htslib 0.0.5 → 0.0.10

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 3eb6fbb228a9fa0642c55cdfa39d5189b311df191725954ae46b8dda135dc8c7
-  data.tar.gz: c90ad39aa4919cefa56e534a706bf9659bba9c6ad69eb374491092ce16ae93f6
+  metadata.gz: 2ecf4c183a1720313371e493fbf7876a08d17639ad853b1d26faeff3addb7c6e
+  data.tar.gz: cd1ed908a5dd8be184c8d9195c6fb5fabda5487d3c35e11476d9bd6d15c7dae8
 SHA512:
-  metadata.gz: 392ea7c6acb5b86e9e1c1a29f066049c54954edfd277be30541aee433a7349f9d74df09e663ec0269f30d6b87ff47d66a4a668470d3aa536e25a2058cf963546
-  data.tar.gz: 9cc235f8876b8fd4339593eb84a0f2cb8b605b9c82b624cd0f97691b4109a672f55ae90447bba17fb5efcb8257279ad8c2d87234bca390b1624c7aa8f9a7014e
+  metadata.gz: 4980914c7b92528055d2d0cf62fad6ca5edcde8ba1b9d426740538aee42fab53c7ef96ebe53adcaa9663368f024546c048b71760b68bd88e7cbb7647e200847c
+  data.tar.gz: a94b5d01702cc9873fdf6d1498658cff15c44ba8c90b08cafe514a883560c6aaa90cfbd83cede8a4ae1a5095c6f99710e76a8aacb0e0851e628c607c37719bd5

data/README.md

@@ -6,14 +6,12 @@
 [![DOI](https://zenodo.org/badge/247078205.svg)](https://zenodo.org/badge/latestdoi/247078205)
 [![Docs Stable](https://img.shields.io/badge/docs-stable-blue.svg)](https://rubydoc.info/gems/htslib)
 
-:dna: [HTSlib](https://github.com/samtools/htslib) - for Ruby
-
-Ruby-htslib is the Ruby bindings to HTSlib, a C library for processing high throughput sequencing (HTS) data. 
-It will provide APIs to read and write file formats such as SAM, BAM, VCF, and BCF.
+Ruby-htslib is the [Ruby](https://www.ruby-lang.org) bindings to [HTSlib](https://github.com/samtools/htslib), a C library for high-throughput sequencing data formats. It allows you to read and write file formats commonly used in genomics, such as [SAM, BAM, VCF, and BCF](http://samtools.github.io/hts-specs/) in the Ruby language.
 
 :apple: Feel free to fork it out if you can develop it! 
 
 :bowtie: alpha stage.
+
 ## Requirements
 
 * [Ruby](https://github.com/ruby/ruby) 2.7 or above.
@@ -38,51 +36,80 @@ export HTSLIBDIR="/your/path/to/htslib" # libhts.so
 
 ## Overview
 
-### Low level API
+### High level API
 
-`HTS::LibHTS` provides native functions. 
+Read SAM / BAM / CRAM - Sequence Alignment Map file
 
 ```ruby
 require 'htslib'
 
-a = HTS::LibHTS.hts_open("a.bam", "r")
-b = HTS::LibHTS.hts_get_format(a)
-p b[:category]
-p b[:format]
+bam = HTS::Bam.open("a.bam")
+
+bam.each do |r|
+  p name:  r.qname,
+    flag:  r.flag,
+    chr:   r.chrom,
+    pos:   r.start + 1,
+    mqual: r.mapping_quality,
+    cigar: r.cigar.to_s,
+    mchr:  r.mate_chrom,
+    mpos:  r.mate_start + 1,
+    isize: r.insert_size,
+    seq:   r.sequence,
+    qual:  r.base_qualities.map { |i| (i + 33).chr }.join,
+    tagMC: r.tag("MC")
+end
+
+bam.close
 ```
 
-Note: Managed struct is not used in ruby-htslib. You may need to free the memory by yourself.
+Read VCF / BCF - Variant Call Format file
 
-### High level API (Plan)
+```ruby
+bcf = HTS::Bcf.open("b.bcf")
+
+bcf.each do |r|
+  p chrom:  r.chrom,
+    pos:    r.pos,
+    id:     r.id,
+    qual:   r.qual.round(2),
+    ref:    r.ref,
+    alt:    r.alt,
+    filter: r.filter
+  # info:   r.info
+  # format: r.format
+end
 
-`Cram` `Bam` `Bcf` `Faidx` `Tabix`
+bcf.close
+```
 
-A high-level API is under development. We will change and improve the API to make it better.
+### Low level API
+
+`HTS::LibHTS` provides native　C functions. 
 
 ```ruby
 require 'htslib'
 
-bam = HTS::Bam.new("a.bam")
-
-bam.each do |r|
-  p name:  r.qname,
-    flag:  r.flag,
-    start: r.start + 1,
-    mpos:  r.mate_pos + 1,
-    mqual: r.mapping_quality,
-    seq:   r.sequence,
-    cigar: r.cigar.to_s,
-    qual:  r.base_qualities.map { |i| (i + 33).chr }.join
-end
+a = HTS::LibHTS.hts_open("a.bam", "r")
+b = HTS::LibHTS.hts_get_format(a)
+p b[:category]
+p b[:format]
 ```
 
+Note: htslib makes extensive use of macro functions for speed. you cannot use C macro functions in Ruby if they are not reimplemented in ruby-htslib. Only small number of C structs are implemented with FFI's ManagedStruct, which frees memory when Ruby's garbage collection fires. Other structs will need to be freed manually. 
+
+### Need more speed?
+
+Try Crystal. [htslib.cr](https://github.com/bio-crystal/htslib.cr) is implemented in Crystal language and provides an API compatible with ruby-htslib. Crsytal language is not as flexible as Ruby language. You can not use eval methods, and you must always be aware of the types. It is not very suitable for writing one-time scripts or experimenting with different code. However, If you have already written code in ruby-htslib, have a clear idea of the manipulations you want to do, and need to execute them many times, then by all means try to implement the command line tool using htslib.cr. The Crystal language is very fast and can perform almost as well as the Rust and C languages.
+
 ## Documentation
 
+* [API Documentation (develop branch)](https://kojix2.github.io/ruby-htslib/)
 * [RubyDoc.info - HTSlib](https://rdoc.info/gems/htslib)
 
 ## Development
 
-To get started with development
+To get started with development:
 
 ```sh
 git clone --recursive https://github.com/kojix2/ruby-htslib
@@ -92,8 +119,13 @@ bundle exec rake htslib:build
 bundle exec rake test
 ```
 
-We plan to actively use the new features of Ruby. Since the number of users is small, backward compatibility is not important.
-On the other hand, we will consider compatibility with [Crystal](https://github.com/bio-crystal/htslib.cr) to some extent.
+[GNU Autotools](https://en.wikipedia.org/wiki/GNU_Autotools) is required to compile htslib.
+
+Many macro functions are used in HTSlib. Since these macro functions cannot be called using FFI, they must be reimplemented in Ruby.
+
+* Actively use the advanced features of Ruby.
+* Remain compatibile with [htslib.cr](https://github.com/bio-crystal/htslib.cr).
+  * The most difficult part is the return value. In the Crystal language, it is convenient for a method to return only one type. In the Ruby language, on the other hand, it is more convenient to return multiple classes. For example, in the Crystal language, it is confusing that a return value can take four types: Int32, Float32, Nil, and String. In Ruby, on the other hand, it is very common and does not cause any problems.
 
 #### FFI Extensions
 
@@ -113,6 +145,16 @@ Ruby-htslib is a library under development, so even small improvements like typo
 * Suggest or add new features
 * [financial contributions](https://github.com/sponsors/kojix2)
 
+```
+Do you need commit rights to ruby-htslib repository?
+Do you want to get admin rights and take over the project?
+If so, please feel free to contact us @kojix2.
+```
+
+#### Why do you implement htslib in a language like Ruby, which is not widely used in the bioinformatics?
+
+One of the greatest joys of using a minor language like Ruby in bioinformatics is that there is nothing stopping you from reinventing the wheel. Reinventing the wheel can be fun. But with languages like Python and R, where many bioinformatics masters work, there is no chance left for beginners to create htslib bindings. Bioinformatics file formats, libraries and tools are very complex and I don't know how to understand them. So I wanted to implement the HTSLib binding myself to better understand how the pioneers of bioinformatics felt when establishing the file format and how they created their tools. And that effort is still going on today...
+
 ## Links
 
 * [samtools/hts-spec](https://github.com/samtools/hts-specs)

data/lib/hts/bam/cigar.rb

@@ -4,7 +4,7 @@
 # https://github.com/quinlan-lab/hts-python
 
 module HTS
-  class Bam
+  class Bam < Hts
     class Cigar
       include Enumerable
 
@@ -18,14 +18,17 @@ module HTS
       end
 
       def to_s
-        to_a.flatten.join
+        map { |op, len| "#{len}#{op}" }.join
       end
 
       def each
+        return to_enum(__method__) unless block_given?
+
         @n_cigar.times do |i|
           c = @pointer[i].read_uint32
-          yield [LibHTS.bam_cigar_oplen(c),
-                 LibHTS.bam_cigar_opchr(c)]
+          op =  LibHTS.bam_cigar_opchr(c)
+          len = LibHTS.bam_cigar_oplen(c)
+          yield [op, len]
         end
       end
     end

data/lib/hts/bam/flag.rb

@@ -4,7 +4,7 @@
 # https://github.com/brentp/hts-nim/blob/master/src/hts/bam/flag.nim
 
 module HTS
-  class Bam
+  class Bam < Hts
     class Flag
       def initialize(flag_value)
         raise TypeError unless flag_value.is_a? Integer
@@ -85,8 +85,12 @@ module HTS
         has_flag? LibHTS::BAM_FSUPPLEMENTARY
       end
 
-      def has_flag?(o)
-        @value[o] != 0
+      def has_flag?(m)
+        (@value & m) != 0
+      end
+
+      def to_s
+        "0x#{format('%x', @value)}\t#{@value}\t#{LibHTS.bam_flag2str(@value)}"
       end
     end
   end

data/lib/hts/bam/header.rb

@@ -4,10 +4,10 @@
 # https://github.com/quinlan-lab/hts-python
 
 module HTS
-  class Bam
+  class Bam < Hts
     class Header
-      def initialize(pointer)
-        @sam_hdr = pointer
+      def initialize(hts_file)
+        @sam_hdr = LibHTS.sam_hdr_read(hts_file)
       end
 
       def struct
@@ -18,16 +18,31 @@ module HTS
         @sam_hdr.to_ptr
       end
 
-      # FIXME: better name?
-      def seqs
-        Array.new(@sam_hdr[:n_targets]) do |i|
+      def target_count
+        @sam_hdr[:n_targets]
+      end
+
+      def target_names
+        Array.new(target_count) do |i|
           LibHTS.sam_hdr_tid2name(@sam_hdr, i)
         end
       end
 
-      def text
+      def target_lengths
+        Array.new(target_count) do |i|
+          LibHTS.sam_hdr_tid2len(@sam_hdr, i)
+        end
+      end
+
+      def to_s
         LibHTS.sam_hdr_str(@sam_hdr)
       end
+
+      private
+
+      def initialize_copy(orig)
+        @sam_hdr = LibHTS.sam_hdr_dup(orig.struct)
+      end
     end
   end
 end

data/lib/hts/bam/record.rb

@@ -4,10 +4,12 @@
 # https://github.com/quinlan-lab/hts-python
 
 module HTS
-  class Bam
+  class Bam < Hts
     class Record
       SEQ_NT16_STR = "=ACMGRSVTWYHKDBN"
 
+      attr_reader :header
+
       def initialize(bam1_t, header)
         @bam1 = bam1_t
         @header = header
@@ -21,16 +23,6 @@ module HTS
         @bam1.to_ptr
       end
 
-      attr_reader :header
-
-      # def initialize_copy
-      #   super
-      # end
-
-      def self.rom_sam_str; end
-
-      def tags; end
-
       # returns the query name.
       def qname
         LibHTS.bam_get_qname(@bam1).read_string
@@ -69,18 +61,22 @@ module HTS
 
       # returns the chromosome or '' if not mapped.
       def chrom
-        tid = @bam1[:core][:tid]
         return "" if tid == -1
 
         LibHTS.sam_hdr_tid2name(@header, tid)
       end
 
+      # returns the chromosome or '' if not mapped.
+      def contig
+        chrom
+      end
+
       # returns the chromosome of the mate or '' if not mapped.
       def mate_chrom
-        tid = @bam1[:core][:mtid]
-        return "" if tid == -1
+        mtid = mate_tid
+        return "" if mtid == -1
 
-        LibHTS.sam_hdr_tid2name(@header, tid)
+        LibHTS.sam_hdr_tid2name(@header, mtid)
       end
 
       def strand
@@ -92,7 +88,7 @@ module HTS
       # end
 
       # insert size
-      def isize
+      def insert_size
         @bam1[:core][:isize]
       end
 
@@ -168,6 +164,11 @@ module HTS
         return nil if aux.null?
 
         t = aux.read_string(1)
+
+        # A (character), B (general array),
+        # f (real number), H (hexadecimal array),
+        # i (integer), or Z (string).
+
         case t
         when "i", "I", "c", "C", "s", "S"
           LibHTS.bam_aux2i(aux)
@@ -175,11 +176,13 @@ module HTS
           LibHTS.bam_aux2f(aux)
         when "Z", "H"
           LibHTS.bam_aux2Z(aux)
-        when "A"
-          LibHTS.bam_aux2A(aux)
+        when "A" # char
+          LibHTS.bam_aux2A(aux).chr
         end
       end
 
+      # def tags; end
+
       def to_s
         kstr = LibHTS::KString.new
         raise "Failed to format bam record" if LibHTS.sam_format1(@header.struct, @bam1, kstr) == -1
@@ -187,9 +190,12 @@ module HTS
         kstr[:s]
       end
 
-      # TODO:
-      # def eql?
-      # def hash
+      private
+
+      def initialize_copy(orig)
+        @header = orig.header
+        @bam = LibHTS.bam_dup1(orig.struct)
+      end
     end
   end
 end

data/lib/hts/bam.rb

@@ -3,6 +3,9 @@
 # Based on hts-python
 # https://github.com/quinlan-lab/hts-python
 
+require_relative "../htslib"
+
+require_relative "hts"
 require_relative "bam/header"
 require_relative "bam/cigar"
 require_relative "bam/flag"
@@ -12,99 +15,141 @@ module HTS
   class Bam
     include Enumerable
 
-    attr_reader :file_path, :mode, :header
-    # HtfFile is FFI::BitStruct
-    attr_reader :htf_file
+    attr_reader :file_name, :index_path, :mode, :header
 
-    class << self
-      alias open new
-    end
+    def self.open(*args, **kw)
+      file = new(*args, **kw) # do not yield
+      return file unless block_given?
 
-    def initialize(file_path, mode = "r", create_index: nil)
-      file_path = File.expand_path(file_path)
+      begin
+        yield file
+      ensure
+        file.close
+      end
+      file
+    end
 
-      unless File.exist?(file_path)
-        message = "No such SAM/BAM file - #{file_path}"
+    def initialize(file_name, mode = "r", index: nil, fai: nil, threads: nil,
+                   create_index: false)
+      if block_given?
+        message = "HTS::Bam.new() dose not take block; Please use HTS::Bam.open() instead"
         raise message
       end
 
-      @file_path = file_path
+      # NOTE: Do not check for the existence of local files, since file_names may be remote URIs.
+
+      @file_name = file_name
       @mode      = mode
-      @htf_file  = LibHTS.hts_open(file_path, mode)
-      @header    = Bam::Header.new(LibHTS.sam_hdr_read(htf_file))
-
-      # read
-      if mode[0] == "r"
-        # load index
-        @idx = LibHTS.sam_index_load(htf_file, file_path)
-        # create index
-        if create_index || (@idx.null? && create_index.nil?)
-          warn "Create index for #{file_path}"
-          LibHTS.sam_index_build(file_path, -1)
-          @idx = LibHTS.sam_index_load(htf_file, file_path)
-        end
-      else
-        # FIXME: implement
-        raise "not implemented yet."
+      @hts_file  = LibHTS.hts_open(@file_name, mode)
+
+      raise Errno::ENOENT, "Failed to open #{@file_name}" if @hts_file.null?
+
+      if fai
+        r = LibHTS.hts_set_fai_filename(@hts_file, fai)
+        raise "Failed to load fasta index: #{fai}" if r < 0
       end
 
-      # IO like API
-      if block_given?
-        begin
-          yield self
-        ensure
-          close
-        end
+      if threads&.> 0
+        r = LibHTS.hts_set_threads(@hts_file, threads)
+        raise "Failed to set number of threads: #{threads}" if r < 0
       end
-    end
 
-    def struct
-      htf_file
+      return if @mode[0] == "w"
+
+      @header = Bam::Header.new(@hts_file)
+
+      create_index(index) if create_index
+
+      @idx = load_index(index)
+
+      @start_position = tell
     end
 
-    def to_ptr
-      htf_file.to_ptr
+    def create_index(index_name = nil)
+      if index
+        warn "Create index for #{@file_name} to #{index_name}"
+        LibHTS.sam_index_build2(@file_name, index_name, -1)
+      else
+        warn "Create index for #{@file_name} to #{index_name}"
+        LibHTS.sam_index_build(@file_name, -1)
+      end
     end
 
-    def write(alns)
-      alns.each do
-        LibHTS.sam_write1(htf_file, header, alns.b) > 0 || raise
+    def load_index(index_name = nil)
+      if index_name
+        LibHTS.sam_index_load2(@hts_file, @file_name, index_name)
+      else
+        LibHTS.sam_index_load3(@hts_file, @file_name, nil, 2) # should be 3 ? (copy remote file to local?)
       end
     end
 
+    def index_loaded?
+      !@idx.null?
+    end
+
     # Close the current file.
     def close
-      LibHTS.hts_close(htf_file)
+      LibHTS.hts_idx_destroy(@idx) if @idx&.null?
+      @idx = nil
+      LibHTS.hts_close(@hts_file)
+      @hts_file = nil
+    end
+
+    def closed?
+      @hts_file.nil? || @hts_file.null?
+    end
+
+    def write_header(header)
+      @header = header.dup
+      LibHTS.hts_set_fai_filename(@hts_file, @file_name)
+      LibHTS.sam_hdr_write(@hts_file, header)
+    end
+
+    def write(aln)
+      aln_dup = aln.dup
+      LibHTS.sam_write1(@hts_file, header, aln_dup) > 0 || raise
     end
 
-    # Flush the current file.
-    def flush
-      # LibHTS.bgzf_flush(@htf_file.fp.bgzf)
+    # Iterate over each record.
+    # Generate a new Record object each time.
+    # Slower than each.
+    def each_copy
+      return to_enum(__method__) unless block_given?
+
+      while LibHTS.sam_read1(@hts_file, header, bam1 = LibHTS.bam_init1) != -1
+        record = Record.new(bam1, header)
+        yield record
+      end
+      self
     end
 
+    # Iterate over each record.
+    # Record object is reused.
+    # Faster than each_copy.
     def each
       # 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.
       return to_enum(__method__) unless block_given?
 
-      while LibHTS.sam_read1(htf_file, header, bam1 = LibHTS.bam_init1) > 0
-        record = Record.new(bam1, header)
-        yield record
-      end
+      bam1 = LibHTS.bam_init1
+      record = Record.new(bam1, header)
+      yield record while LibHTS.sam_read1(@hts_file, header, bam1) != -1
       self
     end
 
     # query [WIP]
     def query(region)
+      raise "Index file is required to call the query method." unless index_loaded?
+
       qiter = LibHTS.sam_itr_querys(@idx, header, region)
       begin
         bam1 = LibHTS.bam_init1
-        slen = LibHTS.sam_itr_next(htf_file, qiter, bam1)
+        slen = LibHTS.sam_itr_next(@hts_file, qiter, bam1)
         while slen > 0
           yield Record.new(bam1, header)
           bam1 = LibHTS.bam_init1
-          slen = LibHTS.sam_itr_next(htf_file, qiter, bam1)
+          slen = LibHTS.sam_itr_next(@hts_file, qiter, bam1)
         end
       ensure
         LibHTS.hts_itr_destroy(qiter)

data/lib/hts/bcf/format.rb

@@ -5,17 +5,37 @@
 # TODO: Make it more like Ruby.
 
 module HTS
-  class Bcf
+  class Bcf < Hts
     class Format
       def initialize(record)
         @record = record
         @p1 = FFI::MemoryPointer.new(:pointer) # FIXME: naming
       end
 
+      # For compatibility with htslib.cr.
+      def get_int(key)
+        get(key, :int)
+      end
+
+      # For compatibility with htslib.cr.
+      def get_float(key)
+        get(key, :float)
+      end
+
+      # For compatibility with htslib.cr.
+      def get_flag(key)
+        get(key, :flag)
+      end
+
+      # For compatibility with htslib.cr.
+      def get_string(key)
+        get(key, :string)
+      end
+
       def get(key, type = nil)
         n = FFI::MemoryPointer.new(:int)
         p1 = @p1
-        h = @record.bcf.header.struct
+        h = @record.header.struct
         r = @record.struct
 
         format_values = proc do |type|
@@ -33,11 +53,13 @@ module HTS
           format_values.call(LibHTS::BCF_HT_REAL)
                        .read_array_of_float(n.read_int)
         when :flag
-          format_values.call(LibHTS::BCF_HT_FLAG)
-                       .read_int == 1
+          raise NotImplementedError, "Flag type not implemented yet."
+          # format_values.call(LibHTS::BCF_HT_FLAG)
+          #              .read_int == 1
         when :string, :str
-          format_values.call(LibHTS::BCF_HT_STR)
-                       .read_pointer.read_string
+          raise NotImplementedError, "String type not implemented yet."
+          # format_values.call(LibHTS::BCF_HT_STR)
+          #              .read_string
         end
       end