X-Git-Url: https://git.donarmstrong.com/?a=blobdiff_plain;f=bam_plcmd.c;h=75f5691c7815184b43d29e9c9072df5856ddff5e;hb=f221655a2515b417a545caece40b6985e927fb83;hp=4bdc351e18a38322bd39dac30433aea3212b3b5a;hpb=0993e913dc402cbb11783bf5d3f96eed1c437e06;p=samtools.git

diff --git a/bam_plcmd.c b/bam_plcmd.c
index 4bdc351..75f5691 100644
--- a/bam_plcmd.c
+++ b/bam_plcmd.c
@@ -91,6 +91,21 @@ static khash_t(64) *load_pos(const char *fn, bam_header_t *h)
 	return hash;
 }
 
+static inline int printw(int c, FILE *fp)
+{
+	char buf[16];
+	int l, x;
+	if (c == 0) return fputc('0', fp);
+	for (l = 0, x = c < 0? -c : c; x > 0; x /= 10) buf[l++] = x%10 + '0';
+	if (c < 0) buf[l++] = '-';
+	buf[l] = 0;
+	for (x = 0; x < l/2; ++x) {
+		int y = buf[x]; buf[x] = buf[l-1-x]; buf[l-1-x] = y;
+	}
+	fputs(buf, fp);
+	return 0;
+}
+
 // an analogy to pileup_func() below
 static int glt3_func(uint32_t tid, uint32_t pos, int n, const bam_pileup1_t *pu, void *data)
 {
@@ -160,29 +175,38 @@ static int glt3_func(uint32_t tid, uint32_t pos, int n, const bam_pileup1_t *pu,
 	return 0;
 }
 
-static void pileup_seq(const bam_pileup1_t *p, int pos, int ref_len, const char *ref)
+static inline void pileup_seq(const bam_pileup1_t *p, int pos, int ref_len, const char *ref)
 {
-	if (p->is_head) printf("^%c", p->b->core.qual > 93? 126 : p->b->core.qual + 33);
+	int j;
+	if (p->is_head) {
+		putchar('^');
+		putchar(p->b->core.qual > 93? 126 : p->b->core.qual + 33);
+	}
 	if (!p->is_del) {
-		int j, rb, c = bam_nt16_rev_table[bam1_seqi(bam1_seq(p->b), p->qpos)];
-		rb = (ref && pos < ref_len)? ref[pos] : 'N';
-		if (c == '=' || toupper(c) == toupper(rb)) c = bam1_strand(p->b)? ',' : '.';
-		else c = bam1_strand(p->b)? tolower(c) : toupper(c);
+		int c = bam_nt16_rev_table[bam1_seqi(bam1_seq(p->b), p->qpos)];
+		if (ref) {
+			int rb = pos < ref_len? ref[pos] : 'N';
+			if (c == '=' || bam_nt16_table[c] == bam_nt16_table[rb]) c = bam1_strand(p->b)? ',' : '.';
+			else c = bam1_strand(p->b)? tolower(c) : toupper(c);
+		} else {
+			if (c == '=') c = bam1_strand(p->b)? ',' : '.';
+			else c = bam1_strand(p->b)? tolower(c) : toupper(c);
+		}
 		putchar(c);
-		if (p->indel > 0) {
-			printf("+%d", p->indel);
-			for (j = 1; j <= p->indel; ++j) {
-				c = bam_nt16_rev_table[bam1_seqi(bam1_seq(p->b), p->qpos + j)];
-				putchar(bam1_strand(p->b)? tolower(c) : toupper(c));
-			}
-		} else if (p->indel < 0) {
-			printf("%d", p->indel);
-			for (j = 1; j <= -p->indel; ++j) {
-				c = (ref && (int)pos+j < ref_len)? ref[pos+j] : 'N';
-				putchar(bam1_strand(p->b)? tolower(c) : toupper(c));
-			}
+	} else putchar(p->is_refskip? (bam1_strand(p->b)? '<' : '>') : '*');
+	if (p->indel > 0) {
+		putchar('+'); printw(p->indel, stdout);
+		for (j = 1; j <= p->indel; ++j) {
+			int c = bam_nt16_rev_table[bam1_seqi(bam1_seq(p->b), p->qpos + j)];
+			putchar(bam1_strand(p->b)? tolower(c) : toupper(c));
 		}
-	} else putchar('*');
+	} else if (p->indel < 0) {
+		printw(p->indel, stdout);
+		for (j = 1; j <= -p->indel; ++j) {
+			int c = (ref && (int)pos+j < ref_len)? ref[pos+j] : 'N';
+			putchar(bam1_strand(p->b)? tolower(c) : toupper(c));
+		}
+	}
 	if (p->is_tail) putchar('$');
 }
 
@@ -256,22 +280,29 @@ static int pileup_func(uint32_t tid, uint32_t pos, int n, const bam_pileup1_t *p
 		}
 	}
 	// print the first 3 columns
-	printf("%s\t%d\t%c\t", d->h->target_name[tid], pos + 1, rb);
+	fputs(d->h->target_name[tid], stdout); putchar('\t');
+	printw(pos+1, stdout); putchar('\t'); putchar(rb); putchar('\t');
 	// print consensus information if required
-	if (d->format & BAM_PLF_CNS)
-		printf("%c\t%d\t%d\t%d\t", bam_nt16_rev_table[cns>>28], cns>>8&0xff, cns&0xff, cns>>16&0xff);
-	// print pileup sequences
-	printf("%d\t", n);
-	rms_aux = 0; // we need to recalculate rms_mapq when -c is not flagged on the command line
-	for (i = 0; i < n; ++i) {
-		const bam_pileup1_t *p = pu + i;
-		int tmp = p->b->core.qual < d->c->cap_mapQ? p->b->core.qual : d->c->cap_mapQ;
-		rms_aux += tmp * tmp;
-		pileup_seq(p, pos, d->len, d->ref);
+	if (d->format & BAM_PLF_CNS) {
+		putchar(bam_nt16_rev_table[cns>>28]); putchar('\t');
+		printw(cns>>8&0xff, stdout); putchar('\t');
+		printw(cns&0xff, stdout); putchar('\t');
+		printw(cns>>16&0xff, stdout); putchar('\t');
 	}
+	// print pileup sequences
+	printw(n, stdout); putchar('\t');
+	for (i = 0; i < n; ++i)
+		pileup_seq(pu + i, pos, d->len, d->ref);
 	// finalize rms_mapq
-	rms_aux = (uint64_t)(sqrt((double)rms_aux / n) + .499);
-	if (rms_mapq < 0) rms_mapq = rms_aux;
+	if (d->format & BAM_PLF_CNS) {
+		for (i = rms_aux = 0; i < n; ++i) {
+			const bam_pileup1_t *p = pu + i;
+			int tmp = p->b->core.qual < d->c->cap_mapQ? p->b->core.qual : d->c->cap_mapQ;
+			rms_aux += tmp * tmp;
+		}
+		rms_aux = (uint64_t)(sqrt((double)rms_aux / n) + .499);
+		if (rms_mapq < 0) rms_mapq = rms_aux;
+	}
 	putchar('\t');
 	// print quality
 	for (i = 0; i < n; ++i) {
@@ -310,7 +341,7 @@ static int pileup_func(uint32_t tid, uint32_t pos, int n, const bam_pileup1_t *p
 		for (i = 0; i < n; ++i) {
 			int x = pu[i].qpos;
 			int l = pu[i].b->core.l_qseq;
-			printf("%d,", x < l/2? x+1 : -((l-1)-x+1));
+			printw(x < l/2? x+1 : -((l-1)-x+1), stdout); putchar(',');
 		}
 	}
 	putchar('\n');
@@ -442,6 +473,7 @@ int bam_pileup(int argc, char *argv[])
 			bam_plp_set_mask(iter, d->mask);
 			while ((ret = samread(fp, b)) >= 0) {
 				int skip = 0;
+				if ((int)b->core.tid < 0) break;
 				// update d->ref if necessary
 				if (d->fai && (int)b->core.tid != d->tid) {
 					free(d->ref);
@@ -496,6 +528,7 @@ int bam_pileup(int argc, char *argv[])
 #define MPLP_NO_COMP 0x20
 #define MPLP_NO_ORPHAN 0x40
 #define MPLP_REALN   0x80
+#define MPLP_SAMINFO 0x100
 
 typedef struct {
 	int max_mq, min_mq, flag, min_baseQ, capQ_thres;
@@ -682,7 +715,7 @@ static int mpileup(mplp_conf_t *conf, int n, char **fn)
 			for (i = 0; i < gplp.n; ++i)
 				bcf_call_glfgen(gplp.n_plp[i], gplp.plp[i], ref16, bca, bcr + i);
 			bcf_call_combine(gplp.n, bcr, ref16, &bc);
-			bcf_call2bcf(tid, pos, &bc, b);
+			bcf_call2bcf(tid, pos, &bc, b, (conf->flag&MPLP_SAMINFO)? bcr : 0);
 			bcf_write(bp, bh, b);
 			bcf_destroy(b);
 		} else {
@@ -737,7 +770,8 @@ int bam_mpileup(int argc, char *argv[])
 	mplp.max_mq = 60;
 	mplp.min_baseQ = 13;
 	mplp.capQ_thres = 0;
-	while ((c = getopt(argc, argv, "gf:r:l:M:q:Q:uaORC:")) >= 0) {
+	mplp.flag = MPLP_NO_ORPHAN | MPLP_REALN;
+	while ((c = getopt(argc, argv, "gf:r:l:M:q:Q:uaORC:BI")) >= 0) {
 		switch (c) {
 		case 'f':
 			mplp.fai = fai_load(optarg);
@@ -748,8 +782,10 @@ int bam_mpileup(int argc, char *argv[])
 		case 'g': mplp.flag |= MPLP_GLF; break;
 		case 'u': mplp.flag |= MPLP_NO_COMP | MPLP_GLF; break;
 		case 'a': mplp.flag |= MPLP_NO_ORPHAN | MPLP_REALN; break;
+		case 'B': mplp.flag &= ~MPLP_REALN & ~MPLP_NO_ORPHAN; break;
 		case 'O': mplp.flag |= MPLP_NO_ORPHAN; break;
 		case 'R': mplp.flag |= MPLP_REALN; break;
+		case 'I': mplp.flag |= MPLP_SAMINFO; break;
 		case 'C': mplp.capQ_thres = atoi(optarg); break;
 		case 'M': mplp.max_mq = atoi(optarg); break;
 		case 'q': mplp.min_mq = atoi(optarg); break;
@@ -767,6 +803,8 @@ int bam_mpileup(int argc, char *argv[])
 		fprintf(stderr, "         -q INT      filter out alignment with MQ smaller than INT [%d]\n", mplp.min_mq);
 		fprintf(stderr, "         -g          generate BCF output\n");
 		fprintf(stderr, "         -u          do not compress BCF output\n");
+		fprintf(stderr, "         -B          disable BAQ computation\n");
+		fprintf(stderr, "         -I          additional individual information (DP and ST)\n");
 		fprintf(stderr, "\n");
 		fprintf(stderr, "Notes: Assuming diploid individuals.\n\n");
 		return 1;