miga-base 1.3.9.8 → 1.3.10.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 6f778c1cf98987c440f71da2da7cfa760279c90009f58cc9abfbe9bb129fec97
-  data.tar.gz: bf128ed11141faa28decc79f80d7426a16e113c8d6b381885815ba98a98e5a4f
+  metadata.gz: '0239e39a0588b73d042da7d970925d2d93a5334c858453e032d51b0af760fa27'
+  data.tar.gz: 81e6903e1feba6571d76fe5d113a60414bd0d3b1b3090d6e26367a93cf8d0da7
 SHA512:
-  metadata.gz: ac2ab474cfdb3667d187f71b787d3cd30685198def25beb20cee58f964fbeec3071733637c14a6722c21c74fd3e1bfa4efb8f0920b36f4d6ebb3b2e0b3eef4c9
-  data.tar.gz: eefb0c999194af805699c7a538d8992c9795447f7a5c81decfc2d85750618e57b116ab56f8c34a2478366ec9efce585649c6f78d1aec44c39c4f8633e3494384
+  metadata.gz: d68e55d5335f3da03eb9cea737aad5fa21a7a272e3958db6130e7260387844c1bed92b0b2f655a5a5133772797212b717559da9701723c9885cc9ee7cffc962f
+  data.tar.gz: 8d27c2f580106c0d1f74e6daaf1cb81ffd7c6fbabf86e7569342027ff9e09b1810c245ad560eff0179c3d06a88b2d1a087344c56e34a2d54c799004e2f6370c0

data/lib/miga/cli/action/download/ncbi.rb

@@ -17,10 +17,7 @@ module MiGA::Cli::Action::Download::Ncbi
     cli.opt_flag(opt, 'chromosome', 'Download complete chromosomes')
     cli.opt_flag(opt, 'scaffold', 'Download genomes in scaffolds')
     cli.opt_flag(opt, 'contig', 'Download genomes in contigs')
-    opt.on(
-      '--all',
-      'Download all genomes (in any status)'
-    ) do
+    opt.on('--all', 'Download all genomes (in any status)') do
       cli[:complete] = true
       cli[:chromosome] = true
       cli[:scaffold] = true
@@ -29,6 +26,10 @@ module MiGA::Cli::Action::Download::Ncbi
     opt.on('--ncbi-list-json STRING', '::HIDE::') do |v|
       cli[:ncbi_list_json] = v
     end
+    opt.on(
+      '--ncbi-taxonomy-dump STRING',
+      'Path to an NCBI Taxonomy dump directory to query instead of API calls'
+    ) { |v| MiGA::RemoteDataset.use_ncbi_taxonomy_dump(v) }
   end
 
   def cli_name_modifiers(opt)

data/lib/miga/cli/action/ncbi_get.rb

@@ -24,14 +24,13 @@ class MiGA::Cli::Action::NcbiGet < MiGA::Cli::Action
       cli_name_modifiers(opt)
       cli_filters(opt)
       cli_save_actions(opt)
-      opt.on(
-        '--api-key STRING',
-        '::HIDE::' # For backwards compatibility
-      ) { |v| ENV['NCBI_API_KEY'] = v }
-      opt.on(
-        '--ncbi-api-key STRING',
-        'NCBI API key'
-      ) { |v| ENV['NCBI_API_KEY'] = v }
+      opt.on('--api-key STRING', '::HIDE::') do |v|
+        warn "The use of --api-key is deprecated, please use --ncbi-api-key"
+        ENV['NCBI_API_KEY'] = v
+      end
+      opt.on('--ncbi-api-key STRING', 'NCBI API key') do |v|
+        ENV['NCBI_API_KEY'] = v
+      end
     end
   end
 

data/lib/miga/remote_dataset.rb

@@ -12,6 +12,55 @@ class MiGA::RemoteDataset < MiGA::MiGA
   # Class-level
 
   class << self
+    ##
+    # Path to a directory with a recent NCBI Taxonomy dump to use instead of
+    # making API calls to NCBI servers, which can be obtained at:
+    # https://ftp.ncbi.nih.gov/pub/taxonomy/taxdump.tar.gz
+    def use_ncbi_taxonomy_dump(path)
+      raise "Directory doesn't exist: #{path}" unless File.directory?(path)
+
+      # Structure: { TaxID => ["name", "rank", parent TaxID] }
+      @ncbi_taxonomy_names = {}
+
+      # Read names.dmp
+      File.open(File.join(path, 'names.dmp')) do |fh|
+        fh.each do |ln|
+          row = ln.split(/\t\|\t?/)
+          next unless row[3] == 'scientific name'
+          @ncbi_taxonomy_names[row[0].to_i] = [row[1].strip]
+        end
+      end
+
+      # Read nodes.dmp
+      File.open(File.join(path, 'nodes.dmp')) do |fh|
+        fh.each do |ln|
+          row = ln.split(/\t\|\t?/)
+          child  = row[0].to_i
+          parent = row[1].to_i
+          @ncbi_taxonomy_names[child][1] = row[2]
+          @ncbi_taxonomy_names[child][2] = parent unless parent == child
+        end
+      end
+    end
+
+    ##
+    # Is a local NCBI Taxonomy dump available?
+    def ncbi_taxonomy_dump?
+      (@ncbi_taxonomy_names ||= nil) ? true : false
+    end
+
+    ##
+    # Get the MiGA::Taxonomy object for the lineage of the taxon with TaxID
+    # +id+ using the local NCBI Taxonomy dump.
+    def taxonomy_from_ncbi_dump(id)
+      MiGA::Taxonomy.new(ns: 'ncbi').tap do |tax|
+        while @ncbi_taxonomy_names[id]
+          tax << { @ncbi_taxonomy_names[id][1] => @ncbi_taxonomy_names[id][0] }
+          id = @ncbi_taxonomy_names[id][2]
+        end
+      end
+    end
+
     ##
     # Translate an NCBI Assembly Accession (+acc+) to corresponding internal
     # NCBI ID, with up to +retrials+ retrials if the returned JSON document
@@ -173,6 +222,10 @@ class MiGA::RemoteDataset < MiGA::MiGA
   def get_ncbi_taxonomy
     tax_id = get_ncbi_taxid or return
 
+    if self.class.ncbi_taxonomy_dump?
+      return self.class.taxonomy_from_ncbi_dump(tax_id)
+    end
+
     lineage = { ns: 'ncbi' }
     doc = MiGA::RemoteDataset.download(:ncbi, :taxonomy, tax_id, :xml)
     doc.scan(%r{<Taxon>(.*?)</Taxon>}m).map(&:first).each do |i|

data/lib/miga/version.rb

@@ -12,7 +12,7 @@ module MiGA
   # - String indicating release status:
   #   - rc* release candidate, not released as gem
   #   - [0-9]+ stable release, released as gem
-  VERSION = [1.3, 9, 8].freeze
+  VERSION = [1.3, 10, 0].freeze
 
   ##
   # Nickname for the current major.minor version.
@@ -20,7 +20,7 @@ module MiGA
 
   ##
   # Date of the current gem relese.
-  VERSION_DATE = Date.new(2024, 1, 30)
+  VERSION_DATE = Date.new(2024, 1, 31)
 
   ##
   # References of MiGA

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: miga-base
 version: !ruby/object:Gem::Version
-  version: 1.3.9.8
+  version: 1.3.10.0
 platform: ruby
 authors:
 - Luis M. Rodriguez-R
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2024-01-30 00:00:00.000000000 Z
+date: 2024-01-31 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: daemons