int main_reheader(int argc, char *argv[])
{
	bam_header_t *h;
	BGZF *in;
	if (argc != 3) {
		fprintf(stderr, "Usage: samtools reheader <in.header.sam> <in.bam>\n");
		return 1;
	}
	{ // read the header
		tamFile fph = sam_open(argv[1]);
		if (fph == 0) {
			fprintf(stderr, "[%s] fail to read the header from %s.\n", __func__, argv[1]);
			return 1;
		}
		h = sam_header_read(fph);
		sam_close(fph);
	}
	in = strcmp(argv[2], "-")? bam_open(argv[2], "r") : bam_dopen(fileno(stdin), "r");
	if (in == 0) {
		fprintf(stderr, "[%s] fail to open file %s.\n", __func__, argv[2]);
		return 1;
	}
	bam_reheader(in, h, fileno(stdout));
	bgzf_close(in);
	return 0;
}
Exemple #2
0
samfile_t* samopen_in(quip_reader_t reader, void* reader_data, bool binary, void* aux)
{
    samfile_t *fp;
    fp = (samfile_t*)calloc(1, sizeof(samfile_t));

    fp->type |= TYPE_READ;
    if (binary) { // binary
        fp->type |= TYPE_BAM;
        fp->x.bam = bam_open_in(reader, reader_data);
        if (fp->x.bam == 0) goto open_err_ret;
        fp->header = bam_header_read(fp->x.bam);
    } else { // text
        fp->x.tamr = sam_open_in(reader, reader_data);
        if (fp->x.tamr == 0) goto open_err_ret;
        fp->header = sam_header_read(fp->x.tamr);
        if (fp->header->n_targets == 0) { // no @SQ fields
            if (aux) { // check if aux is present
                bam_header_t *textheader = fp->header;
                fp->header = sam_header_read2((const char*)aux);
                if (fp->header == 0) goto open_err_ret;
                append_header_text(fp->header, textheader->text, textheader->l_text);
                bam_header_destroy(textheader);
            }
            if (fp->header->n_targets == 0 && bam_verbose >= 1)
                fprintf(stderr, "[samopen] no @SQ lines in the header.\n");
        } else if (bam_verbose >= 2) fprintf(stderr, "[samopen] SAM header is present: %d sequences.\n", fp->header->n_targets);
    }

    return fp;

open_err_ret:
    free(fp);
    return 0;
}
Exemple #3
0
/*static*/ bam_header_t * SAM::update_header_from_list(bam_header_t *header, names_list_t & list) {
	Temporary_File samfile;
	samfile.close_file();
	samfile_t * sf = samopen(samfile.get_filename().c_str(),"wh",header);
	samclose(sf);

	Temporary_File tempfile;
	ofstream &output = tempfile.get_stream();

	ifstream input(samfile.get_filename().c_str());
	string temp;
	while (not input.eof()) {
		getline(input,temp);
		if ((temp.size() >= 3) and (temp[0] != '@' or temp[1] != 'S' or temp[2] != 'Q'))
			output << temp << '\n';
	}

	for (names_list_t::iterator iter = list.begin(); iter != list.end(); iter++)
		output << "@SQ\tSN:" << iter->first << "\tLN:" << iter->second << '\n';
	tempfile.close_file();

	tamFile fp = sam_open(tempfile.get_filename().c_str());

	bam_header_t * newheader = sam_header_read(fp);
	sam_close(fp);

	return newheader;
}
Exemple #4
0
int main_cat(int argc, char *argv[])
{
    bam_header_t *h = 0;
    char *outfn = 0;
    int c, ret;
    while ((c = getopt(argc, argv, "h:o:")) >= 0) {
        switch (c) {
            case 'h': {
                tamFile fph = sam_open(optarg);
                if (fph == 0) {
                    fprintf(stderr, "[%s] ERROR: fail to read the header from '%s'.\n", __func__, argv[1]);
                    return 1;
                }
                h = sam_header_read(fph);
                sam_close(fph);
                break;
            }
            case 'o': outfn = strdup(optarg); break;
        }
    }
    if (argc - optind < 2) {
        fprintf(stderr, "Usage: samtools cat [-h header.sam] [-o out.bam] <in1.bam> <in2.bam> [...]\n");
        return 1;
    }
    ret = bam_cat(argc - optind, argv + optind, h, outfn? outfn : "-");
    free(outfn);
    return ret;
}
Exemple #5
0
void convert_sam_to_bam(char* sam_input, char* bam_input) {
    bam1_t* bam_p = bam_init1();

    LOG_DEBUG("CONVERT-START: sam to bam\n");

    //open SAM file for read
    if (time_flag) {
        start_timer(t1_convert);
    }
    tamFile sam_fd = sam_open(sam_input);

    //open BAM file for write
    bam_file_t* bam_file_p =  bam_fopen_mode(bam_input, NULL, "w");

    //read header from SAM file
    bam_header_t* bam_header_p = sam_header_read(sam_fd);

    //write header to BAM file
    bam_header_write(bam_file_p->bam_fd, bam_header_p);

    //write alignments to BAM file
    while (sam_read1(sam_fd, bam_header_p, bam_p) > 0) {
        bam_write1(bam_file_p->bam_fd, bam_p);
        num_alignments++;
    }

    //close BAM and SAM files, free bam alignment and bam file object
    bam_fclose(bam_file_p);
    sam_close(sam_fd);
    bam_header_destroy(bam_header_p);
    bam_destroy1(bam_p);
    if (time_flag) {
        stop_timer(t1_convert, t2_convert, convert_time);
    }

    //number_of_batchs = 1, convention value for statistics (not real batch)
    number_of_batchs = 1;
}
samfile_t *samopen(const char *fn, const char *mode, const void *aux)
{
	samfile_t *fp;
	fp = (samfile_t*)calloc(1, sizeof(samfile_t));
	if (strchr(mode, 'r')) { // read
		fp->type |= TYPE_READ;
		if (strchr(mode, 'b')) { // binary
			fp->type |= TYPE_BAM;
			fp->x.bam = strcmp(fn, "-")? bam_open(fn, "r") : bam_dopen(fileno(stdin), "r");
			if (fp->x.bam == 0) goto open_err_ret;
			fp->header = bam_header_read(fp->x.bam);
		} else { // text
			fp->x.tamr = sam_open(fn);
			if (fp->x.tamr == 0) goto open_err_ret;
			fp->header = sam_header_read(fp->x.tamr);
			if (fp->header->n_targets == 0) { // no @SQ fields
				if (aux) { // check if aux is present
					bam_header_t *textheader = fp->header;
					fp->header = sam_header_read2((const char*)aux);
					if (fp->header == 0) goto open_err_ret;
					append_header_text(fp->header, textheader->text, textheader->l_text);
					bam_header_destroy(textheader);
				}
				if (fp->header->n_targets == 0 && bam_verbose >= 1)
					fprintf(stderr, "[samopen] no @SQ lines in the header.\n");
			} //else if (bam_verbose >= 2) fprintf(stderr, "[samopen] SAM header is present: %d sequences.\n", fp->header->n_targets);
		}
	} else if (strchr(mode, 'w')) { // write
		fp->header = bam_header_dup((const bam_header_t*)aux);
		if (strchr(mode, 'b')) { // binary
			char bmode[3];
			int i, compress_level = -1;
			for (i = 0; mode[i]; ++i) if (mode[i] >= '0' && mode[i] <= '9') break;
			if (mode[i]) compress_level = mode[i] - '0';
			if (strchr(mode, 'u')) compress_level = 0;
			bmode[0] = 'w'; bmode[1] = compress_level < 0? 0 : compress_level + '0'; bmode[2] = 0;
			fp->type |= TYPE_BAM;
			fp->x.bam = strcmp(fn, "-")? bam_open(fn, bmode) : bam_dopen(fileno(stdout), bmode);
			if (fp->x.bam == 0) goto open_err_ret;
			bam_header_write(fp->x.bam, fp->header);
		} else { // text
			// open file
			fp->x.tamw = strcmp(fn, "-")? fopen(fn, "w") : stdout;
			if (fp->x.tamw == 0) goto open_err_ret;
			if (strchr(mode, 'X')) fp->type |= BAM_OFSTR<<2;
			else if (strchr(mode, 'x')) fp->type |= BAM_OFHEX<<2;
			else fp->type |= BAM_OFDEC<<2;
			// write header
			if (strchr(mode, 'h')) {
				int i;
				bam_header_t *alt;
				// parse the header text 
				alt = bam_header_init();
				alt->l_text = fp->header->l_text; alt->text = fp->header->text;
				sam_header_parse(alt);
				alt->l_text = 0; alt->text = 0;
				// check if there are @SQ lines in the header
				fwrite(fp->header->text, 1, fp->header->l_text, fp->x.tamw); // FIXME: better to skip the trailing NULL
				if (alt->n_targets) { // then write the header text without dumping ->target_{name,len}
					if (alt->n_targets != fp->header->n_targets && bam_verbose >= 1)
						fprintf(stderr, "[samopen] inconsistent number of target sequences. Output the text header.\n");
				} else { // then dump ->target_{name,len}
					for (i = 0; i < fp->header->n_targets; ++i)
						fprintf(fp->x.tamw, "@SQ\tSN:%s\tLN:%d\n", fp->header->target_name[i], fp->header->target_len[i]);
				}
				bam_header_destroy(alt);
			}
		}
	}
	return fp;

open_err_ret:
	free(fp);
	return 0;
}
Exemple #7
0
int bam_cat(int nfn, char * const *fn, const bam_header_t *h, const char* outbam)
{
    BGZF *fp;
    FILE* fp_file;
    uint8_t *buf;
    uint8_t ebuf[BGZF_EMPTY_BLOCK_SIZE];
    const int es=BGZF_EMPTY_BLOCK_SIZE;
    int i;
    
    fp = strcmp(outbam, "-")? bgzf_open(outbam, "w") : bgzf_fdopen(_fileno(stdout), "w");
    if (fp == 0) {
        fprintf(stderr, "[%s] ERROR: fail to open output file '%s'.\n", __FUNCTION__, outbam);
        return 1;
    }
    if (h) bam_header_write(fp, h);
    
    buf = (uint8_t*) malloc(BUF_SIZE);
    for(i = 0; i < nfn; ++i){
        BGZF *in;
        bam_header_t *old;
        int len,j;
        
        in = strcmp(fn[i], "-")? bam_open(fn[i], "r") : bam_dopen(_fileno(stdin), "r");
        if (in == 0) {
            fprintf(stderr, "[%s] ERROR: fail to open file '%s'.\n", __FUNCTION__, fn[i]);
            return -1;
        }
        if (in->open_mode != 'r') return -1;
        
        old = bam_header_read(in);
        if (h == 0 && i == 0) bam_header_write(fp, old);
        
        if (in->block_offset < in->block_length) {
            bgzf_write(fp, (uint8_t*)in->uncompressed_block + in->block_offset, in->block_length - in->block_offset);
            bgzf_flush(fp);
        }
        
        j=0;
#ifdef _USE_KNETFILE
        fp_file=fp->x.fpw;
        while ((len = knet_read(in->x.fpr, buf, BUF_SIZE)) > 0) {
#else  
        fp_file=fp->file;
        while (!feof(in->file) && (len = fread(buf, 1, BUF_SIZE, in->file)) > 0) {
#endif
            if(len<es){
                int diff=es-len;
                if(j==0) {
                    fprintf(stderr, "[%s] ERROR: truncated file?: '%s'.\n", __FUNCTION__, fn[i]);
                    return -1;
                }
                fwrite(ebuf, 1, len, fp_file);
                memcpy(ebuf,ebuf+len,diff);
                memcpy(ebuf+diff,buf,len);
            } else {
                if(j!=0) fwrite(ebuf, 1, es, fp_file);
                len-= es;
                memcpy(ebuf,buf+len,es);
                fwrite(buf, 1, len, fp_file);
            }
            j=1;
        }

        /* check final gzip block */
        {
            const uint8_t gzip1=ebuf[0];
            const uint8_t gzip2=ebuf[1];
            const uint32_t isize=*((uint32_t*)(ebuf+es-4));
            if(((gzip1!=GZIPID1) || (gzip2!=GZIPID2)) || (isize!=0)) {
                fprintf(stderr, "[%s] WARNING: Unexpected block structure in file '%s'.", __FUNCTION__, fn[i]);
                fprintf(stderr, " Possible output corruption.\n");
                fwrite(ebuf, 1, es, fp_file);
            }
        }
        bam_header_destroy(old);
        bgzf_close(in);
    }
    free(buf);
    bgzf_close(fp);
    return 0;
}



int main_cat(int argc, char *argv[])
{
    bam_header_t *h = 0;
    char *outfn = 0;
    int c, ret;
    while ((c = getopt(argc, argv, "h:o:")) >= 0) {
        switch (c) {
            case 'h': {
                tamFile fph = sam_open(optarg);
                if (fph == 0) {
                    fprintf(stderr, "[%s] ERROR: fail to read the header from '%s'.\n", __FUNCTION__, argv[1]);
                    return 1;
                }
                h = sam_header_read(fph);
                sam_close(fph);
                break;
            }
            case 'o': outfn = strdup(optarg); break;
        }
    }
    if (argc - optind < 2) {
        fprintf(stderr, "Usage: samtools cat [-h header.sam] [-o out.bam] <in1.bam> <in2.bam> [...]\n");
        return 1;
    }
    ret = bam_cat(argc - optind, argv + optind, h, outfn? outfn : "-");
    free(outfn);
    return ret;
}
Exemple #8
0
int bam_merge_core2(int by_qname, const char *out, const char *headers, int n, char * const *fn, int flag, const char *reg, int level)
#endif
{
	bamFile fpout, *fp;
	heap1_t *heap;
	bam_header_t *hout = 0;
	bam_header_t *hheaders = NULL;
	int i, j, *RG_len = 0;
	uint64_t idx = 0;
	char **RG = 0, mode[8];
	bam_iter_t *iter = 0;

	if (headers) {
		tamFile fpheaders = sam_open(headers);
		if (fpheaders == 0) {
			const char *message = strerror(errno);
			fprintf(stderr, "[bam_merge_core] cannot open '%s': %s\n", headers, message);
			return -1;
		}
		hheaders = sam_header_read(fpheaders);
		sam_close(fpheaders);
	}

	g_is_by_qname = by_qname;
	fp = (bamFile*)calloc(n, sizeof(bamFile));
	heap = (heap1_t*)calloc(n, sizeof(heap1_t));
	iter = (bam_iter_t*)calloc(n, sizeof(bam_iter_t));
	// prepare RG tag
	if (flag & MERGE_RG) {
		RG = (char**)calloc(n, sizeof(void*));
		RG_len = (int*)calloc(n, sizeof(int));
		for (i = 0; i != n; ++i) {
			int l = strlen(fn[i]);
			const char *s = fn[i];
			if (l > 4 && strcmp(s + l - 4, ".bam") == 0) l -= 4;
			for (j = l - 1; j >= 0; --j) if (s[j] == '/') break;
			++j; l -= j;
			RG[i] = calloc(l + 1, 1);
			RG_len[i] = l;
			strncpy(RG[i], s + j, l);
		}
	}
	// read the first
	for (i = 0; i != n; ++i) {
		bam_header_t *hin;
		fp[i] = bam_open(fn[i], "r");
		if (fp[i] == 0) {
			int j;
			fprintf(stderr, "[bam_merge_core] fail to open file %s\n", fn[i]);
			for (j = 0; j < i; ++j) bam_close(fp[j]);
			free(fp); free(heap);
			// FIXME: possible memory leak
			return -1;
		}
		hin = bam_header_read(fp[i]);
		if (i == 0) { // the first BAM
			hout = hin;
		} else { // validate multiple baf
			int min_n_targets = hout->n_targets;
			if (hin->n_targets < min_n_targets) min_n_targets = hin->n_targets;

			for (j = 0; j < min_n_targets; ++j)
				if (strcmp(hout->target_name[j], hin->target_name[j]) != 0) {
					fprintf(stderr, "[bam_merge_core] different target sequence name: '%s' != '%s' in file '%s'\n",
							hout->target_name[j], hin->target_name[j], fn[i]);
					return -1;
				}

			// If this input file has additional target reference sequences,
			// add them to the headers to be output
			if (hin->n_targets > hout->n_targets) {
				swap_header_targets(hout, hin);
				// FIXME Possibly we should also create @SQ text headers
				// for the newly added reference sequences
			}

			bam_header_destroy(hin);
		}
	}

	if (hheaders) {
		// If the text headers to be swapped in include any @SQ headers,
		// check that they are consistent with the existing binary list
		// of reference information.
		if (hheaders->n_targets > 0) {
			if (hout->n_targets != hheaders->n_targets) {
				fprintf(stderr, "[bam_merge_core] number of @SQ headers in '%s' differs from number of target sequences\n", headers);
				if (!reg) return -1;
			}
			for (j = 0; j < hout->n_targets; ++j)
				if (strcmp(hout->target_name[j], hheaders->target_name[j]) != 0) {
					fprintf(stderr, "[bam_merge_core] @SQ header '%s' in '%s' differs from target sequence\n", hheaders->target_name[j], headers);
					if (!reg) return -1;
				}
		}

		swap_header_text(hout, hheaders);
		bam_header_destroy(hheaders);
	}

	if (reg) {
		int tid, beg, end;
		if (bam_parse_region(hout, reg, &tid, &beg, &end) < 0) {
			fprintf(stderr, "[%s] Malformated region string or undefined reference name\n", __func__);
			return -1;
		}
		for (i = 0; i < n; ++i) {
			bam_index_t *idx;
			idx = bam_index_load(fn[i]);
			iter[i] = bam_iter_query(idx, tid, beg, end);
			bam_index_destroy(idx);
		}
	}

	for (i = 0; i < n; ++i) {
		heap1_t *h = heap + i;
		h->i = i;
		h->b = (bam1_t*)calloc(1, sizeof(bam1_t));
		if (bam_iter_read(fp[i], iter[i], h->b) >= 0) {
			h->pos = ((uint64_t)h->b->core.tid<<32) | (uint32_t)((int32_t)h->b->core.pos+1)<<1 | bam1_strand(h->b);
			h->idx = idx++;
		}
		else h->pos = HEAP_EMPTY;
	}
	if (flag & MERGE_UNCOMP) level = 0;
	else if (flag & MERGE_LEVEL1) level = 1;
	strcpy(mode, "w");
	if (level >= 0) sprintf(mode + 1, "%d", level < 9? level : 9);
	if ((fpout = strcmp(out, "-")? bam_open(out, "w") : bam_dopen(fileno(stdout), "w")) == 0) {
		fprintf(stderr, "[%s] fail to create the output file.\n", __func__);
		return -1;
	}
	bam_header_write(fpout, hout);
	bam_header_destroy(hout);
#ifndef _PBGZF_USE 
	if (!(flag & MERGE_UNCOMP)) bgzf_mt(fpout, n_threads, 256);
#endif

	ks_heapmake(heap, n, heap);
	while (heap->pos != HEAP_EMPTY) {
		bam1_t *b = heap->b;
		if (flag & MERGE_RG) {
			uint8_t *rg = bam_aux_get(b, "RG");
			if (rg) bam_aux_del(b, rg);
			bam_aux_append(b, "RG", 'Z', RG_len[heap->i] + 1, (uint8_t*)RG[heap->i]);
		}
		bam_write1_core(fpout, &b->core, b->data_len, b->data);
		if ((j = bam_iter_read(fp[heap->i], iter[heap->i], b)) >= 0) {
			heap->pos = ((uint64_t)b->core.tid<<32) | (uint32_t)((int)b->core.pos+1)<<1 | bam1_strand(b);
			heap->idx = idx++;
		} else if (j == -1) {
			heap->pos = HEAP_EMPTY;
			free(heap->b->data); free(heap->b);
			heap->b = 0;
		} else fprintf(stderr, "[bam_merge_core] '%s' is truncated. Continue anyway.\n", fn[heap->i]);
		ks_heapadjust(heap, 0, n, heap);
	}

	if (flag & MERGE_RG) {
		for (i = 0; i != n; ++i) free(RG[i]);
		free(RG); free(RG_len);
	}
	for (i = 0; i != n; ++i) {
		bam_iter_destroy(iter[i]);
		bam_close(fp[i]);
	}
	bam_close(fpout);
	free(fp); free(heap); free(iter);
	return 0;
}
Exemple #9
0
samfile_t *samopen(const char *fn, const char *mode, const void *aux)
{
	samfile_t *fp;
	fp = (samfile_t*)calloc(1, sizeof(samfile_t));
	if (mode[0] == 'r') { // read
		fp->type |= TYPE_READ;
		if (mode[1] == 'b') { // binary
			fp->type |= TYPE_BAM;
			fp->x.bam = strcmp(fn, "-")? bam_open(fn, "r") : bam_dopen(fileno(stdin), "r");
			if (fp->x.bam == 0) goto open_err_ret;
			fp->header = bam_header_read(fp->x.bam);
		} else { // text
			fp->x.tamr = sam_open(fn);
			if (fp->x.tamr == 0) goto open_err_ret;
			fp->header = sam_header_read(fp->x.tamr);
			if (fp->header->n_targets == 0) { // no @SQ fields
				if (aux) { // check if aux is present
					bam_header_destroy(fp->header);
					fp->header = sam_header_read2((const char*)aux);
				}
				if (fp->header->n_targets == 0)
					fprintf(stderr, "[samopen] empty header.\n");
			} else fprintf(stderr, "[samopen] SAM header is present: %d sequences.\n", fp->header->n_targets);
		}
	} else if (mode[0] == 'w') { // write
		fp->header = bam_header_dup((const bam_header_t*)aux);
		if (mode[1] == 'b') { // binary
			fp->type |= TYPE_BAM;
			fp->x.bam = strcmp(fn, "-")? bam_open(fn, "w") : bam_dopen(fileno(stdout), "w");
			if (fp->x.bam == 0) goto open_err_ret;
			bam_header_write(fp->x.bam, fp->header);
		} else { // text
			// open file
			fp->x.tamw = strcmp(fn, "-")? fopen(fn, "w") : stdout;
			if (fp->x.tamr == 0) goto open_err_ret;
			// write header
			if (strstr(mode, "h")) {
				int i;
				bam_header_t *alt;
				// parse the header text 
				alt = bam_header_init();
				alt->l_text = fp->header->l_text; alt->text = fp->header->text;
				sam_header_parse(alt);
				alt->l_text = 0; alt->text = 0;
				// check if there are @SQ lines in the header
				if (alt->n_targets) { // then write the header text without dumping ->target_{name,len}
					if (alt->n_targets != fp->header->n_targets)
						fprintf(stderr, "[samopen] inconsistent number of target sequences.\n");
					fwrite(fp->header->text, 1, fp->header->l_text, fp->x.tamw);
				} else { // then dump ->target_{name,len}
					for (i = 0; i < fp->header->n_targets; ++i)
						fprintf(fp->x.tamw, "@SQ\tSN:%s\tLN:%d\n", fp->header->target_name[i], fp->header->target_len[i]);
				}
				bam_header_destroy(alt);
			}
		}
	}
	return fp;

open_err_ret:
	free(fp);
	return 0;
}