3 bam_cat -- efficiently concatenates bam files
\r
5 bam_cat can be used to concatenate BAM files. Under special
\r
6 circumstances, it can be used as an alternative to 'samtools merge' to
\r
7 concatenate multiple sorted files into a single sorted file. For this
\r
8 to work each file must be sorted, and the sorted files must be given
\r
9 as command line arguments in order such that the final read in file i
\r
10 is less than or equal to the first read in file i+1.
\r
12 This code is derived from the bam_reheader function in samtools 0.1.8
\r
13 and modified to perform concatenation by Chris Saunders on behalf of
\r
21 Original SAMtools work copyright (c) 2008-2009 Genome Research Ltd.
\r
22 Modified SAMtools work copyright (c) 2010 Illumina, Inc.
\r
24 Permission is hereby granted, free of charge, to any person obtaining a copy
\r
25 of this software and associated documentation files (the "Software"), to deal
\r
26 in the Software without restriction, including without limitation the rights
\r
27 to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
\r
28 copies of the Software, and to permit persons to whom the Software is
\r
29 furnished to do so, subject to the following conditions:
\r
31 The above copyright notice and this permission notice shall be included in
\r
32 all copies or substantial portions of the Software.
\r
34 THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
\r
35 IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
\r
36 FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
\r
37 AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
\r
38 LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
\r
39 OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
\r
49 CFLAGS+=-g -Wall -O2 -D_FILE_OFFSET_BITS=64 -D_USE_KNETFILE -I$(SAMTOOLS_DIR)
\r
50 LDFLAGS+=-L$(SAMTOOLS_DIR)
\r
65 #define BUF_SIZE 0x10000
\r
70 #define BGZF_EMPTY_BLOCK_SIZE 28
\r
73 int bam_cat(int nfn, char * const *fn, const bam_header_t *h, const char* outbam)
\r
78 uint8_t ebuf[BGZF_EMPTY_BLOCK_SIZE];
\r
79 const int es=BGZF_EMPTY_BLOCK_SIZE;
\r
82 fp = strcmp(outbam, "-")? bgzf_open(outbam, "w") : bgzf_fdopen(fileno(stdout), "w");
\r
84 fprintf(stderr, "[%s] ERROR: fail to open output file '%s'.\n", __func__, outbam);
\r
87 if (h) bam_header_write(fp, h);
\r
89 if ((buf = (uint8_t*) malloc(BUF_SIZE)) == NULL) {
\r
90 fprintf(stderr, "[%s] failed to allocate input buffer.\n", __func__);
\r
93 for(i = 0; i < nfn; ++i){
\r
98 in = strcmp(fn[i], "-")? bam_open(fn[i], "r") : bam_dopen(fileno(stdin), "r");
\r
100 fprintf(stderr, "[%s] ERROR: fail to open file '%s'.\n", __func__, fn[i]);
\r
103 if (in->open_mode != 'r') return -1;
\r
105 old = bam_header_read(in);
\r
106 if (h == 0 && i == 0) bam_header_write(fp, old);
\r
108 if (in->block_offset < in->block_length) {
\r
109 bgzf_write(fp, in->uncompressed_block + in->block_offset, in->block_length - in->block_offset);
\r
114 #ifdef _USE_KNETFILE
\r
116 while ((len = knet_read(in->x.fpr, buf, BUF_SIZE)) > 0) {
\r
119 while (!feof(in->file) && (len = fread(buf, 1, BUF_SIZE, in->file)) > 0) {
\r
124 fprintf(stderr, "[%s] ERROR: truncated file?: '%s'.\n", __func__, fn[i]);
\r
127 if (fwrite(ebuf, 1, len, fp_file) < len) {
\r
128 fprintf(stderr, "[%s] failed to write to output file.\n", __func__);
\r
131 memcpy(ebuf,ebuf+len,diff);
\r
132 memcpy(ebuf+diff,buf,len);
\r
136 if (fwrite(ebuf, 1, es, fp_file) < es) {
\r
137 fprintf(stderr, "[%s] failed to write to output file.\n", __func__);
\r
142 memcpy(ebuf,buf+len,es);
\r
143 if (fwrite(buf, 1, len, fp_file) < len) {
\r
144 fprintf(stderr, "[%s] failed to write to output file.\n", __func__);
\r
151 /* check final gzip block */
\r
153 const uint8_t gzip1=ebuf[0];
\r
154 const uint8_t gzip2=ebuf[1];
\r
155 const uint32_t isize=*((uint32_t*)(ebuf+es-4));
\r
156 if(((gzip1!=GZIPID1) || (gzip2!=GZIPID2)) || (isize!=0)) {
\r
157 fprintf(stderr, "[%s] WARNING: Unexpected block structure in file '%s'.", __func__, fn[i]);
\r
158 fprintf(stderr, " Possible output corruption.\n");
\r
159 if (fwrite(ebuf, 1, es, fp_file) < es) {
\r
160 fprintf(stderr, " Also, failed to write final block.\n");
\r
164 bam_header_destroy(old);
\r
174 int main_cat(int argc, char *argv[])
\r
176 bam_header_t *h = 0;
\r
179 while ((c = getopt(argc, argv, "h:o:")) >= 0) {
\r
182 tamFile fph = sam_open(optarg);
\r
184 fprintf(stderr, "[%s] ERROR: fail to read the header from '%s'.\n", __func__, argv[1]);
\r
187 h = sam_header_read(fph);
\r
191 case 'o': outfn = strdup(optarg); break;
\r
194 if (argc - optind < 2) {
\r
195 fprintf(stderr, "Usage: samtools cat [-h header.sam] [-o out.bam] <in1.bam> <in2.bam> [...]\n");
\r
198 ret = bam_cat(argc - optind, argv + optind, h, outfn? outfn : "-");
\r