/* *******************************************************************
procedure to be called by external program before calling ds_ssort()
using this procedure external programs can choose
the parameters Anchor_dist and Blind_sort_ratio.
The procedure returns 0 if something goes wrong, otherwise
it returns the overshhot, that is the amount of extra space
required at the end of the array contanining the text
******************************************************************** */
int init_ds_ssort(int adist, int bs_ratio)
{
set_global_variables();
Anchor_dist = adist;
Blind_sort_ratio=bs_ratio;
Shallow_limit = Anchor_dist + 50;
if(check_global_variables())
return 0;
return compute_overshoot();
}
//int init_ds_ssort(int adist, int bs_ratio)
Int64 init_ds_ssort(Int64 adist, Int64 bs_ratio)
{
set_global_variables();
Anchor_dist = adist;
Blind_sort_ratio = bs_ratio;
Shallow_limit = Anchor_dist + 50;
//Shallow_limit = Anchor_dist + 1000000; // Shallow_limit = the longest of the lcp-s!
if (check_global_variables()) {
return 0;
}
return compute_overshoot();
}
int main(int argc, char *argv[])
{
void write_sa(char *filename, int *p, int n);
void write_lcp(char *filename, UChar *x,int *p, int n);
void write_bwt(char *filename, UChar *x,int *p, int n);
void check_sa_ordering(UChar *x,int *p, int n, int);
void print_sa_onscreen(UChar *x,int *p, int n, int);
int c, *p, n;
int print_sa, check_sa, num_opt,overshoot;
UChar *x;
clock_t end,start, end_real, start_real;
struct tms r;
double tot_time = 0.0;
double tot_time_real = 0.0;
extern char *optarg;
extern int optind, opterr, optopt;
char *fnam, *sa_filename;
char *lcp_filename,*bwt_filename; // names for (optional) lcp and bwt files
FILE *f;
/* ------------ set default values ------------- */
set_global_variables();
print_sa=check_sa=0;
sa_filename = NULL;
lcp_filename = NULL;
bwt_filename = NULL;
/* ------------- read options from command line ----------- */
num_opt = opterr = 0;
while ((c=getopt(argc, argv, "b:d:l:p:r:w:cvux:f:T:W:B:")) != -1) {
switch (c)
{
case 'b':
bwt_filename = optarg; break;
case 'c':
check_sa++; break;
case 'd':
Anchor_dist = atoi(optarg); break;
case 'l':
Shallow_limit = atoi(optarg); break;
case 'p':
print_sa = atoi(optarg); break;
case 'x':
_ds_Word_size = atoi(optarg); break;
case 'v':
_ds_Verbose++; break;
case 'w':
sa_filename = optarg; break;
case 'f':
Max_pseudo_anchor_offset = atoi(optarg); break;
case 'r':
B2g_ratio = atoi(optarg); break;
case 'u':
Update_anchor_ranks = 1; break;
case 'T':
Mk_qs_thresh = atoi(optarg); break;
case 'W':
lcp_filename = optarg; break;
case 'B':
Blind_sort_ratio = atoi(optarg); break;
case '?':
fprintf(stderr,"Unknown option: %c -main-\n", optopt);
exit(1);
}
num_opt++;
}
if(optind<argc)
fnam=argv[optind];
else {
fprintf(stderr, "Usage:\n\t%s [-b bwtfile][-cuv][-d dist]",argv[0]);
fprintf(stderr, "[-l len][-p num][-f maxoff][-r ratio]\n");
fprintf(stderr,
"\t [-T thresh][-w safile][-W lcpfile][-x wsize][-B ratio]");
fprintf(stderr, " file\n\n");
fprintf(stderr,"\t-b bwtfile write bwt to bwtfile\n");
fprintf(stderr,
"\t-B ratio blind_sort ratio [def. %d]\n",Blind_sort_ratio);
fprintf(stderr,"\t-c check the sa (could be very slow)\n");
fprintf(stderr,"\t-d dist anchor distance [def. %d]\n",Anchor_dist);
fprintf(stderr,"\t-f maxoff Maximum offset for forward ");
fprintf(stderr,"pseudo-anchors [def. %d]\n",Max_pseudo_anchor_offset);
fprintf(stderr,
"\t-l len shallow sort limit [def. %d]\n",Shallow_limit);
fprintf(stderr,
"\t-r ratio bucket to group max ratio [def. %d]\n",B2g_ratio);
fprintf(stderr,"\t-p num print num char of each suffix [def. 0]\n");
fprintf(stderr,
"\t-T thresh Threshold for mk-qs [def. %d]\n", Mk_qs_thresh);
fprintf(stderr,"\t-u updates anchor ranks in get_rank()\n");
fprintf(stderr,"\t-v produces a verbose output\n");
fprintf(stderr,"\t-w safile write sa to safile\n");
fprintf(stderr,
"\t-W lcpfile check sa and write lcp to lcpfile (very slow)\n");
fprintf(stderr,
"\t-x wsize word size in mkqs (default %d)\n\n",_ds_Word_size);
return 0;
}
if(_ds_Verbose) {
fprintf(stderr,"Command line: ");
for(c=0;c<argc;c++)
fprintf(stderr,"%s ",argv[c]);
fprintf(stderr,"\n");
}
/* -------- check parameters ------------- */
if(check_global_variables()) {
exit(1);
}
/* ---------- open file and read text ----------- */
if (! (f=fopen(fnam, "rb"))) {
perror(fnam);
return 1;
}
if (fseek(f, 0L, SEEK_END)) {
perror(fnam);
return 1;
}
n=ftell(f);
if (n==0) {
fprintf(stderr, "%s: file empty\n", fnam);
return 0;
}
// ------ allocate memory for text and sa -------
overshoot = compute_overshoot();
p=malloc((n)*sizeof *p); // sa
x=malloc((n+overshoot)*sizeof *x); // text
if (! p || ! x) {
fprintf(stderr, "malloc failed\n");
return 1;
}
// ------------ read input text ---------------
rewind(f);
c=fread(x, (size_t) 1, (size_t) n, f);
// lseek(fileno(f),0,SEEK_SET);
// c=read(fileno(f), x, (size_t) n);
if(c!=n) {
fprintf(stderr,"Error in read() (%d vs %d) (main)\n",c,n);
perror(fnam);
return 1;
}
fclose(f);
/* --------- start measuring time ------------- */
if(_ds_Verbose)
fprintf(stderr,"Starting sa construction ... \n");
start_real = times(&r);
start = (r.tms_utime+r.tms_stime); /* user + system */
ds_ssort(x, p, n);
end_real = times(&r);
end = (r.tms_utime+r.tms_stime); /* user + system */
// tot_time = ((double) (end-start))/CLK_TCK;
//tot_time_real = ((double) (end_real-start_real))/CLK_TCK;
printf("Elapsed time: %.2f seconds (user+sys). Total real time: %.2f.\n",
tot_time, tot_time_real);
// --------------- write bwt to a file
if(bwt_filename!=NULL)
write_bwt(bwt_filename,x,p,n);
// --------------- write sa to a file
if(sa_filename!=NULL)
write_sa(sa_filename,p,n);
// --------------- write lcp to a file
if(lcp_filename!=NULL)
write_lcp(lcp_filename,x,p,n);
// ------------ check sa --------
if(check_sa)
check_sa_ordering(x,p,n,check_sa);
// ----- display sa -------
if(print_sa)
print_sa_onscreen(x,p,n,print_sa);
// deallocate and exit
free(x); free(p);
return 0;
}