Suffixarray *
readSuffixarray(void *space,
char *idxfilename,
CharSequence **seqs,
Uint len) {
FILE *fp;
char **suffixptr;
Uint nmemb = 0,
idvmemb = 0,
llvmemb = 0,
numofsuffixes,
*suftab = NULL,
*suflink = NULL;
childtab *chldtab = NULL;
unsigned char flags=0,
*lcpctab = NULL;
signed char *id = NULL;
PairUint *llvtab = NULL;
PairSint *idvtab = NULL;
MultiCharSeq *mseq;
Suffixarray *s;
mseq = concatCharSequences(space, seqs, len, (char)254, (char)127);
numofsuffixes = (mseq->totallength - mseq->numofsequences)+1;
suffixptr = getSuffixPtr(space, mseq, (char)254, (char)127);
fp = fopen(idxfilename, "r");
if (fp == NULL) {
fprintf(stderr,"Couldn't open file '%s'. Exit forced.\n", idxfilename);
exit(-1);
}
fread(&nmemb, sizeof(Uint), 1, fp);
suftab = ALLOCMEMORY(NULL, NULL, Uint, nmemb);
fread(suftab, sizeof(Uint), nmemb, fp);
fread(&flags, sizeof(char), 1, fp);
if (flags & LCP_TAB_STORED) {
fprintf(stderr, "reading lcpc/vtab\n");
lcpctab = ALLOCMEMORY(space, NULL, unsigned char, nmemb);
fread(lcpctab, sizeof(unsigned char), nmemb, fp);
fread(&llvmemb, sizeof(Uint), 1, fp);
llvtab = ALLOCMEMORY(space, NULL, PairUint, nmemb);
fread(llvtab, sizeof(PairUint), llvmemb, fp);
}
Suffixarray*
constructSufArr(void *space,
CharSequence **s,
Uint len,
FAlphabet* alphabet)
{
Uint i, numofsuffixes,
*sorted,
*inv_suftab;
char **suffixptr;
MultiCharSeq *mseq;
Suffixarray *arr;
mseq = concatCharSequences(space, s, len, (char)254, (char)127);
numofsuffixes = (mseq->totallength - mseq->numofsequences)+1;
//printf("md5: %s\n", MD5((unsigned char*)"f**k",4,NULL));
fprintf(stderr, "alphabet of size (%d): %s\n", mseq->mapsize, mseq->map);
//fprintf(stderr, "allocating space for %d suffixes (%d bytes)\n", numofsuffixes, 2*numofsuffixes*sizeof(Uint));
inv_suftab = ALLOCMEMORY(space, NULL, Uint , numofsuffixes);
arr = ALLOCMEMORY(space, NULL, Suffixarray, 1);
fprintf(stderr, "constructing suftab.\n");
suffixptr = getSuffixPtr(space, mseq, (char)254, (char)127);
//mseq->sequences[numofsuffixes-1]=0;
sorted = alurusort(space, mseq->sequences, &(numofsuffixes));
// mseq->sequences[numofsuffixes-1]=save;
/* sorted = quickSortMultikey (space, suffixptr, numofsuffixes,
cmpCharSequence, numofsuffixes-1, NULL);
*/
fprintf(stderr, "constructing inv_suftab.\n");
for (i=0; i < numofsuffixes; i++) {
inv_suftab[sorted[i]]=i;
}
arr->seq = mseq;
arr->numofsuffixes = numofsuffixes;
arr->suffixptr = suffixptr;
arr->suftab = sorted;
arr->inv_suftab = inv_suftab;
return arr;
}
Suffixarray*
constructSufArr(void *space,
CharSequence **s,
Uint len,
FAlphabet* alphabet)
{
Uint i, numofsuffixes,
*sorted,
*inv_suftab;
char **suffixptr;
MultiCharSeq *mseq;
Suffixarray *arr;
mseq = concatCharSequences(space, s, len, (char)254, (char)127);
numofsuffixes = (mseq->totallength - mseq->numofsequences)+1;
printf("allocating space for %d suffixes (%d bytes)\n", numofsuffixes, 2*numofsuffixes*sizeof(Uint));
inv_suftab = ALLOCMEMORY(space, NULL, Uint , numofsuffixes);
arr = ALLOCMEMORY(space, NULL, Suffixarray, 1);
printf("constructing suftab.\n");
suffixptr = getSuffixPtr(space, mseq, (char)254, (char)127);
sorted = quickSortMultikey (space, suffixptr, numofsuffixes,
cmpCharSequence, numofsuffixes-1, NULL);
printf("constructing inv_suftab.\n");
for (i=0; i < numofsuffixes; i++) {
inv_suftab[sorted[i]]=i;
}
arr->seq = mseq;
arr->numofsuffixes = numofsuffixes;
arr->suffixptr = suffixptr;
arr->suftab = sorted;
arr->inv_suftab = inv_suftab;
return arr;
}
Suffixarray *
readSuffixarray(void *space,
char *idxfilename,
CharSequence **seqs,
Uint len,
unsigned char silent) {
FILE *fp;
Uint nmemb = 0,
idvmemb = 0,
llvmemb = 0,
numofsuffixes,
*suftab = NULL,
idvi =0;
childtab *chldtab = NULL;
unsigned char flags=0,
*lcpctab = NULL;
unsigned char *mdfive=NULL,
*check=NULL;
PairUint *llvtab = NULL;
PairLSint *idvtab = NULL;
PairSint *idvutab = NULL;
MultiCharSeq *mseq;
Suffixarray *s;
#ifdef SUFLINK_MMAP
int fd;
signed char *id = NULL;
long curiopos, offset;
struct stat sb;
char *suflinkptr;
int pagediff_id;
int pagediff_sl;
#elif SUFLINK_DISKACC
int fd;
off_t off_sl;
off_t off_id;
#else
signed char *id = NULL;
Uint *suflink = NULL;
#endif
mseq = concatCharSequences(space, seqs, len, (char)126, (char)127);
numofsuffixes = mseq->totallength;
fp = fopen(idxfilename, "r");
if (fp == NULL) {
DBG("Couldn't open file '%s'. Exit forced.\n", idxfilename);
exit(-1);
}
fread(&nmemb, sizeof(Uint), 1, fp);
suftab = ALLOCMEMORY(NULL, NULL, Uint, nmemb);
fread(suftab, sizeof(Uint), nmemb, fp);
fread(&flags, sizeof(char), 1, fp);
if (flags & LCP_TAB_STORED) {
if (!silent) MSG("reading lcpc/vtab.\n");
lcpctab = ALLOCMEMORY(space, NULL, unsigned char, nmemb);
fread(lcpctab, sizeof(unsigned char), nmemb, fp);
fread(&llvmemb, sizeof(Uint), 1, fp);
llvtab = ALLOCMEMORY(space, NULL, PairUint, nmemb);
fread(llvtab, sizeof(PairUint), llvmemb, fp);
}
if (flags & CHLD_TAB_STORED) {
if(!silent) MSG("reading childtab.\n");
chldtab = ALLOCMEMORY(space, NULL, childtab, nmemb);
fread(chldtab, sizeof(childtab), nmemb, fp);
}
if ((flags & SUFLINK_TAB_STORED)) {
if(!silent) MSG("reading suflinks.\n");
#ifdef SUFLINK_MMAP
curiopos = ftell(fp);
fd = open(idxfilename, O_RDONLY);
if (fd == -1) {
perror("open");
exit(EXIT_FAILURE);
}
if (fstat(fd, &sb) == -1) {
perror("fstat");
exit(EXIT_FAILURE);
}
offset = curiopos & ~(sysconf(_SC_PAGE_SIZE) - 1);
if (curiopos >= sb.st_size) {
fprintf(stderr, "offset is past end of file\n");
exit(EXIT_FAILURE);
}
pagediff_sl = curiopos - offset;
suflinkptr = mmap(0, nmemb*sizeof(Uint) + pagediff_sl, PROT_READ, MAP_SHARED, fd, offset);
if (suflinkptr == MAP_FAILED) {
perror("mmap");
exit(EXIT_FAILURE);
}
#elif SUFLINK_DISKACC
sl_diskacc = 1;
off_sl = ftell(fp);
fd = open(idxfilename, O_RDONLY);
#else
suflink = ALLOCMEMORY(space, NULL, Uint, nmemb);
fread(suflink, sizeof(Uint), nmemb, fp);
#endif
#ifdef SUFLINK_MMAP
offset = (curiopos+(nmemb*sizeof(Uint))) & ~(sysconf(_SC_PAGE_SIZE) - 1);
if (curiopos >= sb.st_size) {
fprintf(stderr, "offset is past end of file\n");
exit(EXIT_FAILURE);
}
pagediff_id = (curiopos+(nmemb*sizeof(Uint))) - offset;
id = mmap(0, nmemb*sizeof(signed char) + pagediff_id, PROT_READ, MAP_SHARED, fd, offset);
if (id == MAP_FAILED) {
perror("mmap");
exit(EXIT_FAILURE);
}
fseek(fp, nmemb*(sizeof(Uint)+sizeof(signed char)), SEEK_CUR);
#elif SUFLINK_DISKACC
off_id = off_sl+(nmemb*sizeof(Uint));
fseek(fp, nmemb*(sizeof(Uint)+sizeof(signed char)), SEEK_CUR);
#else
id = ALLOCMEMORY(space, NULL, signed char, nmemb);
fread(id, sizeof(signed char), nmemb, fp);
#endif
fread(&idvmemb, sizeof(Uint), 1, fp);
idvtab = ALLOCMEMORY(space, NULL, PairLSint, idvmemb);
if ((flags & LINT_SUFLINKS)) {
if(!silent) MSG("reading lsint id.\n");
fread(idvtab, sizeof(PairLSint), idvmemb, fp);
} else {
idvutab = ALLOCMEMORY(space, NULL, PairSint, idvmemb);
if(!silent) MSG("reading uint id.\n");
fread(idvutab, sizeof(PairUint), idvmemb, fp);
for(idvi=0; idvi < idvmemb; idvi++) {
idvtab[idvi].a = idvutab[idvi].a;
idvtab[idvi].b = idvutab[idvi].b;
}
free(idvutab);
}
}
if ((flags & MD5_STORED)) {
mdfive = ALLOCMEMORY(space, NULL, unsigned char, 16);
fread(mdfive, sizeof(unsigned char), 16, fp);
}
Suffixarray*
constructSufArr(void *space,
CharSequence **s,
Uint len,
FAlphabet* alphabet,
unsigned char silent)
{
Uint i, numofsuffixes,
*sorted,
*inv_suftab;
MultiCharSeq *mseq;
Suffixarray *arr;
unsigned char *temp,
*mdfive=NULL;
mseq = concatCharSequences(space, s, len, (char)126, (char)127);
numofsuffixes = mseq->totallength;
mdfive = ALLOCMEMORY(space, NULL, char, 16);
temp = MD5R((unsigned char*)mseq->sequences, numofsuffixes, NULL);
memmove(mdfive, temp, 16);
if(!silent) NFO("alphabet of size (%d): %s\n", mseq->mapsize, mseq->map);
if(!silent) NFO("size of db sequence: %u\n", numofsuffixes);
inv_suftab = ALLOCMEMORY(space, NULL, Uint , numofsuffixes);
arr = ALLOCMEMORY(space, NULL, Suffixarray, 1);
if(!silent) MSG("constructing suftab.\n");
#ifdef SUF_MKQUICKSORT
sorted = quickSortMultikey (space, mseq->sequences, numofsuffixes,
cmpCharSequence, numofsuffixes-1, NULL);
#else
sorted = alurusort(space, mseq->sequences,
&(numofsuffixes));
#endif
if (!silent)NFO("constructing inv_suftab (%u).\n", numofsuffixes);
for (i=0; i < numofsuffixes; i++) {
if (sorted[i] > numofsuffixes) fprintf(stderr, "construction error? %u: %u\n",i, sorted[i]);
inv_suftab[sorted[i]]=i;
}
if (!silent)MSG("inv_suftab constructed.\n");
arr->seq = mseq;
arr->numofsuffixes = numofsuffixes;
arr->suftab = sorted;
arr->inv_suftab = inv_suftab;
arr->mdfive = mdfive;
arr->lcpctab = NULL;
arr->llvtab = NULL;
arr->id = NULL;
arr->idvtab = NULL;
arr->chldtab = NULL;
arr->bcktab = NULL;
arr->suflink = NULL;
arr->suflink_l = NULL;
arr->suflink_r = NULL;
arr->llint = 1;
return arr;
}
