miga-base 0.4.0.0 → 0.4.1.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 37bf8085cc9b33f88367c8134cd2b01115cfcd9e9116fab7ba3f93abc33c6d6f
-  data.tar.gz: bdefeaa9f965ea991071a70b48239ad7129763bcb05f0cc2ebcd6d66511bd3e6
+  metadata.gz: c3c339de89529e19ad52d1d3522fbe25206045d3db3b030e4ee8c7ddf7d3c1db
+  data.tar.gz: d220804d9a9bc722496f271135da3c4415a77c84d40e49db5d39d015e13d9865
 SHA512:
-  metadata.gz: 9e1463de777f3d77bcb8f804bdfedeb3781ccf0b36330faf21d501b1a87cf9b55f0d03f2668a1889ae165cbaaf2030ab955ad8c0f42a5aaad341a6eb7154b356
-  data.tar.gz: 6738493c3a716f553fea6486841f612616bad38cc4091103ec490417336df327ae03046a04907b9dcaed9c5d81e17ab918d64d841a67a50b23565dc46308736b
+  metadata.gz: 27b5d98711e579c077eda211387457578e70183a30459dc6c6e917b5fc20cdbc730d649dfd58b0a2477ea019e0577c1c4993fc6d29eaee9bbdeb77a753ddb107
+  data.tar.gz: 8a391e9090bdf8dbf555fcdd456c1ae7c909a69b1f59cf5eac039d6f18ffbd0b9caf04bdcad406485f914c1c14681105d086e2b36f82039102df4d6b29baace0

data/lib/miga/dataset.rb

@@ -48,8 +48,10 @@ class MiGA::Dataset < MiGA::MiGA
   # be treated as reference (true, default) or query (false). Pass any
   # additional +metadata+ as a Hash.
   def initialize(project, name, is_ref=true, metadata={})
-    raise "Invalid name '#{name}', please use only alphanumerics and " +
-      "underscores." unless name.miga_name?
+    unless name.miga_name?
+      raise 'Invalid name, please use only alphanumerics and underscores: ' +
+        name.to_s
+    end
     @project = project
     @name = name
     metadata[:ref] = is_ref
@@ -60,8 +62,9 @@ class MiGA::Dataset < MiGA::MiGA
   ##
   # Save any changes you've made in the dataset.
   def save
-    self.metadata[:type] = :metagenome if !metadata[:tax].nil? and
-      !metadata[:tax][:ns].nil? and metadata[:tax][:ns]=="COMMUNITY"
+    if t = metadata[:tax] and n = t[:ns] and n == 'COMMUNITY'
+      self.metadata[:type] = :metagenome
+    end
     self.metadata.save
   end
   
@@ -95,7 +98,7 @@ class MiGA::Dataset < MiGA::MiGA
   # Get standard metadata values for the dataset as Array.
   def info
     MiGA::Dataset.INFO_FIELDS.map do |k|
-      (k=="name") ? self.name : metadata[k.to_sym]
+      (k == 'name') ? self.name : metadata[k.to_sym]
     end
   end
   
@@ -132,11 +135,11 @@ class MiGA::Dataset < MiGA::MiGA
   def ignore_task?(task)
     return true unless is_active?
     return !metadata["run_#{task}"] unless metadata["run_#{task}"].nil?
-    return true if task==:taxonomy and project.metadata[:ref_project].nil?
+    return true if task == :taxonomy and project.metadata[:ref_project].nil?
     pattern = [true, false]
-    ( [@@_EXCLUDE_NOREF_TASKS_H[task], is_ref?     ]==pattern or
-      [@@_ONLY_MULTI_TASKS_H[task],    is_multi?   ]==pattern or
-      [@@_ONLY_NONMULTI_TASKS_H[task], is_nonmulti?]==pattern )
+    ( [@@_EXCLUDE_NOREF_TASKS_H[task], is_ref?     ] == pattern or
+      [@@_ONLY_MULTI_TASKS_H[task],    is_multi?   ] == pattern or
+      [@@_ONLY_NONMULTI_TASKS_H[task], is_nonmulti?] == pattern )
   end
 
   ##
@@ -146,13 +149,14 @@ class MiGA::Dataset < MiGA::MiGA
   # This function is currently only supported for query datasets when
   # +ref_project+ is false (default), and only for reference dataset when
   # +ref_project+ is true. It returns +nil+ if this analysis is not supported.
-  def closest_relatives(how_many=1, ref_project=false)
+  def closest_relatives(how_many = 1, ref_project = false)
     return nil if (is_ref? != ref_project) or is_multi?
     r = result(ref_project ? :taxonomy : :distances)
     return nil if r.nil?
     db = SQLite3::Database.new(r.file_path :aai_db)
-    db.execute("SELECT seq2, aai FROM aai WHERE seq2 != ? " +
-      "GROUP BY seq2 ORDER BY aai DESC LIMIT ?", [name, how_many])
+    db.execute(
+      'SELECT seq2, aai FROM aai WHERE seq2 != ? ' \
+      'GROUP BY seq2 ORDER BY aai DESC LIMIT ?', [name, how_many])
   end
 
 end # class MiGA::Dataset

data/lib/miga/remote_dataset.rb

@@ -42,7 +42,7 @@ class MiGA::RemoteDataset < MiGA::MiGA
     @db = db.to_sym
     @universe = universe.to_sym
     @metadata = {}
-    @metadata[:"#{universe}_#{db}"] = ids.join(",")
+    @metadata[:"#{universe}_#{db}"] = ids.join(',')
     @@UNIVERSE.keys.include?(@universe) or
       raise "Unknown Universe: #{@universe}. Try: #{@@UNIVERSE.keys}"
     @@UNIVERSE[@universe][:dbs].include?(@db) or
@@ -130,7 +130,7 @@ class MiGA::RemoteDataset < MiGA::MiGA
   def get_ncbi_taxonomy
     tax_id = get_ncbi_taxid
     return nil if tax_id.nil?
-    lineage = {}
+    lineage = {ns: 'ncbi'}
     doc = MiGA::RemoteDataset.download(:ncbi, :taxonomy, tax_id, :xml)
     doc.scan(%r{<Taxon>(.*?)</Taxon>}m).map(&:first).each do |i|
       name = i.scan(%r{<ScientificName>(.*)</ScientificName>}).first.to_a.first

data/lib/miga/taxonomy.rb

@@ -38,7 +38,9 @@ class MiGA::Taxonomy < MiGA::MiGA
 
   ##
   # Initialize from JSON-derived Hash +o+.
-  def self.json_create(o) new(o['str']) ; end
+  def self.json_create(o)
+    new(o['str'], nil, o['alt'])
+  end
 
   ##
   # Returns cannonical rank (Symbol) for the +rank+ String.
@@ -63,8 +65,10 @@ class MiGA::Taxonomy < MiGA::MiGA
   # space-delimited entries, the array is a vector of entries. Each entry can be
   # either a rank:value pair (if +ranks+ is nil), or just values in the same
   # order as ther ranks in +ranks+. Alternatively, +str+ as a Hash with rank =>
-  # value pairs is also supported.
-  def initialize(str, ranks = nil)
+  # value pairs is also supported. If +alt+ is passed, it must be an Array of
+  # String, Array, or Hash entries as defined above (except +ranks+ are not
+  # allowed).
+  def initialize(str, ranks = nil, alt = [])
     @ranks = {}
     if ranks.nil?
       case str when Array, Hash
@@ -74,12 +78,13 @@ class MiGA::Taxonomy < MiGA::MiGA
       end
     else
       ranks = ranks.split(/\s+/) unless ranks.is_a? Array
-      str = str.split(/\s/) unless str.is_a? Array
+      str = str.split(/\s+/) unless str.is_a? Array
       raise "Unequal number of ranks (#{ranks.size}) " +
         "and names (#{str.size}):#{ranks} => #{str}" unless
         ranks.size==str.size
       (0 .. str.size).each{ |i| self << "#{ranks[i]}:#{str[i]}" }
     end
+    @alt = (alt || []).map { |i| Taxonomy.new(i) }
   end
   
   ##
@@ -89,7 +94,7 @@ class MiGA::Taxonomy < MiGA::MiGA
     if value.is_a? Hash
       value.each_pair do |rank_i, name_i|
         next if name_i.nil? or name_i == ""
-        @ranks[ Taxonomy.normalize_rank rank_i ] = name_i.tr("_"," ")
+        @ranks[ Taxonomy.normalize_rank rank_i ] = name_i.tr('_',' ')
       end
     elsif value.is_a? Array
       value.each{ |v| self << v }
@@ -104,6 +109,26 @@ class MiGA::Taxonomy < MiGA::MiGA
   ##
   # Get +rank+ value.
   def [](rank) @ranks[ rank.to_sym ] ; end
+
+  ##
+  # Get the alternative taxonomies.
+  # - If +which+ is nil (default), returns all alternative taxonomies as Array
+  #   (not including the master taxonomy).
+  # - If +which+ is Integer, returns the indexed taxonomy
+  #   (starting with 0, the master taxonomy).
+  # - Otherwise, returns the first taxonomy with namespace +which+ (coerced as
+  #   String), including the master taxonomy.
+  # In the latter two cases it can be nil.
+  def alternative(which = nil)
+    case which
+    when nil
+      @alt
+    when Integer
+      ([self] + @alt)[which]
+    else
+      ([self] + @alt).find{ |i| i.namespace.to_s == which.to_s }
+    end
+  end
   
   ##
   # Evaluates if the loaded taxonomy includes +taxon+. It assumes that +taxon+
@@ -116,30 +141,50 @@ class MiGA::Taxonomy < MiGA::MiGA
 
   ##
   # Sorted list of ranks, as an Array of two-entry Arrays (rank and value).
-  def sorted_ranks
+  # If +force_ranks+ is true, it returns all standard ranks even if undefined.
+  # If +with_namespace+ is true, it includes also the namespace.
+  def sorted_ranks(force_ranks = false, with_namespace = false)
     @@KNOWN_RANKS.map do |r|
-      ranks[r].nil? ? nil : [r, ranks[r]]
+      next if r == :ns and not with_namespace
+      next if ranks[r].nil? and not force_ranks
+      [r, ranks[r]]
     end.compact
   end
+
+  ##
+  # Namespace of the taxonomy (a String) or +nil+.
+  def namespace; self[ :ns ] ; end
   
   ##
   # Get the most general rank as a two-entry Array (rank and value).
-  def highest; sorted_ranks.first ; end
+  # If +force_ranks+ is true, it always returns the value for domain (d)
+  # even if undefined.
+  def highest(force_ranks = false)
+    sorted_ranks.first
+  end
 
   ##
   # Get the most specific rank as a two-entry Array (rank and value).
-  def lowest; sorted_ranks.last ; end
+  # If +force_ranks+ is true, it always returns the value for dataset (ds)
+  # even if undefined.
+  def lowest(force_ranks = false)
+    sorted_ranks(force_ranks).last
+  end
   
   ##
-  # Generate cannonical String for the taxonomy.
-  def to_s
-    sorted_ranks.map{ |r| "#{r[0]}:#{r[1].gsub(/[\s:]/,"_")}" }.join(" ")
+  # Generate cannonical String for the taxonomy. If +force_ranks+ is true,
+  # it returns all the standard ranks even if undefined.
+  def to_s(force_ranks = false)
+    sorted_ranks(force_ranks, true).
+      map{ |r| "#{r[0]}:#{(r[1] || '').gsub(/[\s:]/, '_')}" }.join(' ')
   end
   
   ##
   # Generate JSON-formated String representing the taxonomy.
   def to_json(*a)
-    { JSON.create_id => self.class.name, "str" => self.to_s }.to_json(*a)
+    hsh = { JSON.create_id => self.class.name, 'str' => self.to_s }
+    hsh['alt'] = alternative.map(&:to_s) unless alternative.empty?
+    hsh.to_json(*a)
   end
   
 end

data/lib/miga/version.rb

@@ -10,7 +10,7 @@ module MiGA
   # - Float representing the major.minor version.
   # - Integer representing gem releases of the current version.
   # - Integer representing minor changes that require new version number.
-  VERSION = [0.4, 0, 0]
+  VERSION = [0.4, 1, 0]
   
   ##
   # Nickname for the current major.minor version.
@@ -18,7 +18,7 @@ module MiGA
   
   ##
   # Date of the current gem release.
-  VERSION_DATE = Date.new(2019, 8, 15)
+  VERSION_DATE = Date.new(2019, 8, 17)
   
   ##
   # Reference of MiGA.

data/test/remote_dataset_test.rb

@@ -24,28 +24,35 @@ class RemoteDatasetTest < Test::Unit::TestCase
   end
 
   def test_bad_remote_dataset
-    assert_raise { MiGA::RemoteDataset.new("ids", :embl, :marvel) }
-    assert_raise { MiGA::RemoteDataset.new("ids", :google, :ebi) }
+    assert_raise { MiGA::RemoteDataset.new('ids', :embl, :marvel) }
+    assert_raise { MiGA::RemoteDataset.new('ids', :google, :ebi) }
   end
 
   def test_rest
-    hiv2 = "M30502.1"
+    hiv2 = 'M30502.1'
     {embl: :ebi, nuccore: :ncbi}.each do |db, universe|
       rd = MiGA::RemoteDataset.new(hiv2, db, universe)
       assert_equal([hiv2], rd.ids)
-      omit_if(!$remote_tests, "Remote access is error-prone.")
+      omit_if(!$remote_tests, 'Remote access is error-prone.')
       tx = rd.get_ncbi_taxonomy
-      assert_equal(MiGA::Taxonomy, tx.class, "Failed on #{universe}:#{db}")
-      assert_equal("Lentivirus", tx[:g], "Failed on #{universe}:#{db}")
+      msg = "Failed on #{universe}:#{db}"
+      assert_equal(MiGA::Taxonomy, tx.class, msg)
+      assert_equal('Lentivirus', tx[:g], msg)
+      assert_equal('ns:ncbi o:Ortervirales f:Retroviridae ' \
+        'g:Lentivirus s:Human_immunodeficiency_virus_2', tx.to_s, msg)
+      assert_equal('ns:ncbi d: k: p: c: o:Ortervirales f:Retroviridae ' \
+        'g:Lentivirus s:Human_immunodeficiency_virus_2 ssp: str: ds:',
+        tx.to_s(true), msg)
+      assert_equal('ncbi', tx.namespace, msg)
     end
   end
 
   def test_net_ftp
-    cjac = "ftp://ftp.ebi.ac.uk/pub/databases/ena/tsa/public/ga/GAPJ01.fasta.gz"
-    n = "Cjac_L14"
+    cjac = 'ftp://ftp.ebi.ac.uk/pub/databases/ena/tsa/public/ga/GAPJ01.fasta.gz'
+    n = 'Cjac_L14'
     rd = MiGA::RemoteDataset.new(cjac, :assembly_gz, :web)
     assert_equal([cjac], rd.ids)
-    omit_if(!$remote_tests, "Remote access is error-prone.")
+    omit_if(!$remote_tests, 'Remote access is error-prone.')
     p = $p1
     assert_nil(p.dataset(n))
     rd.save_to(p, n)

data/test/taxonomy_test.rb

@@ -6,22 +6,32 @@ class TaxonomyTest < Test::Unit::TestCase
   def test_ranks
     assert_respond_to(MiGA::Taxonomy, :KNOWN_RANKS)
     assert(MiGA::Taxonomy.KNOWN_RANKS.include? :s)
-    assert_nil(MiGA::Taxonomy.normalize_rank "No Rank")
-    assert_nil(MiGA::Taxonomy.normalize_rank "Captain")
+    assert_nil(MiGA::Taxonomy.normalize_rank 'No Rank')
+    assert_nil(MiGA::Taxonomy.normalize_rank 'Captain')
     assert_equal(:f, MiGA::Taxonomy.normalize_rank(:Family))
   end
 
   def test_json
-    js = '{"json_class":"MiGA::Taxonomy",' +
-      '"str":"k:Fantasia f:Dragonaceae s:Dragonia_azura"}'
-    tx = JSON.parse(js, {:symbolize_names=>false, :create_additions=>true})
+    txt = 'k:Fantasia f:Dragonaceae s:Dragonia_azura'
+    js = '{"json_class":"MiGA::Taxonomy","str":"' + txt + '"}'
+    tx = JSON.parse(js, {symbolize_names: false, create_additions: true})
     assert_equal(MiGA::Taxonomy, tx.class)
-    assert_equal("Dragonaceae", tx[:f])
+    assert_equal('Dragonaceae', tx[:f])
     assert_equal(js, tx.to_json)
   end
 
+  def test_namespace
+    txt = 'ns:Irrealis k:Fantasia f:Dragonaceae s:Dragonia_azura'
+    tx = MiGA::Taxonomy.new(txt)
+    assert_equal(txt, tx.to_s)
+    assert_equal(
+      [[:k, 'Fantasia'],[:f, 'Dragonaceae'],[:s, 'Dragonia azura']],
+      tx.sorted_ranks)
+    assert_equal('Irrealis', tx.namespace)
+  end
+
   def test_append
-    tx = MiGA::Taxonomy.new ""
+    tx = MiGA::Taxonomy.new ''
     assert_equal("", "#{tx}")
     tx << ["domain:Public","family:GNU"]
     assert_equal("GNU", tx[:f])
@@ -44,4 +54,22 @@ class TaxonomyTest < Test::Unit::TestCase
     end
   end
 
+  def test_alt
+    tx = MiGA::Taxonomy.new('ns:a s:Arnie', nil,
+      ['ns:b s:Bernie','ns:c s:Cornie','s:Darnie'])
+    assert_equal('ns:a s:Arnie', tx.to_s)
+    assert_equal([[:s, 'Arnie']], tx.sorted_ranks)
+    assert_equal('ns:a s:Arnie', tx.alternative(0).to_s)
+    assert_equal('ns:b s:Bernie', tx.alternative(1).to_s)
+    assert_equal('ns:c s:Cornie', tx.alternative(:c).to_s)
+    assert_equal('s:Darnie', tx.alternative('').to_s)
+    assert_nil(tx.alternative(:x))
+    assert_equal(3, tx.alternative.size)
+    js = tx.to_json
+    tx_js = JSON.parse(js, {symbolize_names: false, create_additions: true})
+    assert_equal(tx.to_s, tx_js.to_s)
+    assert_equal(tx.alternative(2).to_s, tx_js.alternative(2).to_s)
+    assert_equal(tx.alternative.size, tx_js.alternative.size)
+  end
+
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: miga-base
 version: !ruby/object:Gem::Version
-  version: 0.4.0.0
+  version: 0.4.1.0
 platform: ruby
 authors:
 - Luis M. Rodriguez-R
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2019-08-15 00:00:00.000000000 Z
+date: 2019-08-17 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: daemons