Module: MiGA::Cli::Action::Download::Seqcode

Includes:
Base
Included in:
SeqcodeGet
Defined in:
lib/miga/cli/action/download/seqcode.rb

Overview

Helper module including download functions for the seqcode_get action

Instance Method Summary collapse

Methods included from Base

#cli_base_flags, #cli_filters, #cli_save_actions, #discard_excluded, #download_entries, #exclude_newer, #finalize_tasks, #generic_perform, #impose_limit, #load_ncbi_taxonomy_dump, #load_tasks, #save_entry, #unlink_entries

Instance Method Details

#cli_name_modifiers(opt) ⇒ Object 



13
14
15
16
17
18
 
# File 'lib/miga/cli/action/download/seqcode.rb', line 13

def cli_name_modifiers(opt)
  opt.on(
    '--no-version-name',
    'Do not add sequence version to the dataset name'
  ) { |v| cli[:add_version] = v }
end

#cli_task_flags(_opt) ⇒ Object 



10
11
 
# File 'lib/miga/cli/action/download/seqcode.rb', line 10

def cli_task_flags(_opt)
end

#remote_list ⇒ Object 



24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
 
# File 'lib/miga/cli/action/download/seqcode.rb', line 24

def remote_list
  cli.say 'Downloading genome list'
  current_page = 1
  total_pages  = 1
  ds = {}

  while current_page <= total_pages
    json = MiGA::RemoteDataset.download(
      :seqcode, :'type-genomes', nil, :json, nil, page: current_page
    )
    doc = MiGA::Json.parse(json, contents: true)
    current_page = doc[:current_page] + 1
    total_pages  = doc[:total_pages]

    doc[:values].each do |name|
      next unless name[:type_material]
      acc = name[:type_material].values.first
      db  = name[:type_material].keys.first
      next unless %i[assembly nuccore].include?(db) # No INSDC genome, ignore

      classif = name[:classification] || {}
      tax = MiGA::Taxonomy.new(Hash[classif.map { |i| [i[:rank], i[:name]] }])
      tax << { 'ns' => 'seqcode', name[:rank] => name[:name] }
      d = {
        ids: [acc], db: db, universe: :seqcode,
        md: {
          type: :genome, tax: tax, is_type: true,
          type_rel: 'SeqCode type genome',
          seqcode_url: "https://seqco.de/i:#{name[:id]}"
        }
      }
      d[:md][:get_ncbi_taxonomy] = true if cli[:get_ncbi_taxonomy]
      ds[remote_row_name(tax, db, acc)] = d
    end
  end
  ds
end

#remote_row_name(tax, db, asm) ⇒ Object 



62
63
64
65
66
67
 
# File 'lib/miga/cli/action/download/seqcode.rb', line 62

def remote_row_name(tax, db, asm)
  acc = asm.to_s
  acc.gsub!(/\.\d+\Z/, '') unless cli[:add_version]
  db_short = { assembly: 'asm', nuccore: 'gb' }[db]
  "#{tax.lowest[1]}_#{db_short}_#{acc}".miga_name
end

#sanitize_cli ⇒ Object 



20
21
22
 
# File 'lib/miga/cli/action/download/seqcode.rb', line 20

def sanitize_cli
  cli[:save_every] = 1 if cli[:dry]
end