parse_fasta 1.6.2 → 1.7.0

checksums.yaml

@@ -1,7 +1,15 @@
 ---
-SHA1:
-  metadata.gz: 704dd834a4d948441636422507ca52a9fd44141b
-  data.tar.gz: 83d5170a4636337ba9dff578c0de90683f610172
+!binary "U0hBMQ==":
+  metadata.gz: !binary |-
+    YjUzNjQwMGYzNDQzNWNiYzczOTFmYmE3OGY2YWI0MmJkNjhkNmE0ZA==
+  data.tar.gz: !binary |-
+    ODFjMmIxNWE5MDc1ODE0MGUyOWRmYTg5YmFkNWJkMjE1MzUxMjg1Yg==
 SHA512:
-  metadata.gz: ec6f5c7bca2f75e2ec82d18efe31f08e72ca5e3a67f7a7c2c61f5cc5392813396246d601b95c34697dd8a3b133f9655608733ff05e36912f33a6858fe2815c37
-  data.tar.gz: ba8e6522891519acf8a22d019500cd4ec46e01fa39b954f52f3b4c2c68586ae6e431a6eb2375799ab52fed0ada667c731be0b34a54ccce732d0f06f1d8dfa7f6
+  metadata.gz: !binary |-
+    NmFmNjcwODhhM2E0NmM2NzE1OTA0NmYwOGVmZTM0NjA3Mzg4ZmUyNWUwYjNm
+    ZTRjNmE0NWI2MTc1NWM5NTMwM2I5YjEzYzQxMjcwYTJhMjdiMmVkZmJmNWVj
+    ZTU1ODY1OGE1ZGRiOGQ3ZGU1NTIyYWI0ZTJlN2I4YTBiMGIzOTg=
+  data.tar.gz: !binary |-
+    ZTM2YTBhNGU4N2E4ZDJkYTI1OWNiMTI5YjQ5Njc1ZTlmYTI4MzMxYWRmZDNk
+    N2U0Zjk1N2M1M2YzZWMxNjk0NDZlNmIzYjFjYTAzNWQ2YjllYWQxYjk2ZDAx
+    NDRiYTc3ZjFiNzIxM2VmNjQxODlkMjgxZWQ4MTRlY2M0NjRkOWQ=

data/.yardopts

@@ -0,0 +1 @@
+--markup-provider rdiscount

data/README.md

@@ -27,7 +27,7 @@ lightweight than BioRuby. And more fun! ;)
 ## Documentation ##
 
 Checkout
-[parse_fasta docs](http://rubydoc.info/gems/parse_fasta/1.6.2/frames)
+[parse_fasta docs](http://rubydoc.info/gems/parse_fasta/1.7.0/frames)
 for the full api documentation.
 
 ## Usage ##
@@ -60,8 +60,16 @@ What if you don't care if the input is a fastA or a fastQ? No problem!
 	  puts [header, seq].join "\t"
 	end
 
+Read fasta file into a hash.
+
+    seqs = FastaFile.open(ARGV[0]).to_hash
+
 ## Versions ##
 
+### 1.7 ###
+
+Add `SeqFile#to_hash`, `FastaFile#to_hash` and `FastqFile#to_hash`.
+
 ### 1.6 ###
 
 Added `SeqFile` class, which accepts either fastA or fastQ files. It

data/lib/parse_fasta/fasta_file.rb

@@ -43,6 +43,23 @@ class FastaFile < File
     super
   end
 
+  # Returns the records in the fasta file as a hash map with the
+  # headers as keys and the Sequences as values.
+  #
+  # @example Read a fastA into a hash table.
+  #   seqs = FastaFile.open('reads.fa').to_hash
+  #
+  # @return [Hash] A hash with headers as keys, sequences as the
+  #   values (Sequence objects)
+  def to_hash
+    hash = {}
+    self.each_record do |head, seq|
+      hash[head] = seq
+    end
+
+    hash
+  end
+
   # Analagous to IO#each_line, #each_record is used to go through a
   # fasta file record by record. It will accept gzipped files as well.
   #

data/lib/parse_fasta/fastq_file.rb

@@ -22,6 +22,25 @@ require 'zlib'
 # format files. Gzipped files are no problem.
 class FastqFile < File
 
+  # Returns the records in the fastq file as a hash map with the
+  # headers as keys pointing to a hash map like so
+  # { "seq1" => { head: "seq1", seq: "ACTG", desc: "", qual: "II3*"} }
+  #
+  # @example Read a fastQ into a hash table.
+  #   seqs = FastqFile.open('reads.fq.gz').to_hash
+  #
+  # @return [Hash] A hash with headers as keys, and a hash map as the
+  #   value with keys :head, :seq, :desc, :qual, for header, sequence,
+  #   description, and quality.
+  def to_hash
+    hash = {}
+    self.each_record do |head, seq, desc, qual|
+      hash[head] = { head: head, seq: seq, desc: desc, qual: qual }
+    end
+
+    hash
+  end
+
   # Analagous to IO#each_line, #each_record is used to go through a
   # fastq file record by record. It will accept gzipped files as well.
   #
@@ -33,7 +52,7 @@ class FastqFile < File
   #   FastqFile.open('reads.fq.gz').each_record do |head, seq, desc, qual|
   #     # do some fun stuff here!
   #   end
-  # 
+  #
   # @yield The header, sequence, description and quality string for
   #   each record in the fastq file to the block
   # @yieldparam header [String] The header of the fastq record without
@@ -54,8 +73,8 @@ class FastqFile < File
       f = Zlib::GzipReader.open(self)
     rescue Zlib::GzipFile::Error => e
       f = self
-    end      
-    
+    end
+
     f.each_line do |line|
       line.chomp!
 
@@ -70,10 +89,10 @@ class FastqFile < File
         quality = Quality.new(line)
         yield(header, sequence, description, quality)
       end
-      
+
       count += 1
     end
-    
+
     f.close if f.instance_of?(Zlib::GzipReader)
     return f
   end

data/lib/parse_fasta/seq_file.rb

@@ -20,6 +20,27 @@
 # depending on what the user provides. Handles, gzipped files.
 class SeqFile < File
 
+  # Returns the records in the sequence file as a hash map with the
+  # headers as keys and the Sequences as values. For a fastq file,
+  # acts the same as `FastaFile#to_hash`
+  #
+  # @example Read a fastA into a hash table.
+  #   seqs = SeqFile.open('reads.fa').to_hash
+  #
+  # @return [Hash] A hash with headers as keys, sequences as the
+  #   values (Sequence objects)
+  def to_hash
+    first_char = get_first_char(self)
+
+    if first_char == '>'
+      FastaFile.open(self).to_hash
+    elsif first_char == '@'
+      FastqFile.open(self).to_hash
+    else
+      raise ArgumentError, "Input does not look like FASTA or FASTQ"
+    end
+  end
+
   # Analagous to IO#each_line, #each_record will go through a fastA or
   # fastQ file record by record.
   #
@@ -54,7 +75,7 @@ class SeqFile < File
   # @yieldparam sequence [Sequence] The sequence of the record.
   def each_record
     first_char = get_first_char(self)
-    
+
     if first_char == '>'
       FastaFile.open(self).each_record do |header, sequence|
         yield(header, sequence)
@@ -65,7 +86,7 @@ class SeqFile < File
       end
     else
       raise ArgumentError, "Input does not look like FASTA or FASTQ"
-    end      
+    end
   end
 
   private
@@ -75,7 +96,7 @@ class SeqFile < File
       handle = Zlib::GzipReader.open(f)
     rescue Zlib::GzipFile::Error => e
       handle = f
-    end      
+    end
 
     handle.each_line.peek[0]
   end

data/lib/parse_fasta/version.rb

@@ -17,5 +17,5 @@
 # along with parse_fasta.  If not, see <http://www.gnu.org/licenses/>.
 
 module ParseFasta
-  VERSION = "1.6.2"
+  VERSION = "1.7.0"
 end

data/parse_fasta.gemspec

@@ -28,5 +28,6 @@ Gem::Specification.new do |spec|
   spec.add_development_dependency "rspec", "~> 2.14"
   spec.add_development_dependency "bio", "~> 1.4"
   spec.add_development_dependency "yard", "~> 0.8"
+  spec.add_development_dependency "rdiscount"
   spec.add_development_dependency "coveralls", "~> 0.7"
 end

data/spec/lib/fasta_file_spec.rb

@@ -44,6 +44,22 @@ describe FastaFile do
     end
   end
 
+  describe "#to_hash" do
+    let(:records) { Helpers::RECORDS_MAP }
+    let(:fname) { "#{File.dirname(__FILE__)}/../../test_files/test.fa.gz" }
+    let(:fasta) { FastaFile.open(fname) }
+
+    it "reads the records into a hash: header as key and seq as val" do
+      expect(fasta.to_hash).to eq records
+    end
+
+    it "passes the values as Sequence objects" do
+      expect(
+        fasta.to_hash.values.all? { |val| val.instance_of? Sequence }
+      ).to eq true
+    end
+  end
+
   describe "#each_record" do
     let(:records) { Helpers::RECORDS }
 

data/spec/lib/fastq_file_spec.rb

@@ -43,9 +43,40 @@ describe FastqFile do
         expect(qual).to be_an_instance_of Quality
       end
     end
+  end
+
+  describe "#to_hash" do
+    let(:records) {
+      { "seq1" => { head: "seq1",
+                    seq: "AACCTTGG",
+                    desc: "",
+                    qual: ")#3gTqN8" },
+        "seq2 apples" => { head: "seq2 apples",
+                           seq: "ACTG",
+                           desc: "seq2 apples",
+                           qual: "*ujM" }
+      }
+    }
+    let(:fname) { "#{File.dirname(__FILE__)}/../../test_files/test.fq.gz" }
+    let(:fastq) { FastqFile.open(fname) }
+
+    it "reads the records into a hash: header as key and seq as val" do
+      expect(fastq.to_hash).to eq records
+    end
+
+    it "passes the seqs as Sequence objects" do
+      expect(
+        fastq.to_hash.values.all? { |val| val[:seq].instance_of? Sequence }
+      ).to eq true
+    end
 
+    it "passes the quals as Quality objects" do
+      expect(
+        fastq.to_hash.values.all? { |val| val[:qual].instance_of? Quality }
+      ).to eq true
+    end
   end
-  
+
   context "with a 4 line per record fastq file" do
     describe "#each_record" do
       context "with a gzipped file" do
@@ -80,6 +111,5 @@ describe FastqFile do
         end
       end
     end
-  end        
+  end
 end
-

data/spec/lib/seq_file_spec.rb

@@ -19,6 +19,57 @@
 require 'spec_helper'
 
 describe SeqFile do
+
+  describe "#to_hash" do
+    context "when input is a fasta file" do
+      let(:records) { Helpers::RECORDS_MAP }
+      let(:fname) { "#{File.dirname(__FILE__)}/../../test_files/test.fa.gz" }
+      let(:fasta) { SeqFile.open(fname) }
+
+      it "reads the records into a hash: header as key and seq as val" do
+        expect(fasta.to_hash).to eq records
+      end
+
+      it "passes the values as Sequence objects" do
+        expect(
+          fasta.to_hash.values.all? { |val| val.instance_of? Sequence }
+        ).to eq true
+      end
+    end
+
+    context "when input is a fastq file" do
+      let(:records) {
+        { "seq1" => { head: "seq1",
+                      seq: "AACCTTGG",
+                      desc: "",
+                      qual: ")#3gTqN8" },
+          "seq2 apples" => { head: "seq2 apples",
+                             seq: "ACTG",
+                             desc: "seq2 apples",
+                             qual: "*ujM" }
+        }
+      }
+      let(:fname) { "#{File.dirname(__FILE__)}/../../test_files/test.fq.gz" }
+      let(:fastq) { SeqFile.open(fname) }
+
+      it "reads the records into a hash: header as key and seq as val" do
+        expect(fastq.to_hash).to eq records
+      end
+
+      it "passes the seqs as Sequence objects" do
+        expect(
+          fastq.to_hash.values.all? { |val| val[:seq].instance_of? Sequence }
+        ).to eq true
+      end
+
+      it "passes the quals as Quality objects" do
+        expect(
+          fastq.to_hash.values.all? { |val| val[:qual].instance_of? Quality }
+        ).to eq true
+      end
+    end
+  end
+
   describe "#each_record" do
 
     context "when input is a fasta file" do
@@ -138,8 +189,8 @@ describe SeqFile do
         err_msg = "Input does not look like FASTA or FASTQ"
 
         expect { SeqFile.open(fname).each_record do |h, s|
-            puts [h, s].join ' '
-          end
+                   puts [h, s].join ' '
+                 end
         }.to raise_error(ArgumentError, err_msg)
       end
     end

data/spec/spec_helper.rb

@@ -31,6 +31,17 @@ module Helpers
              ["seq3", "yyyyyyyyyyyyyyyNNN"],
              ["empty seq at end", ""]]
 
+  RECORDS_MAP = {
+    "empty seq at beginning" => "",
+    "seq1 is fun" => "AACTGGNNN",
+    "seq2" => "AATCCTGNNN",
+    "empty seq 1" => "",
+    "empty seq 2" => "",
+    "seq3" => "yyyyyyyyyyyyyyyNNN",
+    "empty seq at end" => ""
+  }
+
+
   TRUTHY_RECORDS = [["empty seq at beginning", []],
                     ["seq1 is fun", ["AACTGGNNN"]],
                     ["seq2", ["AAT", "CCTGNNN"]],

metadata

@@ -1,97 +1,111 @@
 --- !ruby/object:Gem::Specification
 name: parse_fasta
 version: !ruby/object:Gem::Version
-  version: 1.6.2
+  version: 1.7.0
 platform: ruby
 authors:
 - Ryan Moore
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2015-05-22 00:00:00.000000000 Z
+date: 2015-07-11 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: bundler
   requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - "~>"
+    - - ~>
       - !ruby/object:Gem::Version
         version: '1.6'
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - "~>"
+    - - ~>
       - !ruby/object:Gem::Version
         version: '1.6'
 - !ruby/object:Gem::Dependency
   name: rake
   requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - "~>"
+    - - ~>
       - !ruby/object:Gem::Version
         version: '10.3'
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - "~>"
+    - - ~>
       - !ruby/object:Gem::Version
         version: '10.3'
 - !ruby/object:Gem::Dependency
   name: rspec
   requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - "~>"
+    - - ~>
       - !ruby/object:Gem::Version
         version: '2.14'
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - "~>"
+    - - ~>
       - !ruby/object:Gem::Version
         version: '2.14'
 - !ruby/object:Gem::Dependency
   name: bio
   requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - "~>"
+    - - ~>
       - !ruby/object:Gem::Version
         version: '1.4'
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - "~>"
+    - - ~>
       - !ruby/object:Gem::Version
         version: '1.4'
 - !ruby/object:Gem::Dependency
   name: yard
   requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - "~>"
+    - - ~>
       - !ruby/object:Gem::Version
         version: '0.8'
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - "~>"
+    - - ~>
       - !ruby/object:Gem::Version
         version: '0.8'
+- !ruby/object:Gem::Dependency
+  name: rdiscount
+  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: coveralls
   requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - "~>"
+    - - ~>
       - !ruby/object:Gem::Version
         version: '0.7'
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - "~>"
+    - - ~>
       - !ruby/object:Gem::Version
         version: '0.7'
 description: Provides nice, programmatic access to fasta and fastq files, as well
@@ -102,9 +116,10 @@ executables: []
 extensions: []
 extra_rdoc_files: []
 files:
-- ".coveralls.yml"
-- ".gitignore"
-- ".travis.yml"
+- .coveralls.yml
+- .gitignore
+- .travis.yml
+- .yardopts
 - COPYING
 - Gemfile
 - README.md
@@ -131,7 +146,7 @@ files:
 - test_files/test.fq.gz
 homepage: https://github.com/mooreryan/parse_fasta
 licenses:
-- 'GPLv3: http://www.gnu.org/licenses/gpl.txt'
+- ! 'GPLv3: http://www.gnu.org/licenses/gpl.txt'
 metadata: {}
 post_install_message: 
 rdoc_options: []
@@ -139,17 +154,17 @@ require_paths:
 - lib
 required_ruby_version: !ruby/object:Gem::Requirement
   requirements:
-  - - ">="
+  - - ! '>='
     - !ruby/object:Gem::Version
       version: 1.9.3
 required_rubygems_version: !ruby/object:Gem::Requirement
   requirements:
-  - - ">="
+  - - ! '>='
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
 rubyforge_project: 
-rubygems_version: 2.4.6
+rubygems_version: 2.4.8
 signing_key: 
 specification_version: 4
 summary: Easy-peasy parsing of fasta & fastq files!