htslib 0.0.0 → 0.0.1

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 8b19cf8cd36bbb9ffeb34a9c92b98adba85716f57d24ec6990cc95c53c8f658d
-  data.tar.gz: e678bdbe86be9c73c8fb3e378f86236a0a6eaf36ce047368af9263753bdca439
+  metadata.gz: 5c0c00377d4485fa702cc429c3005e98d1d095842d51fa6d27346fad9891c280
+  data.tar.gz: 56e680d3f890ef1df37f2d99644c13bde250a130ca46e5d6b756ca16366fe81e
 SHA512:
-  metadata.gz: cf6a74ee14a7f0d3bbff603c1af232d7a106a0c7d70e4c8e4c7ac4270ab98866a971611a793a1b6d994fde2b5d0a400b0e73fe04a99c9b2abfa16d5e0d82a2c5
-  data.tar.gz: d24f1fa238f3ad8d541f649a03cede9580bcdedad52e81a97f7d33a44a9c3d8fbfcf162b5912fcc9f7e5c832f09979d59cf7a4bc13fc6da7aa4facd0102fdfbb
+  metadata.gz: 1a38ecbb819edb2f81e5b76250c8ef070bf5d252925fecb859a4083462a7934feefd8ac50b2628f9f3912915982354bd61d424224302b45e9e2911321469bba6
+  data.tar.gz: 01d5ac7826b5db6b6838986d05801b69c76320bbb4a89029cf21a6f01031c9bcc23c287526fbda02c8a2c0a32453e53ed6d689373754680c584f969214c58f6f

data/README.md

@@ -1,8 +1,11 @@
 # HTSlib
 
 [![Gem Version](https://badge.fury.io/rb/htslib.svg)](https://badge.fury.io/rb/htslib)
-![CI](https://github.com/kojix2/ruby-htslib/workflows/CI/badge.svg?branch=master)
+![CI](https://github.com/kojix2/ruby-htslib/workflows/CI/badge.svg)
 [![The MIT License](https://img.shields.io/badge/license-MIT-blue.svg)](LICENSE.txt)
+[![DOI](https://zenodo.org/badge/247078205.svg)](https://zenodo.org/badge/latestdoi/247078205)
+
+:dna: [HTSlib](https://github.com/samtools/htslib) - high-throughput sequencing data manipulation - for Ruby
 
 :apple: Feel free to fork it out if you can develop it! 
 
@@ -26,6 +29,8 @@ export HTSLIBDIR="/your/path/to/htslib"
 
 ## Usage
 
+HTS::FFI - Low-level API 
+
 ```ruby
 require 'htslib'
 
@@ -35,13 +40,36 @@ p b[:category]
 p b[:format]
 ```
 
+A high-level API based on [hts-python](https://github.com/quinlan-lab/hts-python) is under development.
+
+## Documentation
+
+* [RubyDoc.info - HTSlib](https://rdoc.info/gems/htslib)
+
 ## Development
 
+To get started with development
+
+```sh
+git clone --recurse-submodules https://github.com/kojix2/ruby-htslib
+cd ruby-htslib
+bundle install
+bundle exec rake htslib:compile
+bundle exec rake spec
+```
+
 ## Contributing
 
-Bug reports and pull requests are welcome on GitHub at https://github.com/kojix2/ruby-htslib.
+* [Report bugs](https://github.com/kojix2/ruby-htslib/issues)
+* Fix bugs and [submit pull requests](https://github.com/kojix2/ruby-htslib/pulls)
+* Write, clarify, or fix documentation
+* Suggest or add new features
+
+## Links
 
+* [samtools/hts-spec](https://github.com/samtools/hts-specs)
+* [c2ffi](https://github.com/rpav/c2ffi)
 
 ## License
 
-The gem is available as open source under the terms of the [MIT License](https://opensource.org/licenses/MIT).
+[MIT License](https://opensource.org/licenses/MIT).

data/lib/hts/bam.rb

@@ -1,87 +1,79 @@
 # frozen_string_literal: true
 
-# Create a skeleton using hts-python as a reference.
+# Based on hts-python
 # https://github.com/quinlan-lab/hts-python
 
-class BamHeader
-  def initialize; end
-
-  def seqs; end
-end
-
-class Cigar
-  def initialize; end
-
-  def to_s; end
-
-  def inspect; end
-end
-
-class Alignment
-  def initialize; end
-
-  def self.rom_sam_str; end
-
-  def tags; end
-
-  def qname; end
-
-  def qname=; end
-
-  def rnext; end
-
-  def pnext; end
-
-  def rname; end
-
-  def strand; end
-
-  def base_qualities; end
-
-  def pos; end
-
-  def pos=; end
-
-  def isize; end
-
-  def mapping_quality; end
-
-  def cigar; end
-
-  def qlen; end
-
-  def rlen; end
-
-  def seqs; end
-
-  def flag_str; end
-
-  def flag; end
-
-  # def eql?
-  # def hash
-
-  def inspect; end
-
-  def to_s; end
-end
-
-class Bam
-  def initialize; end
-
-  def self.header_from_fasta; end
-
-  def inspect; end
-
-  def write; end
-
-  def close; end
-
-  def flush; end
-
-  def to_s; end
-
-  def each; end
-
-  # def call
+require_relative 'bam/header'
+require_relative 'bam/cigar'
+require_relative 'bam/alignment'
+
+module HTS
+  class Bam
+    include Enumerable
+    attr_reader :fname, :mode, :header, :htf
+
+    def initialize(fname, mode = 'r', create_index: nil, header: nil, fasta: nil)
+      @fname = File.expand_path(fname)
+      File.exist?(@fname) || raise("No such SAM/BAM file - #{@fname}")
+
+      @mode = mode
+      @htf = FFI.hts_open(@fname, mode)
+
+      if mode[0] == 'r'
+        @idx = FFI.sam_index_load(@htf, @fname)
+        if (@idx.null? && create_index.nil?) || create_index
+          FFI.sam_index_build(fname, -1)
+          @idx = FFI.sam_index_load(@htf, @fname)
+          warn 'NO querying'
+        end
+        @header = Bam::Header.new(FFI.sam_hdr_read(@htf))
+        @b = FFI.bam_init1
+
+      else
+        # FIXME
+        raise 'not implemented yet.'
+
+      end
+    end
+
+    def self.header_from_fasta; end
+
+    def write(alns)
+      alns.each do
+        FFI.sam_write1(@htf, @header, alns.b) > 0 || raise
+      end
+    end
+
+    # Close the current file.
+    def close
+      FFI.hts_close(@htf)
+    end
+
+    # Flush the current file.
+    def flush
+      raise
+      # FFI.bgzf_flush(@htf.fp.bgzf)
+    end
+
+    def each(&block)
+      # Each does not always start at the beginning of the file.
+      # This is the common behavior of IO objects in Ruby.
+      # This may change in the future.
+      block.call(Alignment.new(@b, @header.h)) while FFI.sam_read1(@htf, @header.h, @b) > 0
+    end
+
+    # query [WIP]
+    def query(region)
+      qiter = FFI.sam_itr_querys(@idx, @header.h, region)
+      begin
+        slen = FFI.sam_itr_next(@htf, qiter, @b)
+        while slen > 0
+          yield Alignment.new(@b, @header.h)
+          slen = FFI.sam_itr_next(@htf, qiter, @b)
+        end
+      ensure
+        FFI.hts_itr_destroy(qiter)
+      end
+    end
+  end
 end

data/lib/hts/bam/alignment.rb

@@ -0,0 +1,156 @@
+# frozen_string_literal: true
+
+# Based on hts-python
+# https://github.com/quinlan-lab/hts-python
+
+module HTS
+  # A cigar object usually created from `Alignment`.
+  class Bam
+    class Alignment
+      def initialize(bam1_t, bam_hdr_t)
+        @b = bam1_t
+        @h = bam_hdr_t
+      end
+
+      # def initialize_copy
+      #   super
+      # end
+
+      def self.rom_sam_str; end
+
+      def tags; end
+
+      # Read (query) name.
+      def qname
+        FFI.bam_get_qname(@b).read_string
+      end
+
+      # Set (query) name.
+      # def qname=(name)
+      #   raise 'Not Implemented'
+      # end
+
+      # returns the chromosome of the mate or '' if not mapped.
+      def mate_chrom
+        tid = @b[:core][:mtid]
+        return '' if tid == -1
+
+        FFI.sam_hdr_tid2name(@h, tid)
+      end
+
+      # returns the tid of the mate or -1 if not mapped.
+      def mate_tid
+        @b[:core][:mtid]
+      end
+
+      # returns the tid of the alignment or -1 if not mapped.
+      def tid
+        @b[:core][:tid]
+      end
+
+      # mate position
+      def mate_pos
+        @b[:core][:mpos]
+      end
+
+      # returns 0-based start position.
+      def start
+        @b[:core][:pos]
+      end
+
+      # returns end position of the read.
+      def stop
+        FFI.bam_endpos @b
+      end
+
+      # returns the chromosome or '' if not mapped.
+      def chrom
+        tid = @b[:core][:tid]
+        return '' if tid == -1
+
+        FFI.sam_hdr_tid2name(@h, tid)
+      end
+
+      def strand
+        FFI.bam_is_rev(@b) ? '-' : '+'
+      end
+
+      # def start=(v)
+      #   raise 'Not Implemented'
+      # end
+
+      # insert size
+      def isize
+        @b[:core][:isize]
+      end
+
+      # mapping quality
+      def mapping_quality
+        @b[:core][:qual]
+      end
+
+      # returns a `Cigar` object.
+      def cigar
+        Cigar.new(FFI.bam_get_cigar(@b), @b[:core][:n_cigar])
+      end
+
+      def qlen
+        FFI.bam_cigar2qlen(
+          @b[:core][:n_cigar],
+          FFI.bam_get_cigar(@b)
+        )
+      end
+
+      def rlen
+        FFI.bam_cigar2rlen(
+          @b[:core][:n_cigar],
+          FFI.bam_get_cigar(@b)
+        )
+      end
+
+      # return the read sequence
+      def sequence
+        seq_nt16_str = '=ACMGRSVTWYHKDBN'
+        r = FFI.bam_get_seq(@b)
+        Array.new(@b[:core][:l_qseq]) do |i|
+          seq_nt16_str[FFI.bam_seqi(r, i)]
+        end.join
+      end
+
+      def base_at(n)
+        n += @b[:core][:l_qseq] if n < 0
+        seq_nt16_str = '=ACMGRSVTWYHKDBN'
+        return '.' if (n >= @b[:core][:l_qseq]) || (n < 0) # eg. base_at(-1000)
+
+        r = FFI.bam_get_seq(@b)
+        seq_nt16_str[FFI.bam_seqi(r, n)]
+      end
+
+      def base_qualities
+        q_ptr = FFI.bam_get_qual(@b)
+        q_ptr.read_array_of_uint8(@b[:core][:l_qseq])
+      end
+
+      def base_quality_at(n)
+        n += @b[:core][:l_qseq] if n < 0 # eg. base_quality_at(-1000)
+        return 0 if (n >= @b[:core][:l_qseq]) || (n < 0)
+
+        q_ptr = FFI.bam_get_qual(@b)
+        q_ptr.get_uint8(n)
+      end
+
+      def flag_str
+        FFI.bam_flag2str(flag)
+      end
+
+      # returns a `Flag` object.
+      def flag
+        @b[:core][:flag]
+      end
+
+      # TODO:
+      # def eql?
+      # def hash
+    end
+  end
+end

data/lib/hts/bam/cigar.rb

@@ -0,0 +1,30 @@
+# frozen_string_literal: true
+
+# Based on hts-python
+# https://github.com/quinlan-lab/hts-python
+
+module HTS
+  class Bam
+    class Cigar
+      include Enumerable
+      OPS = 'MIDNSHP=XB'
+
+      def initialize(cigar, n_cigar)
+        @c = cigar
+        @n_cigar = n_cigar
+      end
+
+      def to_s
+        to_a.flatten.join
+      end
+
+      def each
+        @n_cigar.times do |i|
+          c = @c[i].read_uint32
+          yield [FFI.bam_cigar_oplen(c),
+                 FFI.bam_cigar_opchr(c)]
+        end
+      end
+    end
+  end
+end

data/lib/hts/bam/header.rb

@@ -0,0 +1,27 @@
+# frozen_string_literal: true
+
+# Based on hts-python
+# https://github.com/quinlan-lab/hts-python
+
+module HTS
+  class Bam
+    class Header
+      attr_reader :h
+
+      def initialize(h)
+        @h = h
+      end
+
+      # FIXME: better name?
+      def seqs
+        Array.new(@h[:n_targets]) do |i|
+          FFI.sam_hdr_tid2name(@h, i)
+        end
+      end
+
+      def text
+        FFI.sam_hdr_str(@h)
+      end
+    end
+  end
+end

data/lib/hts/fai.rb

@@ -1,16 +1,18 @@
 # frozen_string_literal: true
 
-# Create a skeleton using hts-python as a reference.
+# Based on hts-python
 # https://github.com/quinlan-lab/hts-python
 
-class Fai
-  def initialize; end
+module HTS
+  class Fai
+    def initialize; end
 
-  # def call
+    # def call
 
-  def nseqs; end
+    def nseqs; end
 
-  def include?; end
+    def include?; end
 
-  # __iter__
+    # __iter__
+  end
 end

data/lib/hts/ffi.rb

@@ -7,7 +7,7 @@ module HTS
     begin
       ffi_lib HTS.ffi_lib
     rescue LoadError => e
-      raise LoadError, "Could not find HTSlib.#{FFI::Platform::LIBSUFFIX}"
+      raise LoadError, "#{e}\nCould not find #{HTS.ffi_lib}"
     end
 
     def self.attach_function(*)
@@ -18,8 +18,19 @@ module HTS
   end
 end
 
-require_relative 'ffi/struct'
-require_relative 'ffi_constants'
+module FFI
+  class Struct
+    def self.union_layout(*args)
+      Class.new(FFI::Union) { layout(*args) }
+    end
+
+    def self.struct_layout(*args)
+      Class.new(FFI::Struct) { layout(*args) }
+    end
+  end
+end
+
+require_relative 'ffi/constants'
 
 # alphabetical order
 require_relative 'ffi/bgzf'

data/lib/hts/{ffi_constants.rb → ffi/constants.rb}

@@ -141,7 +141,8 @@ module HTS
         )
     end
 
-    class SamHdr < ::FFI::Struct # HtsFile
+    # HtsFile
+    class SamHdr < ::FFI::Struct
       layout \
         :n_targets,      :int32,
         :ignore_sam_err, :int32,
@@ -225,6 +226,7 @@ module HTS
           :a,            :pointer
         )
     end
+
     class Bam1Core < ::FFI::Struct
       layout \
         :pos,            :hts_pos_t,

data/lib/hts/ffi/sam.rb

@@ -12,8 +12,8 @@ module HTS
         c >> BAM_CIGAR_SHIFT
       end
 
-      def bam_cigar_opchar(c)
-        _BAM_CIGAR_STR_PADDED[bam_cigar_op(c)]
+      def bam_cigar_opchr(c)
+        (BAM_CIGAR_STR + '??????')[bam_cigar_op(c)]
       end
 
       def bam_cigar_gen(l, o)
@@ -60,7 +60,9 @@ module HTS
         b[:l_data] - (b[:core][:n_cigar] << 2) - b[:core][:l_qname] - b[:core][:l_qseq] - ((b[:core][:l_qseq] + 1) >> 1)
       end
 
-      # def bam_seqi(s, i)
+      def bam_seqi(s, i)
+        s[(i) >> 1].read_uint8 >> ((~i & 1) << 2) & 0xf
+      end
 
       # def bam_set_seqi(s, i, b)
     end
@@ -405,6 +407,14 @@ module HTS
       [HtsIdx, SamHdr, :pointer, :uint],
       HtsItr.by_ref
 
+    # Get the next read from a SAM/BAM/CRAM iterator
+    def self.sam_itr_next(htsfp, itr, r)
+      # FIXME: check if htsfp is compressed BGZF
+      hts_log_error('Null iterator') if itr.null?
+      # FIXME: check multi
+      hts_itr_next(htsfp[:fp][:bgzf], itr, r, htsfp)
+    end
+
     attach_function \
       :sam_parse_region,
       [SamHdr, :string, :pointer, :pointer, :pointer, :int],

data/lib/hts/tbx.rb

@@ -1,14 +1,16 @@
 # frozen_string_literal: true
 
-# Create a skeleton using hts-python as a reference.
+# Based on hts-python
 # https://github.com/quinlan-lab/hts-python
 
-class Tbx
-  def initialize; end
+module HTS
+  class Tbx
+    def initialize; end
 
-  def build; end
+    def build; end
 
-  def sequences; end
+    def sequences; end
 
-  # def __call__\
+    # def __call__\
+  end
 end

data/lib/hts/vcf.rb

@@ -1,30 +1,32 @@
 # frozen_string_literal: true
 
-# Create a skeleton using hts-python as a reference.
+# Based on hts-python
 # https://github.com/quinlan-lab/hts-python
 
-class VCF
-  def initialize; end
+module HTS
+  class VCF
+    def initialize; end
 
-  def inspect; end
+    def inspect; end
 
-  def each; end
+    def each; end
 
-  def seq; end
+    def seq; end
 
-  def n_samples; end
-end
+    def n_samples; end
+  end
 
-class Variant
-  def initialize; end
+  class Variant
+    def initialize; end
 
-  def inspect; end
+    def inspect; end
 
-  def formats; end
+    def formats; end
 
-  def genotypes; end
-end
+    def genotypes; end
+  end
 
-class Format
-  def initialize; end
+  class Format
+    def initialize; end
+  end
 end

data/lib/hts/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module HTS
-  VERSION = '0.0.0'
+  VERSION = '0.0.1'
 end

data/lib/htslib.rb

@@ -10,9 +10,21 @@ module HTS
   class << self
     attr_accessor :ffi_lib
   end
-  self.ffi_lib = File.expand_path("libhts.#{FFI::Platform::LIBSUFFIX}", ENV['HTSLIBDIR'])
+
+  suffix = ::FFI::Platform::LIBSUFFIX
+
+  self.ffi_lib = if ENV['HTSLIBDIR']
+                   File.expand_path("libhts.#{suffix}", ENV['HTSLIBDIR'])
+                 else
+                   File.expand_path("../vendor/libhts.#{suffix}", __dir__)
+                 end
   autoload :FFI, 'hts/ffi'
 end
 
 # alias
 HTSlib = HTS
+
+require_relative 'hts/bam'
+require_relative 'hts/fai'
+require_relative 'hts/tbx'
+require_relative 'hts/vcf'

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: htslib
 version: !ruby/object:Gem::Version
-  version: 0.0.0
+  version: 0.0.1
 platform: ruby
 authors:
 - kojix2
-autorequire: 
+autorequire:
 bindir: bin
 cert_chain: []
-date: 2020-10-21 00:00:00.000000000 Z
+date: 2021-01-25 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: ffi
@@ -38,6 +38,20 @@ dependencies:
     - - ">="
       - !ruby/object:Gem::Version
         version: '0'
+- !ruby/object:Gem::Dependency
+  name: minitest
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
 - !ruby/object:Gem::Dependency
   name: pry
   requirement: !ruby/object:Gem::Requirement
@@ -67,7 +81,7 @@ dependencies:
       - !ruby/object:Gem::Version
         version: '0'
 - !ruby/object:Gem::Dependency
-  name: rspec
+  name: rubocop
   requirement: !ruby/object:Gem::Requirement
     requirements:
     - - ">="
@@ -81,7 +95,7 @@ dependencies:
       - !ruby/object:Gem::Version
         version: '0'
 - !ruby/object:Gem::Dependency
-  name: rubocop
+  name: simplecov
   requirement: !ruby/object:Gem::Requirement
     requirements:
     - - ">="
@@ -94,7 +108,7 @@ dependencies:
     - - ">="
       - !ruby/object:Gem::Version
         version: '0'
-description: 
+description:
 email:
 - 2xijok@gmail.com
 executables: []
@@ -104,18 +118,20 @@ files:
 - LICENSE.txt
 - README.md
 - lib/hts/bam.rb
+- lib/hts/bam/alignment.rb
+- lib/hts/bam/cigar.rb
+- lib/hts/bam/header.rb
 - lib/hts/fai.rb
 - lib/hts/ffi.rb
 - lib/hts/ffi/bgzf.rb
+- lib/hts/ffi/constants.rb
 - lib/hts/ffi/faidx.rb
 - lib/hts/ffi/hfile.rb
 - lib/hts/ffi/hts.rb
 - lib/hts/ffi/kfunc.rb
 - lib/hts/ffi/sam.rb
-- lib/hts/ffi/struct.rb
 - lib/hts/ffi/tbx.rb
 - lib/hts/ffi/vcf.rb
-- lib/hts/ffi_constants.rb
 - lib/hts/tbx.rb
 - lib/hts/vcf.rb
 - lib/hts/version.rb
@@ -124,7 +140,7 @@ homepage: https://github.com/kojix2/ruby-htslib
 licenses:
 - MIT
 metadata: {}
-post_install_message: 
+post_install_message:
 rdoc_options: []
 require_paths:
 - lib
@@ -139,8 +155,8 @@ required_rubygems_version: !ruby/object:Gem::Requirement
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
-rubygems_version: 3.1.4
-signing_key: 
+rubygems_version: 3.0.3
+signing_key:
 specification_version: 4
 summary: HTSlib bindings for Ruby
 test_files: []

data/lib/hts/ffi/struct.rb

@@ -1,14 +0,0 @@
-# frozen_string_literal: true
-
-# This should be removed if you get the better way...
-module FFI
-  class Struct
-    def self.union_layout(*args)
-      Class.new(FFI::Union) { layout(*args) }
-    end
-
-    def self.struct_layout(*args)
-      Class.new(FFI::Struct) { layout(*args) }
-    end
-  end
-end