X-Git-Url: http://woldlab.caltech.edu/gitweb/?p=samtools.git;a=blobdiff_plain;f=bam.h;h=8e26ea69d3b7c6b008aba6112052310a4072b08a;hp=83c03ada21098eda911caabb0e56db3864599e73;hb=cefa18095b2479339b08111936313066ec548657;hpb=b27e00385f41769d03a8cca4dbd71275fc9fa906 diff --git a/bam.h b/bam.h index 83c03ad..8e26ea6 100644 --- a/bam.h +++ b/bam.h @@ -45,21 +45,7 @@ #include #include -#define _IOLIB 2 - -#if _IOLIB == 1 && !defined(_NO_RAZF) -#define BAM_TRUE_OFFSET -#include "razf.h" -/*! @abstract BAM file handler */ -typedef RAZF *bamFile; -#define bam_open(fn, mode) razf_open(fn, mode) -#define bam_dopen(fd, mode) razf_dopen(fd, mode) -#define bam_close(fp) razf_close(fp) -#define bam_read(fp, buf, size) razf_read(fp, buf, size) -#define bam_write(fp, buf, size) razf_write(fp, buf, size) -#define bam_tell(fp) razf_tell(fp) -#define bam_seek(fp, pos, dir) razf_seek(fp, pos, dir) -#elif _IOLIB == 2 +#ifndef BAM_LITE #define BAM_VIRTUAL_OFFSET16 #include "bgzf.h" /*! @abstract BAM file handler */ @@ -71,18 +57,15 @@ typedef BGZF *bamFile; #define bam_write(fp, buf, size) bgzf_write(fp, buf, size) #define bam_tell(fp) bgzf_tell(fp) #define bam_seek(fp, pos, dir) bgzf_seek(fp, pos, dir) -#elif _IOLIB == 3 -#define BAM_VIRTUAL_OFFSET16 -#include "razf.h" -/*! @abstract BAM file handler */ -typedef RAZF *bamFile; -#define bam_open(fn, mode) razf_open2(fn, mode) -#define bam_dopen(fd, mode) razf_dopen2(fd, mode) -#define bam_close(fp) razf_close(fp) -#define bam_read(fp, buf, size) razf_read(fp, buf, size) -#define bam_write(fp, buf, size) razf_write(fp, buf, size) -#define bam_tell(fp) razf_tell2(fp) -#define bam_seek(fp, pos, dir) razf_seek2(fp, pos, dir) +#else +#define BAM_TRUE_OFFSET +#include +typedef gzFile bamFile; +#define bam_open(fn, mode) gzopen(fn, mode) +#define bam_dopen(fd, mode) gzdopen(fd, mode) +#define bam_close(fp) gzclose(fp) +#define bam_read(fp, buf, size) gzread(fp, buf, size) +/* no bam_write/bam_tell/bam_seek() here */ #endif /*! @typedef @@ -90,6 +73,7 @@ typedef RAZF *bamFile; @field n_targets number of reference sequences @field target_name names of the reference sequences @field target_len lengths of the referene sequences + @field dict header dictionary @field hash hash table for fast name lookup @field rg2lib hash table for @RG-ID -> LB lookup @field l_text length of the plain text in the header @@ -102,8 +86,8 @@ typedef struct { int32_t n_targets; char **target_name; uint32_t *target_len; - void *hash, *rg2lib; - int l_text; + void *dict, *hash, *rg2lib; + size_t l_text, n_text; char *text; } bam_header_t; @@ -130,6 +114,10 @@ typedef struct { /*! @abstract optical or PCR duplicate */ #define BAM_FDUP 1024 +#define BAM_OFDEC 0 +#define BAM_OFHEX 1 +#define BAM_OFSTR 2 + /*! @abstract defautl mask for pileup */ #define BAM_DEF_MASK (BAM_FUNMAP | BAM_FSECONDARY | BAM_FQCFAIL | BAM_FDUP) @@ -202,6 +190,8 @@ typedef struct { uint8_t *data; } bam1_t; +typedef struct __bam_iter_t *bam_iter_t; + #define bam1_strand(b) (((b)->core.flag&BAM_FREVERSE) != 0) #define bam1_mstrand(b) (((b)->core.flag&BAM_FMREVERSE) != 0) @@ -284,6 +274,10 @@ extern char bam_nt16_nt4_table[]; extern "C" { #endif + /********************* + * Low-level SAM I/O * + *********************/ + /*! @abstract TAM file handler */ typedef struct __tamFile_t *tamFile; @@ -335,6 +329,7 @@ extern "C" { be destroyed in the first place. */ int sam_header_parse(bam_header_t *h); + int32_t bam_get_tid(const bam_header_t *header, const char *seq_name); /*! @abstract Parse @RG lines a update a header struct @@ -348,12 +343,22 @@ extern "C" { #define sam_write1(header, b) bam_view1(header, b) + + /******************************** + * APIs for string dictionaries * + ********************************/ + int bam_strmap_put(void *strmap, const char *rg, const char *lib); const char *bam_strmap_get(const void *strmap, const char *rg); void *bam_strmap_dup(const void*); void *bam_strmap_init(); void bam_strmap_destroy(void *strmap); + + /********************* + * Low-level BAM I/O * + *********************/ + /*! @abstract Initialize a header structure. @return the pointer to the header structure @@ -436,8 +441,8 @@ extern "C" { @abstract Free the memory allocated for an alignment. @param b pointer to an alignment */ -#define bam_destroy1(b) do { \ - free((b)->data); free(b); \ +#define bam_destroy1(b) do { \ + if (b) { free((b)->data); free(b); } \ } while (0) /*! @@ -448,6 +453,15 @@ extern "C" { */ char *bam_format1(const bam_header_t *header, const bam1_t *b); + char *bam_format1_core(const bam_header_t *header, const bam1_t *b, int of); + + const char *bam_get_library(bam_header_t *header, const bam1_t *b); + + + /*************** + * pileup APIs * + ***************/ + /*! @typedef @abstract Structure for one alignment covering the pileup position. @field b pointer to the alignment @@ -469,11 +483,25 @@ extern "C" { uint32_t is_del:1, is_head:1, is_tail:1; } bam_pileup1_t; - struct __bam_plbuf_t; - /*! @abstract pileup buffer */ - typedef struct __bam_plbuf_t bam_plbuf_t; + typedef int (*bam_plp_auto_f)(void *data, bam1_t *b); - void bam_plbuf_set_mask(bam_plbuf_t *buf, int mask); + struct __bam_plp_t; + typedef struct __bam_plp_t *bam_plp_t; + + bam_plp_t bam_plp_init(bam_plp_auto_f func, void *data); + int bam_plp_push(bam_plp_t iter, const bam1_t *b); + const bam_pileup1_t *bam_plp_next(bam_plp_t iter, int *_tid, int *_pos, int *_n_plp); + const bam_pileup1_t *bam_plp_auto(bam_plp_t iter, int *_tid, int *_pos, int *_n_plp); + void bam_plp_set_mask(bam_plp_t iter, int mask); + void bam_plp_reset(bam_plp_t iter); + void bam_plp_destroy(bam_plp_t iter); + + struct __bam_mplp_t; + typedef struct __bam_mplp_t *bam_mplp_t; + + bam_mplp_t bam_mplp_init(int n, bam_plp_auto_f func, void **data); + void bam_mplp_destroy(bam_mplp_t iter); + int bam_mplp_auto(bam_mplp_t iter, int *_tid, int *_pos, int *n_plp, const bam_pileup1_t **plp); /*! @typedef @abstract Type of function to be called by bam_plbuf_push(). @@ -486,46 +514,20 @@ extern "C" { */ typedef int (*bam_pileup_f)(uint32_t tid, uint32_t pos, int n, const bam_pileup1_t *pl, void *data); - /*! - @abstract Reset a pileup buffer for another pileup process - @param buf the pileup buffer to be reset - */ - void bam_plbuf_reset(bam_plbuf_t *buf); + typedef struct { + bam_plp_t iter; + bam_pileup_f func; + void *data; + } bam_plbuf_t; - /*! - @abstract Initialize a buffer for pileup. - @param func fucntion to be called by bam_pileup_core() - @param data user provided data - @return pointer to the pileup buffer - */ + void bam_plbuf_set_mask(bam_plbuf_t *buf, int mask); + void bam_plbuf_reset(bam_plbuf_t *buf); bam_plbuf_t *bam_plbuf_init(bam_pileup_f func, void *data); - - /*! - @abstract Destroy a pileup buffer. - @param buf pointer to the pileup buffer - */ void bam_plbuf_destroy(bam_plbuf_t *buf); - - /*! - @abstract Push an alignment to the pileup buffer. - @param b alignment to be pushed - @param buf pileup buffer - @see bam_plbuf_init() - @return always 0 currently - - @discussion If all the alignments covering a particular site have - been collected, this function will call the user defined function - as is provided to bam_plbuf_init(). The coordinate of the site and - all the alignments will be transferred to the user defined - function as function parameters. - - When all the alignments are pushed to the buffer, this function - needs to be called with b equal to NULL. This will flush the - buffer. A pileup buffer can only be reused when bam_plbuf_reset() - is called. - */ int bam_plbuf_push(const bam1_t *b, bam_plbuf_t *buf); + int bam_pileup_file(bamFile fp, int mask, bam_pileup_f func, void *func_data); + struct __bam_lplbuf_t; typedef struct __bam_lplbuf_t bam_lplbuf_t; @@ -540,8 +542,10 @@ extern "C" { /*! @abstract bam_plbuf_push() equivalent with level calculated. */ int bam_lplbuf_push(const bam1_t *b, bam_lplbuf_t *buf); - /*! @abstract bam_plbuf_file() equivalent with level calculated. */ - int bam_lpileup_file(bamFile fp, int mask, bam_pileup_f func, void *func_data); + + /********************* + * BAM indexing APIs * + *********************/ struct __bam_index_t; typedef struct __bam_index_t bam_index_t; @@ -591,6 +595,10 @@ extern "C" { */ int bam_fetch(bamFile fp, const bam_index_t *idx, int tid, int beg, int end, void *data, bam_fetch_f func); + bam_iter_t bam_iter_query(const bam_index_t *idx, int tid, int beg, int end); + int bam_iter_read(bamFile fp, bam_iter_t iter, bam1_t *b); + void bam_iter_destroy(bam_iter_t iter); + /*! @abstract Parse a region in the format: "chr2:100,000-200,000". @discussion bam_header_t::hash will be initialized if empty. @@ -603,6 +611,11 @@ extern "C" { */ int bam_parse_region(bam_header_t *header, const char *str, int *ref_id, int *begin, int *end); + + /************************** + * APIs for optional tags * + **************************/ + /*! @abstract Retrieve data of a tag @param b pointer to an alignment struct @@ -622,10 +635,15 @@ extern "C" { char bam_aux2A(const uint8_t *s); char *bam_aux2Z(const uint8_t *s); + int bam_aux_del(bam1_t *b, uint8_t *s); void bam_aux_append(bam1_t *b, const char tag[2], char type, int len, uint8_t *data); - uint8_t *bam_aux_get_core(bam1_t *b, const char tag[2]); // an alias of bam_aux_get() + + /***************** + * Miscellaneous * + *****************/ + /*! @abstract Calculate the rightmost coordinate of an alignment on the reference genome. @@ -644,14 +662,6 @@ extern "C" { */ int32_t bam_cigar2qlen(const bam1_core_t *c, const uint32_t *cigar); - typedef struct { - int32_t qbeg, qend; - int32_t tbeg, tend; - int32_t cbeg, cend; - } bam_segreg_t; - - int bam_segreg(int32_t pos, const bam1_core_t *c, const uint32_t *cigar, bam_segreg_t *reg); - #ifdef __cplusplus } #endif