X-Git-Url: https://git.donarmstrong.com/?a=blobdiff_plain;f=bam_aux.c;h=fbcd9822b233dab64f6c861fb332846acb951ce7;hb=6ee6fd65332260af63fbfbf0a625ea3847371eda;hp=d6f18ea281712a7d89007d7d87291cbf13718c1b;hpb=63e4e54189bc9e7f43852edf865f95e713b93dc4;p=samtools.git

diff --git a/bam_aux.c b/bam_aux.c
index d6f18ea..fbcd982 100644
--- a/bam_aux.c
+++ b/bam_aux.c
@@ -1,7 +1,9 @@
 #include <ctype.h>
 #include "bam.h"
 #include "khash.h"
+typedef char *str_p;
 KHASH_MAP_INIT_STR(s, int)
+KHASH_MAP_INIT_STR(r2l, str_p)
 
 void bam_aux_append(bam1_t *b, const char tag[2], char type, int len, uint8_t *data)
 {
@@ -18,24 +20,46 @@ void bam_aux_append(bam1_t *b, const char tag[2], char type, int len, uint8_t *d
 	memcpy(b->data + ori_len + 3, data, len);
 }
 
-uint8_t *bam_aux_get(bam1_t *b, const char tag[2])
+uint8_t *bam_aux_get_core(bam1_t *b, const char tag[2])
+{
+	return bam_aux_get(b, tag);
+}
+
+#define __skip_tag(s) do { \
+		int type = toupper(*(s));										\
+		++(s);															\
+		if (type == 'C' || type == 'A') ++(s);							\
+		else if (type == 'S') (s) += 2;									\
+		else if (type == 'I' || type == 'F') (s) += 4;					\
+		else if (type == 'D') (s) += 8;									\
+		else if (type == 'Z' || type == 'H') { while (*(s)) ++(s); ++(s); } \
+	} while (0)
+
+uint8_t *bam_aux_get(const bam1_t *b, const char tag[2])
 {
 	uint8_t *s;
 	int y = tag[0]<<8 | tag[1];
 	s = bam1_aux(b);
 	while (s < b->data + b->data_len) {
-		int type, x = (int)s[0]<<8 | s[1];
+		int x = (int)s[0]<<8 | s[1];
 		s += 2;
 		if (x == y) return s;
-		type = toupper(*s); ++s;
-		if (type == 'C') ++s;
-		else if (type == 'S') s += 2;
-		else if (type == 'I' || type == 'F') s += 4;
-		else if (type == 'D') s += 8;
-		else if (type == 'Z' || type == 'H') { while (*s) putchar(*s++); ++s; }
+		__skip_tag(s);
 	}
 	return 0;
 }
+// s MUST BE returned by bam_aux_get()
+int bam_aux_del(bam1_t *b, uint8_t *s)
+{
+	uint8_t *p, *aux;
+	aux = bam1_aux(b);
+	p = s - 2;
+	__skip_tag(s);
+	memmove(p, s, b->l_aux - (s - aux));
+	b->data_len -= s - p;
+	b->l_aux -= s - p;
+	return 0;
+}
 
 void bam_init_header_hash(bam_header_t *header)
 {
@@ -65,7 +89,7 @@ int32_t bam_get_tid(const bam_header_t *header, const char *seq_name)
 	return k == kh_end(h)? -1 : kh_value(h, k);
 }
 
-void bam_parse_region(bam_header_t *header, const char *str, int *ref_id, int *begin, int *end)
+int bam_parse_region(bam_header_t *header, const char *str, int *ref_id, int *begin, int *end)
 {
 	char *s, *p;
 	int i, l, k;
@@ -74,7 +98,7 @@ void bam_parse_region(bam_header_t *header, const char *str, int *ref_id, int *b
 
 	bam_init_header_hash(header);
 	h = (khash_t(s)*)header->hash;
-	
+
 	l = strlen(str);
 	p = s = (char*)malloc(l+1);
 	/* squeeze out "," */
@@ -86,12 +110,12 @@ void bam_parse_region(bam_header_t *header, const char *str, int *ref_id, int *b
 	iter = kh_get(s, h, s); /* get the ref_id */
 	if (iter == kh_end(h)) { // name not found
 		*ref_id = -1; free(s);
-		return;
+		return -1;
 	}
 	*ref_id = kh_value(h, iter);
 	if (i == k) { /* dump the whole sequence */
 		*begin = 0; *end = 1<<29; free(s);
-		return;
+		return 0;
 	}
 	for (p = s + i + 1; i != k; ++i) if (s[i] == '-') break;
 	*begin = atoi(p);
@@ -100,8 +124,12 @@ void bam_parse_region(bam_header_t *header, const char *str, int *ref_id, int *b
 		*end = atoi(p);
 	} else *end = 1<<29;
 	if (*begin > 0) --*begin;
-	assert(*begin <= *end);
 	free(s);
+	if (*begin > *end) {
+		fprintf(stderr, "[bam_parse_region] invalid region.\n");
+		return -1;
+	}
+	return 0;
 }
 
 int32_t bam_aux2i(const uint8_t *s)