ lsx_writebuf(ft, "\000\001A", (size_t) 3); /* Dummy file name "A" */ lsx_padbytes(ft, (size_t) 65-3); lsx_writes(ft, "FSSD"); lsx_padbytes(ft, (size_t) 83-69); lsx_writedw(ft, (unsigned) compressed_len); /* compressed_data size */ lsx_writedw(ft, 0); /* rsrc size */ lsx_padbytes(ft, (size_t) 128 - 91); if (lsx_error(ft)) { lsx_fail_errno(ft, errno, "write error in HCOM header"); rc = SOX_EOF; } else if (lsx_writebuf(ft, compressed_data, compressed_len) != compressed_len) { /* Write the compressed_data fork */ lsx_fail_errno(ft, errno, "can't write compressed HCOM data"); rc = SOX_EOF; } free(compressed_data); if (rc == SOX_SUCCESS) /* Pad the compressed_data fork to a multiple of 128 bytes */ lsx_padbytes(ft, 128u - (compressed_len % 128)); return rc; } LSX_FORMAT_HANDLER(hcom) { static char const * const names[] = {"hcom", NULL}; static sox_rate_t const write_rates[] = {22050,22050/2,22050/3,22050/4.,0}; static unsigned const write_encodings[] = { SOX_ENCODING_HCOM, 8, 0, 0}; static sox_format_handler_t handler = {SOX_LIB_VERSION_CODE, "Mac FSSD files with Huffman compression", names, SOX_FILE_BIG_END|SOX_FILE_MONO, startread, read_samples, stopread, startwrite, write_samples, stopwrite, NULL, write_encodings, write_rates, sizeof(priv_t) }; return &handler; }