static YAP_Bool
mpi_gc(void) {
//write_msg(__FUNCTION__,__FILE__,__LINE__,"MPI_gc>: requests=%d\n",requests->n_entries);
CONT_TIMER();
init_hash_traversal(requests);
gc(requests);
init_hash_traversal(broadcasts);
gc(broadcasts);
//write_msg(__FUNCTION__,__FILE__,__LINE__,"MPI_gc<: requests=%d\n",requests->n_entries);
PAUSE_TIMER();
return true;
}
int main(int argc, char **argv ) {
long paired=0;
//printf("%d",sizeof(struct index_entry));
if (argc!=6) {
fprintf(stderr,"Usage: fastqinterleaved2pair.c fastq fastq1 fastq2\n");
//fprintf(stderr,"%d",argc);
exit(1);
}
FILE *fd=open_fastq(argv[1]);
// ************************************************************
off_t cur_offset=1;
unsigned long cline=1;
hashtable index=new_hashtable(HASHSIZE);
index_mem+=sizeof(hashtable);
index_file(argv[1],index,0,-1);
printf("\n");
// print some info
printf("Reads indexed: %ld\n",index->n_entries);
printf("Memory used in indexing: %ld MB\n",index_mem/1024/1024);
//
char *p1=argv[3];
char *p2=argv[4];
char *p3=argv[5];
fd1=open_fastq(argv[1]);
fd2=open_fastq(argv[2]);
FILE *fdw1=fopen(p1,"w");
FILE *fdw2=fopen(p2,"w");
FILE *fdw3=fopen(p3,"w");
unsigned long up2=0;
if ( fdw1==NULL || fdw2==NULL || fdw3==NULL ) {
fprintf(stderr,"Unable to create output files\n");
exit(1);
}
// read the entry using another fd
cline=1;
while(!feof(fd2)) {
long start_pos=ftell(fd2);
char *hdr=READ_LINE(fd2);
if ( hdr==NULL) break;
if ( hdr[0]!='@' ) {
fprintf(stderr,"line %ul: error in header %s",cline,hdr);
return 1;
}
// discard @ (one less byte)
hdr=&hdr[1];
int len=strlen(hdr);
len--;
hdr[len-1]='\0'; //
// lookup hdr in index
INDEX_ENTRY* e=lookup_header(index,hdr);
if (e==NULL) {
++up2;
copy_read(start_pos,fd2,fdw3);
} else {
long key=hashit(hdr);
// pair found
++paired;
copy_read(start_pos,fd2,fdw2);
copy_read(e->entry_start,fd1,fdw1);
// remove entry from index
if (delete(index,key,e)!=e) {
fprintf(stderr,"Unable to delete entry from index\n");
exit(1);
}
free_indexentry(e);
}
PRINT_READS_PROCESSED(cline/4);
cline+=4;
}
printf("\n");
printf("Recording %ld unpaired reads from %s\n",index->n_entries,argv[1]);fflush(stdout);
fclose(fd1);
// record the unpaired from argv[1]
fd1=open_fastq(argv[1]);
#ifndef SEQDISKACCESS
init_hash_traversal(index);
INDEX_ENTRY* e;
cline=1;
while((e=(INDEX_ENTRY*)next_hash_object(index))!=NULL) {
copy_read(e->entry_start,fd1,fdw3);
PRINT_READS_PROCESSED(cline);
++cline;
}
//
#else
//sequential disk access
//
cline=1;
unsigned long remaining=index->n_entries;
while(!feof(fd1) && remaining ) {
//long start_pos=ftell(fd2);
char *hdr=READ_LINE(fd1);
if ( hdr==NULL) break;
if ( hdr[0]!='@' ) {
fprintf(stderr,"line %ld %s: error in header %s",cline,argv[1],hdr);
return 1;
}
// discard @ (one less byte)
hdr=&hdr[1];
int len=strlen(hdr);
len--;
hdr[len-1]='\0'; //
// lookup hdr in index
INDEX_ENTRY* e=lookup_header(index,hdr);
if (e!=NULL) {
copy_read(e->entry_start,fd1,fdw3);
remaining--;
} else {
READ_LINE(fd1);//seq
READ_LINE(fd1);//qual
READ_LINE(fd1);//qual
}
PRINT_READS_PROCESSED(cline/4);
cline+=4;
}
fclose(fd1);
#endif
printf("\n");
printf("Unpaired from %s: %ld\n",argv[1],index->n_entries);
printf("Unpaired from %s: %ld\n",argv[2],up2);
printf("Paired: %ld\n",paired);
/*fseek(fd2,start_pos,SEEK_SET);
printf("%s",READ_LINE(fd2));
printf("%s",READ_LINE(fd2));
printf("%s",READ_LINE(fd2));
printf("%s",READ_LINE(fd2));
*/
fclose(fdw1);
fclose(fdw2);
fclose(fdw3);
if ( paired == 0 ) {
fprintf(stderr,"!!!WARNING!!! 0 paired reads! are the headers ok?\n");
exit(1);
}
exit(0);
}
