miga-base 1.3.0.0 → 1.3.1.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: '0832ca5217f409e381a60a13fb077540c6850b3eac0410eb3f077c7e8c82b8ae'
-  data.tar.gz: 44e8b3dfdda624c8fd11e7c2b2e142594383b64a912772b1ae6df8df7a0ef33b
+  metadata.gz: d73a7b2585ba65f467f6648e9ebd49033cd822cfcf4392fa201939b591886743
+  data.tar.gz: 468d835c418fae7bb5b3c90bb357783cbfa8133dfadf2bde5d702d648524e000
 SHA512:
-  metadata.gz: 5dcc61a0bfe803218b94b262f44ee88f2a591ba5416cbedf755f932811825aa91780967e6852aabbbfd38c436e56deef997635209efa7eb8d67fdcd31f9e1a59
-  data.tar.gz: cadef92e9f5a6bc029323660962404c13e6344ef6c662dd2894e364f030b9f8d7d2f2c6ded2617f6de857e485e2e1eff3abafe71f4cb1e4013078dca86c7bf8a
+  metadata.gz: bef438ae52da81802456e7bfac5d6f1ad4e367d20342f5c2c3fc79f4c05d46a7b4023097570ddd127cc80fe119a2114506aa27472512c1a9dce42fbe638f8d9b
+  data.tar.gz: d8426a5097b0bba3f09c3804dd904dae1cae4d5fe1a364647965d71c0ba25eb9ec9da0315a0e856ca4061dba38db6f4d0a720fcd853a784b94bfd199a15729c0

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

@@ -194,7 +194,7 @@ class MiGA::Cli::Action::Init < MiGA::Cli::Action
     req_libraries = {
       r: %w[ape cluster vegan],
       ruby: %w[sqlite3 daemons json],
-      python: %w[numpy sqlite3 pyhmmer]
+      python: %w[numpy sqlite3 pyhmmer pyrodigal]
     }
 
     req_libraries.each do |language, libraries|

data/lib/miga/result/stats.rb

@@ -231,15 +231,16 @@ module MiGA::Result::Stats
 
   # Fix estimates based on essential genes based on taxonomy
   def fix_essential_genes_by_domain
-    return if (tax = source.metadata[:tax]).nil? ||
-       !%w[Archaea Bacteria].include?(tax[:d]) ||
+    tax = source.metadata[:tax]
+    return if (!tax.nil? && !%w[Archaea Bacteria].include?(tax[:d])) ||
        file_path(:raw_report)
 
+    domain = tax.nil? ? 'AB' : tax[:d][0]
     MiGA::MiGA.DEBUG "Fixing essential genes by domain"
     scr = File.join(MiGA::MiGA.root_path, 'utils', 'domain-ess-genes.rb')
     rep = file_path(:report)
     $stderr.print MiGA::MiGA.run_cmd(
-      ['ruby', scr, rep, "#{rep}.domain", tax[:d][0]],
+      ['ruby', scr, rep, "#{rep}.domain", domain],
       return: :output, err2out: true, source: :miga
     )
     add_file(:raw_report, "#{source.name}.ess/log")

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, 0, 0].freeze
+  VERSION = [1.3, 1, 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(2023, 3, 20)
+  VERSION_DATE = Date.new(2023, 3, 21)
 
   ##
   # References of MiGA

data/utils/distance/commands.rb

@@ -2,7 +2,7 @@ module MiGA::DistanceRunner::Commands
   ##
   # Estimates AAI against +targets+ using hAAI
   def haai(targets)
-    puts "[#{Time.now}] hAAI: #{dataset.name} vs #{targets.size} targets"
+    $stderr.puts "[#{Time.now}] hAAI: #{dataset.name} vs #{targets.size} targets"
     empty_vals = targets.map { |_i| nil }
     return empty_vals if opts[:haai_p] == 'no'
 
@@ -19,7 +19,7 @@ module MiGA::DistanceRunner::Commands
   ##
   # Estimates or calculates AAI against +targets+
   def aai(targets)
-    puts "[#{Time.now}] AAI: #{dataset.name} vs #{targets.size} targets"
+    $stderr.puts "[#{Time.now}] AAI: #{dataset.name} vs #{targets.size} targets"
 
     # Try hAAI first
     haai(targets)
@@ -41,7 +41,7 @@ module MiGA::DistanceRunner::Commands
   ##
   # Calculates ANI against +targets+
   def ani(targets)
-    puts "[#{Time.now}] ANI: #{dataset.name} vs #{targets.size} targets"
+    $stderr.puts "[#{Time.now}] ANI: #{dataset.name} vs #{targets.size} targets"
     empty_vals = targets.map { |_i| nil }
     return empty_vals unless File.size?(tmp_file('largecontigs.fa'))
 
@@ -227,7 +227,7 @@ module MiGA::DistanceRunner::Commands
         aai_data[out[1]] = [out[6], 0, 0, 0]
       end
     end
-    puts "Results: #{haai_data.size} | Inferences: #{aai_data.size}"
+    $stderr.puts "Results: #{haai_data.size} | Inferences: #{aai_data.size}"
     batch_data_to_db(:haai, haai_data)
     batch_data_to_db(:aai, aai_data)
 

data/utils/distance/database.rb

@@ -104,6 +104,9 @@ module MiGA::DistanceRunner::Database
       ).first
     end if File.size?(db)
     y
+  rescue SQLite3::CorruptException => e
+    $stderr.puts "Corrupt database: #{db}"
+    raise e
   end
 
   ##

data/utils/distance/temporal.rb

@@ -42,10 +42,19 @@ module MiGA::DistanceRunner::Temporal
   # Copies temporal databases back to the MiGA Project
   def checkpoint!(metric)
     $stderr.puts "Checkpoint (metric = #{metric})"
+
+    # This is simply to test database consistency before overwriting the
+    # previous persistent version
     SQLite3::Database.new(tmp_dbs[metric]) do |conn|
       conn.execute("select count(*) from #{metric == :haai ? :aai : metric}")
     end
-    FileUtils.cp(tmp_dbs[metric], dbs[metric])
+
+    # This reduces the probability of other threads failing due to incomplete
+    # databases, +cp+ can be slower, and that effect is increased by the fact
+    # that tmp_dbs -> dbs could involve a transfer between filesystems, whereas
+    # +move+ within the same filesystem is nearly instantaenous
+    FileUtils.cp(tmp_dbs[metric], "#{dbs[metric]}.tmp")
+    FileUtils.move("#{dbs[metric]}.tmp", dbs[metric])
     @db_counts[metric] = 0
   end
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: miga-base
 version: !ruby/object:Gem::Version
-  version: 1.3.0.0
+  version: 1.3.1.0
 platform: ruby
 authors:
 - Luis M. Rodriguez-R
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2023-03-20 00:00:00.000000000 Z
+date: 2023-03-21 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: daemons