sequenceserver 1.1.0.beta11 → 1.1.0.beta12

data/lib/sequenceserver/api_errors.rb

@@ -1,5 +1,4 @@
 module SequenceServer
-
   # API errors have an http status, title, message, and additional information
   # like stacktrace or information from program output.
   class APIError < StandardError
@@ -38,10 +37,10 @@ module SequenceServer
     end
 
     def message
-<<MSG
-Looks like there's a problem with one of the query sequences, selected
-databases, or advanced parameters.
-MSG
+      <<~MSG
+        Looks like there's a problem with one of the query sequences, selected
+        databases, or advanced parameters.
+      MSG
     end
 
     attr_reader :more_info
@@ -63,12 +62,12 @@ MSG
     end
 
     def message
-<<MSG
-Looks like there is a problem with the server. Try visiting the page again
-after a while. If this message persists, please report the problem on our
-<a href="https://github.com/wurmlab/sequenceserver/issues" target="_blank">
-issue tracker</a>.
-MSG
+      <<~MSG
+        Looks like there is a problem with the server. Try visiting the page again
+        after a while. If this message persists, please report the problem on our
+        <a href="https://github.com/wurmlab/sequenceserver/issues" target="_blank">
+        issue tracker</a>.
+      MSG
     end
 
     attr_reader :more_info

data/lib/sequenceserver/blast/constants.rb

@@ -4,18 +4,18 @@ module SequenceServer
   module BLAST
     ERROR_LINE = /Error:\s(.*)/
 
-    ALGORITHMS = %w(blastn blastp blastx tblastn tblastx)
+    ALGORITHMS = %w[blastn blastp blastx tblastn tblastx].freeze
 
-    OUTFMT_SPECIFIERS = %w(qseqid qgi qacc sseqid sallseqid sgi sallgi sacc
+    OUTFMT_SPECIFIERS = %w[qseqid qgi qacc sseqid sallseqid sgi sallgi sacc
                            sallacc qstart qend sstart send qseq sseq evalue
                            bitscore score length length pident nident
                            mismatch positive gapopen gaps ppos frames
                            qframe hframe btop staxids sscinames scomnames
                            sblastnames sskingdoms stitle salltitles sstrand
-                           qcovs qcovhsp).join(' ')
+                           qcovs qcovhsp].join(' ').freeze
 
-    STDREP = %w(qseqid sseqid sscinames pident length mismatch gapopen qstart qend
-                sstart send evalue bitscore qcovs qcovhsp).join(' ')
+    STDREP = %w[qseqid sseqid sscinames pident length mismatch gapopen qstart
+                qend sstart send evalue bitscore qcovs qcovhsp].join(' ').freeze
     OUTFMT = {
       'pairwise'        => [0, :txt],
       'qa'              => [1, :txt],
@@ -30,7 +30,7 @@ module SequenceServer
       'asn_binary'      => [9, :asn],
       'csv'             => [10, :csv],
       'archive'         => [11, :txt]
-    }
+    }.freeze
   end
 end
 

data/lib/sequenceserver/blast/formatter.rb

@@ -9,7 +9,7 @@ module SequenceServer
     # return saved ouput.
     class Formatter
       class << self
-        alias_method :run, :new
+        alias run new
       end
 
       extend Forwardable
@@ -41,13 +41,12 @@ module SequenceServer
         return if File.exist?(file)
         command = "blast_formatter -archive '#{job.stdout}'" \
           " -outfmt '#{format} #{specifiers}'"
-        sys(command, path: config[:bin], dir: DOTDIR,
-            stdout: file)
+        sys(command, path: config[:bin], dir: DOTDIR, stdout: file)
       rescue CommandFailed => e
         # Mostly we will never get here: empty archive file,
         # file permissions, broken BLAST binaries, etc. will
         # have been caught before reaching here.
-        fail SystemError, e.stderr
+        raise SystemError, e.stderr
       end
     end
   end

data/lib/sequenceserver/blast/hit.rb

@@ -59,8 +59,11 @@ module SequenceServer
       end
 
       def to_json(*args)
-        [:number, :id, :accession, :title, :length, :score, :identity, :qcovs,
-         :sciname, :evalue, :hsps, :links].inject({}) { |h, k| h[k] = send(k); h }.to_json(*args)
+        %i[number id accession title length score identity qcovs
+         sciname evalue hsps links].inject({}) { |h, k|
+          h[k] = send(k)
+          h
+        }.to_json(*args)
       end
 
       private

data/lib/sequenceserver/blast/hsp.rb

@@ -5,8 +5,8 @@ module SequenceServer
     HSP = Struct.new(:hit, :number, :bit_score, :score, :evalue, :qstart, :qend,
                      :sstart, :send, :qframe, :sframe, :identity, :positives,
                      :gaps, :length, :qcovhsp, :qseq, :sseq, :midline) do
-      INTEGER_ARGS = [1, 3].concat((5..15).to_a)
-      FLOAT_ARGS   = [2, 4]
+      INTEGER_ARGS = [1, 3].concat((5..15).to_a).freeze
+      FLOAT_ARGS   = [2, 4].freeze
 
       def initialize(*args)
         INTEGER_ARGS.each do |i|
@@ -21,9 +21,12 @@ module SequenceServer
       end
 
       def to_json(*args)
-        [:number, :bit_score, :score, :evalue, :qstart, :qend,
-         :sstart, :send, :qframe, :sframe, :identity, :positives,
-         :gaps, :length, :qcovhsp, :qseq, :sseq, :midline].inject({}) { |h, k| h[k] = self[k]; h }.to_json(*args)
+        %i[number bit_score score evalue qstart qend
+           sstart send qframe sframe identity positives
+           gaps length qcovhsp qseq sseq midline].inject({}) { |h, k|
+          h[k] = self[k]
+          h
+        }.to_json(*args)
       end
     end
   end

data/lib/sequenceserver/blast/job.rb

@@ -1,12 +1,10 @@
 require 'sequenceserver/job'
 
 module SequenceServer
-
   # BLAST module.
   module BLAST
     # Extends SequenceServer::Job to describe a BLAST job.
     class Job < Job
-
       def initialize(params)
         validate params
         super do
@@ -33,9 +31,11 @@ module SequenceServer
 
       # Override Job#raise! to raise specific API errors based on exitstatus
       # and using contents of stderr to provide context about the error.
+      #
+      # rubocop:disable Metrics/CyclomaticComplexity
       def raise!
         # Return true exit status is 0 and stdout is not empty.
-        return true if exitstatus == 0 && !File.zero?(stdout)
+        return true if exitstatus.zero? && !File.zero?(stdout)
 
         # Handle error. See [1].
         case exitstatus
@@ -47,16 +47,16 @@ module SequenceServer
         when 4
           # Out of memory. User can retry with a shorter search, so raising
           # InputError here instead of SystemError.
-          fail InputError, <<MSG
-Ran out of memory. Please try a smaller query, or searching fewer and smaller
-databases, or limiting the output by using advanced options.
-MSG
+          fail InputError, <<~MSG
+            Ran out of memory. Please try a smaller query, or searching fewer and smaller
+            databases, or limiting the output by using advanced options.
+            MSG
         when 6
           # Error creating output files. It can't be a permission issue as that
           # would have been caught while creating job directory. But we can run
           # out of storage after creating the job directory and while running
           # the job. This is a SystemError.
-          fail SystemError, "Ran out of disk space."
+          fail SystemError, 'Ran out of disk space.'
         else
           # I am not sure what the exit codes 2 & 3 means and we should note
           # encounter exit code 5. The only other error that I know can happen
@@ -64,16 +64,17 @@ MSG
           # macOS updates. So raise SystemError, include the exit status in the
           # message, and say that that the "most likely" reason is broken BLAST+
           # binaries.
-          fail SystemError, <<MSG
-BLAST failed abruptly (exit status: #{exitstatus}). Most likely there is a
-problem with the BLAST+ binaries.
-MSG
+          fail SystemError, <<~MSG
+            BLAST failed abruptly (exit status: #{exitstatus}). Most likely there is a
+            problem with the BLAST+ binaries.
+            MSG
         end
       end
+      # rubocop:enable Metrics/CyclomaticComplexity
 
       private
 
-      def parse_advanced param_line
+      def parse_advanced(param_line)
         param_list = (param_line || '').split(' ')
         res = {}
 

data/lib/sequenceserver/blast/query.rb

@@ -10,7 +10,7 @@ module SequenceServer
       def initialize(*args)
         args[1] = args[1].to_i
         if args[2] == 'No definition line' ||
-            args[2] == 'unnamed protein product'
+           args[2] == 'unnamed protein product'
           args[2] = "Query_#{args[1]}"
         end
         args[3] = args[3].to_i
@@ -22,8 +22,10 @@ module SequenceServer
       attr_reader :id, :title
 
       def to_json(*args)
-        [:number, :id, :title, :length, :hits]
-          .inject({}) { |h, k| h[k] = send(k); h }.to_json(*args)
+        %i[number id title length hits].inject({}) do |h, k|
+          h[k] = send(k)
+          h
+        end.to_json(*args)
       end
     end
   end

data/lib/sequenceserver/blast/report.rb

@@ -1,5 +1,5 @@
 require 'ox'
-Ox.default_options = {skip: :skip_none}
+Ox.default_options = { skip: :skip_none }
 
 require 'sequenceserver/report'
 require 'sequenceserver/links'
@@ -37,8 +37,10 @@ module SequenceServer
 
       def to_json
         [:querydb, :program, :program_version, :params, :stats,
-         :queries].inject({}) { |h, k| h[k] = send(k); h }.
-        update(search_id: job.id, submitted_at: job.submitted_at.utc).to_json
+         :queries].inject({}) { |h, k|
+          h[k] = send(k)
+          h
+        }.update(search_id: job.id, submitted_at: job.submitted_at.utc).to_json
       end
 
       private
@@ -82,13 +84,13 @@ module SequenceServer
       def extract_stats(ir)
         stats  = ir[8].first[5][0]
         @stats = {
-          :nsequences   => stats[0],
-          :ncharacters  => stats[1],
-          :hsp_length   => stats[2],
-          :search_space => stats[3],
-          :kappa        => stats[4],
-          :labmda       => stats[5],
-          :entropy      => stats[6]
+          nsequences:   stats[0],
+          ncharacters:  stats[1],
+          hsp_length:   stats[2],
+          search_space: stats[3],
+          kappa:        stats[4],
+          labmda:       stats[5],
+          entropy:      stats[6]
         }
       end
 
@@ -115,7 +117,7 @@ module SequenceServer
             n[2] = defline.join(' ')
           end
           hit = Hit.new(query, n[0], n[1], n[3], n[2], n[4],
-                        tsv_ir[n[1]][0], tsv_ir[n[1]][1],[])
+                        tsv_ir[n[1]][0], tsv_ir[n[1]][1], [])
           extract_hsps(n[5], tsv_ir[n[1]][2], hit)
           query.hits << hit
         end
@@ -133,23 +135,23 @@ module SequenceServer
       def parse_xml(xml)
         node_to_array Ox.parse(xml).root
       rescue Ox::ParseError
-        fail InputError, <<MSG
-BLAST generated incorrect XML output. This can happen if sequence ids in your
-databases are not unique across all files. As a temporary workaround, you can
-repeat the search with one database at a time. Proper fix is to recreate the
-following databases with unique sequence ids:
+        fail InputError, <<~MSG
+          BLAST generated incorrect XML output. This can happen if sequence ids in your
+          databases are not unique across all files. As a temporary workaround, you can
+          repeat the search with one database at a time. Proper fix is to recreate the
+          following databases with unique sequence ids:
 
-    #{querydb.map(&:title).join(', ')}
+              #{querydb.map(&:title).join(', ')}
 
-If you are not the one managing this server, try to let the manager know
-about this.
-MSG
+          If you are not the one managing this server, try to let the manager know
+          about this.
+        MSG
       end
 
-      PARSEABLE_AS_HASH  = %w(Parameters)
-      PARSEABLE_AS_ARRAY = %w(BlastOutput_param Iteration_stat Statistics
+      PARSEABLE_AS_HASH  = %w[Parameters].freeze
+      PARSEABLE_AS_ARRAY = %w[BlastOutput_param Iteration_stat Statistics
                               Iteration_hits BlastOutput_iterations
-                              Iteration Hit Hit_hsps Hsp)
+                              Iteration Hit Hit_hsps Hsp].freeze
 
       def node_to_hash(element)
         Hash[*element.nodes.map { |n| [n.name, node_to_value(n)] }.flatten]

data/lib/sequenceserver/config.rb

@@ -60,6 +60,20 @@ module SequenceServer
         data[:database_dir] ||= database_dir
       end
 
+      # We would like a persistent :options: key in the config file, explicitly
+      # stating the parameters to use. Recommended values are written to config
+      # file on the first run. However, existing users won't have :options: key
+      # available from before. For them, we retain the old behaviour of
+      # automatically adding `-task blastn` for BLASTN searches.
+      if blast_opts = data.dig(:options, :blastn)
+        unless blast_opts.join.match('-task')
+          # Issue a warning.
+          logger.info "BLASTN will be run using '-task blastn' option." +
+                      " You can override this through configuration file."
+          data[:options][:blastn].push '-task blastn'
+        end
+      end
+
       data
     end
 
@@ -67,11 +81,10 @@ module SequenceServer
     # otherwise.
     def parse_config_file
       unless file? config_file
-        logger.debug "Configuration file not found: #{config_file}"
+        logger.info "Configuration file not found: #{config_file}"
         return {}
       end
-
-      logger.debug "Reading configuration file: #{config_file}."
+      logger.info "Reading configuration file: #{config_file}."
       normalize YAML.load_file(config_file)
     rescue => error
       raise CONFIG_FILE_ERROR.new(config_file, error)
@@ -84,9 +97,16 @@ module SequenceServer
     # Default configuration data.
     def defaults
       {
-        :num_threads  => 1,
-        :port         => 4567,
-        :host         => '0.0.0.0'
+        host: '0.0.0.0',
+        port: 4567,
+        options: {
+          blastn:  ['-task blastn', '-evalue 1e-5'],
+          blastp:  ['-evalue 1e-5'],
+          blastx:  ['-evalue 1e-5'],
+          tblastx: ['-evalue 1e-5'],
+          tblastn: ['-evalue 1e-5']
+        },
+        num_threads: 1
       }
     end
   end

data/lib/sequenceserver/database.rb

@@ -38,7 +38,7 @@ module SequenceServer
 
     def include?(accession)
       cmd = "blastdbcmd -entry '#{accession}' -db #{name}"
-      out, _ = sys(cmd, path: config[:bin])
+      out, = sys(cmd, path: config[:bin])
       !out.empty?
     end
 
@@ -159,7 +159,7 @@ module SequenceServer
           next if Database.include? file
           next unless probably_fasta? file
           sequence_type = guess_sequence_type_in_fasta file
-          if [:protein, :nucleotide].include?(sequence_type)
+          if %i[protein nucleotide].include?(sequence_type)
             list << [file, sequence_type]
           end
         end
@@ -229,8 +229,9 @@ module SequenceServer
       # /home/ben/pd.ben/sequenceserver/db/nr => no
       # /home/ben/pd.ben/sequenceserver/db/img3.5.finished.faa.01 => yes
       # /home/ben/pd.ben/sequenceserver/db/nr00 => no
+      # /mnt/blast-db/refseq_genomic.100 => yes
       def multipart_database_name?(db_name)
-        !(db_name.match(%r{.+/\S+\.\d{2}$}).nil?)
+        !(db_name.match(%r{.+/\S+\.\d{2,3}$}).nil?)
       end
 
       # Returns true if first character of the file is '>'.

data/lib/sequenceserver/doctor.rb

@@ -34,8 +34,8 @@ module SequenceServer
 
           out = `blastdbcmd -entry all -db #{db.name} -outfmt "%i" 2> /dev/null`
           {
-            :db     => db,
-            :seqids => out.to_s.split
+            db:     db,
+            seqids: out.to_s.split
           }
         end.compact
       end
@@ -66,33 +66,33 @@ module SequenceServer
 
         case error
         when ERROR_PARSE_SEQIDS
-          puts <<MSG
-*** Doctor has found improperly formatted database:
-#{bullet_list(values)}
-    Please reformat your databases with -parse_seqids switch (or use
-    sequenceserver -m) for using SequenceServer as the current format
-    may cause problems.
+          puts <<~MSG
+            *** Doctor has found improperly formatted database:
+            #{bullet_list(values)}
+            Please reformat your databases with -parse_seqids switch (or use
+            sequenceserver -m) for using SequenceServer as the current format
+            may cause problems.
 
-    These databases are ignored in further checks.
-MSG
+            These databases are ignored in further checks.
+          MSG
 
         when ERROR_NUMERIC_IDS
-          puts <<MSG
-*** Doctor has found databases with numeric sequence ids:
-#{bullet_list(values)}
-    Note that this may cause problems with sequence retrieval.
-MSG
+          puts <<~MSG
+            *** Doctor has found databases with numeric sequence ids:
+            #{bullet_list(values)}
+            Note that this may cause problems with sequence retrieval.
+          MSG
 
         when ERROR_PROBLEMATIC_IDS
-          puts <<MSG
-*** Doctor has found databases with problematic sequence ids:
-#{bullet_list(values)}
-    This causes some sequence to contain extraneous words like `gnl|`
-    appended to their id string.
-MSG
+          puts <<~MSG
+            *** Doctor has found databases with problematic sequence ids:
+            #{bullet_list(values)}
+            This causes some sequence to contain extraneous words like `gnl|`
+            appended to their id string.
+          MSG
         end
       end
-      # rubocop:disable Metrics/MethodLength
+      # rubocop:enable Metrics/MethodLength
     end
 
     def initialize