minimap2/python/cmappy.pxd

from libc.stdint cimport int8_t, uint8_t, int32_t, int64_t, uint32_t, uint64_t

cdef extern from "minimap.h":
	#
	# Options
	#
	ctypedef struct mm_idxopt_t:
		short k, w, flag, bucket_bits
		int mini_batch_size
		uint64_t batch_size

	ctypedef struct mm_mapopt_t:
		int seed
		int sdust_thres
		int flag
		int bw
		int max_gap, max_gap_ref
		int max_frag_len
		int max_chain_skip
		int min_cnt
		int min_chain_score
		float mask_level
		float pri_ratio
		int best_n
		int max_join_long, max_join_short
		int min_join_flank_sc
		float min_join_flank_ratio;
		int a, b, q, e, q2, e2
		int sc_ambi
		int noncan
		int zdrop, zdrop_inv
		int end_bonus
		int min_dp_max
		int min_ksw_len
		int anchor_ext_len, anchor_ext_shift
		float max_clip_ratio
		int pe_ori, pe_bonus
		float mid_occ_frac
		int32_t min_mid_occ
		int32_t mid_occ
		int32_t max_occ
		int mini_batch_size

	int mm_set_opt(char *preset, mm_idxopt_t *io, mm_mapopt_t *mo)
	int mm_verbose

	#
	# Indexing
	#
	ctypedef struct mm_idx_seq_t:
		char *name
		uint64_t offset
		uint32_t len

	ctypedef struct mm_idx_bucket_t:
		pass

	ctypedef struct mm_idx_t:
		int32_t b, w, k, flag
		uint32_t n_seq
		mm_idx_seq_t *seq
		uint32_t *S
		mm_idx_bucket_t *B
		void *km
		void *h

	ctypedef struct mm_idx_reader_t:
		pass

	mm_idx_reader_t *mm_idx_reader_open(const char *fn, const mm_idxopt_t *opt, const char *fn_out)
	mm_idx_t *mm_idx_reader_read(mm_idx_reader_t *r, int n_threads)
	void mm_idx_reader_close(mm_idx_reader_t *r)
	void mm_idx_destroy(mm_idx_t *mi)
	void mm_mapopt_update(mm_mapopt_t *opt, const mm_idx_t *mi)

	int mm_idx_index_name(mm_idx_t *mi)

	#
	# Mapping (key struct defined in cmappy.h below)
	#
	ctypedef struct mm_reg1_t:
		pass

	ctypedef struct mm_tbuf_t:
		pass

	mm_tbuf_t *mm_tbuf_init()
	void mm_tbuf_destroy(mm_tbuf_t *b)

#
# Helper header (because it is hard to expose mm_reg1_t with Cython)
#
cdef extern from "cmappy.h":
	ctypedef struct mm_hitpy_t:
		const char *ctg
		int32_t ctg_start, ctg_end
		int32_t qry_start, qry_end
		int32_t blen, mlen, NM, ctg_len
		uint8_t mapq, is_primary
		int8_t strand, trans_strand
		int32_t seg_id
		int32_t n_cigar32
		uint32_t *cigar32

	void mm_reg2hitpy(const mm_idx_t *mi, mm_reg1_t *r, mm_hitpy_t *h)
	void mm_free_reg1(mm_reg1_t *r)
	mm_reg1_t *mm_map_aux(const mm_idx_t *mi, const char *seq1, const char *seq2, int *n_regs, mm_tbuf_t *b, const mm_mapopt_t *opt)
	char *mappy_fetch_seq(const mm_idx_t *mi, const char *name, int st, int en, int *l)

	ctypedef struct kstring_t:
		unsigned l, m
		char *s

	ctypedef struct kstream_t:
		pass

	ctypedef struct kseq_t:
		kstring_t name, comment, seq, qual
		int last_char
		kstream_t *f

	kseq_t *mm_fastx_open(const char *fn)
	void mm_fastx_close(kseq_t *ks)
	int kseq_read(kseq_t *seq)

	char *mappy_revcomp(int l, const uint8_t *seq)
	int mm_verbose_level(int v)
	void mm_reset_timer()
fixed a bug on rev strand; added example 2017-09-17 05:51:13 +08:00			`from libc.stdint cimport int8_t, uint8_t, int32_t, int64_t, uint32_t, uint64_t`
first python version 2017-09-16 20:44:47 +08:00
			`cdef extern from "minimap.h":`
			`#`
			`# Options`
			`#`
			`ctypedef struct mm_idxopt_t:`
r573: prepare to generalize index 2017-11-12 08:54:06 +08:00			`short k, w, flag, bucket_bits`
first python version 2017-09-16 20:44:47 +08:00			`int mini_batch_size`
			`uint64_t batch_size`

			`ctypedef struct mm_mapopt_t:`
updated the mappy 2017-10-06 05:10:07 +08:00			`int seed`
first python version 2017-09-16 20:44:47 +08:00			`int sdust_thres`
			`int flag`
			`int bw`
			`int max_gap, max_gap_ref`
updated mappy and example.c 2017-10-16 23:15:07 +08:00			`int max_frag_len`
first python version 2017-09-16 20:44:47 +08:00			`int max_chain_skip`
			`int min_cnt`
			`int min_chain_score`
			`float mask_level`
			`float pri_ratio`
			`int best_n`
			`int max_join_long, max_join_short`
			`int min_join_flank_sc`
r763: fine control long join flank len (#128) 2018-03-30 02:16:58 +08:00			`float min_join_flank_ratio;`
first python version 2017-09-16 20:44:47 +08:00			`int a, b, q, e, q2, e2`
r770: matching N bases not working properly (#155) 2018-05-01 07:55:23 +08:00			`int sc_ambi`
first python version 2017-09-16 20:44:47 +08:00			`int noncan`
r708: implemented double Z-drop thresholds (#112) When aligning long reads, we would prefer to align through low-quality regions. This requires a large Z-drop threshold. However, to find small inversions, we need to use a small Z-drop. This commit address this conflict with two Z-drop thresholds. When Z-drop exceeds the smaller threshold, we perform a local alignment to check if there is a potential inversion. If there is one, we break the alignment; otherwise we break the alignment only if Z-drop excess the larger threshold. This commit also fixes a bug that reported wrong coordinates when the inversion is on the forward strand (#112). 2018-02-15 06:10:02 +08:00			`int zdrop, zdrop_inv`
updated mappy and example.c 2017-10-16 23:15:07 +08:00			`int end_bonus`
first python version 2017-09-16 20:44:47 +08:00			`int min_dp_max`
			`int min_ksw_len`
Release minimap2-2.6 (r623) 2017-12-12 23:39:20 +08:00			`int anchor_ext_len, anchor_ext_shift`
r703: added --max-clip-ratio still testing the option 2018-02-13 02:29:18 +08:00			`float max_clip_ratio`
updated the mappy 2017-10-06 05:10:07 +08:00			`int pe_ori, pe_bonus`
first python version 2017-09-16 20:44:47 +08:00			`float mid_occ_frac`
r741: added --min-occ-floor to improve #107 2018-03-13 02:32:27 +08:00			`int32_t min_mid_occ`
first python version 2017-09-16 20:44:47 +08:00			`int32_t mid_occ`
updated the mappy 2017-10-06 05:10:07 +08:00			`int32_t max_occ`
first python version 2017-09-16 20:44:47 +08:00			`int mini_batch_size`

			`int mm_set_opt(char preset, mm_idxopt_t io, mm_mapopt_t *mo)`
exposed timer and verbose-level to mappy 2017-09-18 03:21:36 +08:00			`int mm_verbose`
first python version 2017-09-16 20:44:47 +08:00
			`#`
			`# Indexing`
			`#`
			`ctypedef struct mm_idx_seq_t:`
			`char *name`
			`uint64_t offset`
			`uint32_t len`

			`ctypedef struct mm_idx_bucket_t:`
			`pass`

			`ctypedef struct mm_idx_t:`
r573: prepare to generalize index 2017-11-12 08:54:06 +08:00			`int32_t b, w, k, flag`
first python version 2017-09-16 20:44:47 +08:00			`uint32_t n_seq`
			`mm_idx_seq_t *seq`
			`uint32_t *S`
			`mm_idx_bucket_t *B`
fixed two mappy warnings 2018-03-27 03:12:54 +08:00			`void *km`
			`void *h`
first python version 2017-09-16 20:44:47 +08:00
			`ctypedef struct mm_idx_reader_t:`
			`pass`

			`mm_idx_reader_t mm_idx_reader_open(const char fn, const mm_idxopt_t opt, const char fn_out)`
			`mm_idx_t mm_idx_reader_read(mm_idx_reader_t r, int n_threads)`
			`void mm_idx_reader_close(mm_idx_reader_t *r)`
			`void mm_idx_destroy(mm_idx_t *mi)`
			`void mm_mapopt_update(mm_mapopt_t opt, const mm_idx_t mi)`

r718: retrieve sequence from the index 2018-02-23 23:18:26 +08:00			`int mm_idx_index_name(mm_idx_t *mi)`

first python version 2017-09-16 20:44:47 +08:00			`#`
this is embarrassing: rename again to mappy 2017-09-17 12:05:30 +08:00			`# Mapping (key struct defined in cmappy.h below)`
first python version 2017-09-16 20:44:47 +08:00			`#`
			`ctypedef struct mm_reg1_t:`
			`pass`

			`ctypedef struct mm_tbuf_t:`
			`pass`

			`mm_tbuf_t *mm_tbuf_init()`
			`void mm_tbuf_destroy(mm_tbuf_t *b)`

minor 2017-09-16 21:30:00 +08:00			`#`
exposed fasta/q reader to mappy 2017-09-18 02:41:59 +08:00			`# Helper header (because it is hard to expose mm_reg1_t with Cython)`
minor 2017-09-16 21:30:00 +08:00			`#`
this is embarrassing: rename again to mappy 2017-09-17 12:05:30 +08:00			`cdef extern from "cmappy.h":`
first python version 2017-09-16 20:44:47 +08:00			`ctypedef struct mm_hitpy_t:`
fixed a bug on rev strand; added example 2017-09-17 05:51:13 +08:00			`const char *ctg`
			`int32_t ctg_start, ctg_end`
			`int32_t qry_start, qry_end`
updated mappy and example.c 2017-10-16 23:15:07 +08:00			`int32_t blen, mlen, NM, ctg_len`
fixed a bug on rev strand; added example 2017-09-17 05:51:13 +08:00			`uint8_t mapq, is_primary`
			`int8_t strand, trans_strand`
r670: added PE support to mappy and minor code cleanup 2018-02-01 00:33:08 +08:00			`int32_t seg_id`
fixed a bug on rev strand; added example 2017-09-17 05:51:13 +08:00			`int32_t n_cigar32`
			`uint32_t *cigar32`
load python README.rst into setup.py 2017-09-17 10:43:52 +08:00
improvement to the python binding 2017-09-16 23:14:01 +08:00			`void mm_reg2hitpy(const mm_idx_t mi, mm_reg1_t r, mm_hitpy_t *h)`
first python version 2017-09-16 20:44:47 +08:00			`void mm_free_reg1(mm_reg1_t *r)`
r670: added PE support to mappy and minor code cleanup 2018-02-01 00:33:08 +08:00			`mm_reg1_t mm_map_aux(const mm_idx_t mi, const char seq1, const char seq2, int n_regs, mm_tbuf_t b, const mm_mapopt_t *opt)`
r718: retrieve sequence from the index 2018-02-23 23:18:26 +08:00			`char mappy_fetch_seq(const mm_idx_t mi, const char name, int st, int en, int l)`
exposed fasta/q reader to mappy 2017-09-18 02:41:59 +08:00
			`ctypedef struct kstring_t:`
			`unsigned l, m`
			`char *s`

			`ctypedef struct kstream_t:`
			`pass`

			`ctypedef struct kseq_t:`
			`kstring_t name, comment, seq, qual`
			`int last_char`
			`kstream_t *f`

			`kseq_t mm_fastx_open(const char fn)`
			`void mm_fastx_close(kseq_t *ks)`
			`int kseq_read(kseq_t *seq)`
exposed timer and verbose-level to mappy 2017-09-18 03:21:36 +08:00
bugfix: in-place revcomp() not working 2018-02-21 00:05:54 +08:00			`char mappy_revcomp(int l, const uint8_t seq)`
exposed timer and verbose-level to mappy 2017-09-18 03:21:36 +08:00			`int mm_verbose_level(int v)`
			`void mm_reset_timer()`