Class: Bulkrax::CsvParser

Inherits:
ApplicationParser show all
Includes:
ErroredEntries, ExportBehavior
Defined in:
app/parsers/bulkrax/csv_parser.rb

Overview

rubocop:disable Metrics/ClassLength

Direct Known Subclasses

BagitParser

Instance Attribute Summary collapse

Attributes inherited from ApplicationParser

#headers, #importerexporter

Class Method Summary collapse

Instance Method Summary collapse

Methods included from ExportBehavior

#build_export_metadata, #build_for_exporter, #filename, #hyrax_record

Methods included from ErroredEntries

#build_errored_entry_row, #setup_errored_entries_file, #write_errored_entries_file

Methods inherited from ApplicationParser

#base_path, #exporter?, #find_or_create_entry, #generated_metadata_mapping, #get_field_mapping_hash_for, #import_file_path, import_supported?, #importer?, #initialize, #invalid_record, #limit_reached?, #model_field_mappings, #new_entry, parser_fields, #path_for_import, #perform_method, #record, #record_has_source_identifier, #related_children_parsed_mapping, #related_children_raw_mapping, #related_parents_parsed_mapping, #related_parents_raw_mapping, #required_elements, #source_identifier, #unzip, #visibility, #work_identifier, #write, #write_import_file, #zip

Constructor Details

This class inherits a constructor from Bulkrax::ApplicationParser

Instance Attribute Details

#collections ⇒ Object 



56
57
58
59
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 56

def collections
  build_records if @collections.nil?
  @collections
end

#file_sets ⇒ Object 



66
67
68
69
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 66

def file_sets
  build_records if @file_sets.nil?
  @file_sets
end

#works ⇒ Object 



61
62
63
64
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 61

def works
  build_records if @works.nil?
  @works
end

Class Method Details

.export_supported? ⇒ Boolean

Returns:

  • (Boolean) 


10
11
12
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 10

def self.export_supported?
  true
end

Instance Method Details

#build_records ⇒ Object 



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
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 26

def build_records
  @collections = []
  @works = []
  @file_sets = []

  if model_field_mappings.map { |mfm| mfm.to_sym.in?(records.first.keys) }.any?
    records.map do |r|
      model_field_mappings.map(&:to_sym).each do |model_mapping|
        next unless r.key?(model_mapping)

        model = r[model_mapping].nil? ? "" : r[model_mapping].strip
        if model.casecmp('collection').zero?
          @collections << r
        elsif model.casecmp('fileset').zero?
          @file_sets << r
        else
          @works << r
        end
      end
    end
    @collections = @collections.flatten.compact.uniq
    @file_sets = @file_sets.flatten.compact.uniq
    @works = @works.flatten.compact.uniq
  else # if no model is specified, assume all records are works
    @works = records.flatten.compact.uniq
  end

  true
end

#collection_entry_class ⇒ Object 



209
210
211
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 209

def collection_entry_class
  CsvCollectionEntry
end

#collections_total ⇒ Object 



71
72
73
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 71

def collections_total
  collections.size
end

#create_collections ⇒ Object 



116
117
118
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 116

def create_collections
  create_objects(['collection'])
end

#create_entry_and_job(current_record, type) ⇒ Object 



155
156
157
158
159
160
161
162
163
164
165
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 155

def create_entry_and_job(current_record, type)
  new_entry = find_or_create_entry(send("#{type}_entry_class"),
                                   current_record[source_identifier],
                                   'Bulkrax::Importer',
                                   current_record.to_h)
  if current_record[:delete].present?
    "Bulkrax::Delete#{type.camelize}Job".constantize.send(perform_method, new_entry, current_run)
  else
    "Bulkrax::Import#{type.camelize}Job".constantize.send(perform_method, new_entry.id, current_run.id)
  end
end

#create_file_sets ⇒ Object 



124
125
126
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 124

def create_file_sets
  create_objects(['file_set'])
end

#create_new_entries ⇒ Object Also known as: create_from_collection, create_from_importer, create_from_worktype, create_from_all 



186
187
188
189
190
191
192
193
194
195
196
197
198
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 186

def create_new_entries
  # NOTE: The each method enforces the limit, as it can best optimize the underlying queries.
  current_records_for_export.each do |id, entry_class|
    new_entry = find_or_create_entry(entry_class, id, 'Bulkrax::Exporter')
    begin
      entry = ExportWorkJob.perform_now(new_entry.id, current_run.id)
    rescue => e
      Rails.logger.info("#{e.message} was detected during export")
    end

    self.headers |= entry.parsed_metadata.keys if entry
  end
end

#create_objects(types_array = nil) ⇒ Object 



132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 132

def create_objects(types_array = nil)
  index = 0
  (types_array || %w[collection work file_set relationship]).each do |type|
    if type.eql?('relationship')
      ScheduleRelationshipsJob.set(wait: 5.minutes).perform_later(importer_id: importerexporter.id)
      next
    end
    send(type.pluralize).each do |current_record|
      next unless record_has_source_identifier(current_record, index)
      break if limit_reached?(limit, index)

      seen[current_record[source_identifier]] = true
      create_entry_and_job(current_record, type)
      increment_counters(index, "#{type}": true)
      index += 1
    end
    importer.record_status
  end
  true
rescue StandardError => e
  set_status_info(e)
end

#create_relationships ⇒ Object 



128
129
130
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 128

def create_relationships
  create_objects(['relationship'])
end

#create_works ⇒ Object 



120
121
122
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 120

def create_works
  create_objects(['work'])
end

#current_records_for_export ⇒ Object 



179
180
181
182
183
184
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 179

def current_records_for_export
  @current_records_for_export ||= Bulkrax::ParserExportRecordSet.for(
    parser: self,
    export_from: importerexporter.export_from
  )
end

#entry_class ⇒ Object Also known as: work_entry_class 



204
205
206
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 204

def entry_class
  CsvEntry
end

#export_headers ⇒ Object

All possible column names



316
317
318
319
320
321
322
323
324
325
326
327
328
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 316

def export_headers
  headers = sort_headers(self.headers)

  # we don't want access_control_id exported and we want file at the end
  headers.delete('access_control_id') if headers.include?('access_control_id')

  # add the headers below at the beginning or end to maintain the preexisting export behavior
  headers.prepend('model')
  headers.prepend(source_identifier.to_s)
  headers.prepend('id')

  headers.uniq
end

#export_key_allowed(key) ⇒ Object 



310
311
312
313
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 310

def export_key_allowed(key)
  new_entry(entry_class, 'Bulkrax::Exporter').field_supported?(key) &&
    key != source_identifier.to_s
end

#file_paths ⇒ Object

Retrieve file paths for [:file] mapping in records

and check all listed files exist.

Raises:

  • (StandardError) 


375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 375

def file_paths
  raise StandardError, 'No records were found' if records.blank?
  return [] if importerexporter.metadata_only?

  @file_paths ||= records.map do |r|
    file_mapping = Bulkrax.field_mappings.dig(self.class.to_s, 'file', :from)&.first&.to_sym || :file
    next if r[file_mapping].blank?

    r[file_mapping].split(Bulkrax.multi_value_element_split_on).map do |f|
      file = File.join(path_to_files, f.tr(' ', '_'))
      if File.exist?(file) # rubocop:disable Style/GuardClause
        file
      else
        raise "File #{file} does not exist"
      end
    end
  end.flatten.compact.uniq
end

#file_set_entry_class ⇒ Object 



213
214
215
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 213

def file_set_entry_class
  CsvFileSetEntry
end

#file_sets_total ⇒ Object 



79
80
81
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 79

def file_sets_total
  file_sets.size
end

#import_fields ⇒ Object

We could use CsvEntry#fields_from_data(data) but that would mean re-reading the data



84
85
86
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 84

def import_fields
  @import_fields ||= records.inject(:merge).keys.compact.uniq
end

#missing_elements(record) ⇒ Object 



92
93
94
95
96
97
98
99
100
101
102
103
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 92

def missing_elements(record)
  keys_from_record = keys_without_numbers(record.reject { |_, v| v.blank? }.keys.compact.uniq.map(&:to_s))
  keys = []
  # Because we're persisting the mapping in the database, these are likely string keys.
  # However, there's no guarantee.  So, we need to ensure that by running stringify.
  importerexporter.mapping.stringify_keys.map do |k, v|
    Array.wrap(v['from']).each do |vf|
      keys << k if keys_from_record.include?(vf)
    end
  end
  required_elements.map(&:to_s) - keys.uniq.map(&:to_s)
end

#object_names ⇒ Object 



330
331
332
333
334
335
336
337
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 330

def object_names
  return @object_names if @object_names

  @object_names = mapping.values.map { |value| value['object'] }
  @object_names.uniq!&.delete(nil)

  @object_names
end

#path_to_files(**args) ⇒ Object

Retrieve the path where we expect to find the files



395
396
397
398
399
400
401
402
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 395

def path_to_files(**args)
  filename = args.fetch(:filename, '')

  return @path_to_files if @path_to_files.present? && filename.blank?
  @path_to_files = File.join(
      zip? ? importer_unzip_path : File.dirname(import_file_path), 'files', filename
    )
end

#records(_opts = {}) ⇒ Object 



14
15
16
17
18
19
20
21
22
23
24
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 14

def records(_opts = {})
  return @records if @records.present?

  file_for_import = only_updates ? parser_fields['partial_import_file_path'] : import_file_path
  # data for entry does not need source_identifier for csv, because csvs are read sequentially and mapped after raw data is read.
  csv_data = entry_class.read_data(file_for_import)
  importer.parser_fields['total'] = csv_data.count
  importer.save

  @records = csv_data.map { |record_data| entry_class.data_for_entry(record_data, nil, self) }
end

#records_split_count ⇒ Object 



238
239
240
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 238

def records_split_count
  1000
end

#required_elements?(record) ⇒ Boolean

Returns:

  • (Boolean) 


88
89
90
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 88

def required_elements?(record)
  missing_elements(record).blank?
end

#retrieve_cloud_files(files) ⇒ Object

TODO:

  • investigate getting directory structure

TODO:

  • investigate using perform_later, and having the importer check for

DownloadCloudFileJob before it starts



245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 245

def retrieve_cloud_files(files)
  files_path = File.join(path_for_import, 'files')
  FileUtils.mkdir_p(files_path) unless File.exist?(files_path)
  files.each_pair do |_key, file|
    # fixes bug where auth headers do not get attached properly
    if file['auth_header'].present?
      file['headers'] ||= {}
      file['headers'].merge!(file['auth_header'])
    end
    # this only works for uniquely named files
    target_file = File.join(files_path, file['file_name'].tr(' ', '_'))
    # Now because we want the files in place before the importer runs
    # Problematic for a large upload
    Bulkrax::DownloadCloudFileJob.perform_now(file, target_file)
  end
  return nil
end

#setup_export_file(folder_count) ⇒ Object

in the parser as it is specific to the format



366
367
368
369
370
371
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 366

def setup_export_file(folder_count)
  path = File.join(importerexporter.exporter_export_path, folder_count.to_s)
  FileUtils.mkdir_p(path) unless File.exist?(path)

  File.join(path, "export_#{importerexporter.export_source}_from_#{importerexporter.export_from}_#{folder_count}.csv")
end

#sort_entries(entries) ⇒ Object 



339
340
341
342
343
344
345
346
347
348
349
350
351
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 339

def sort_entries(entries)
  # always export models in the same order: work, collection, file set
  entries.sort_by do |entry|
    case entry.type
    when 'Bulkrax::CsvCollectionEntry'
      '1'
    when 'Bulkrax::CsvFileSetEntry'
      '2'
    else
      '0'
    end
  end
end

#sort_headers(headers) ⇒ Object 



353
354
355
356
357
358
359
360
361
362
363
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 353

def sort_headers(headers)
  # converting headers like creator_name_1 to creator_1_name so they get sorted by numerical order
  # while keeping objects grouped together
  headers.sort_by do |item|
    number = item.match(/\d+/)&.[](0) || 0.to_s
    sort_number = number.rjust(4, "0")
    object_prefix = object_names.detect { |o| item.match(/^#{o}/) } || item
    remainder = item.gsub(/^#{object_prefix}_/, '').gsub(/_#{number}/, '')
    "#{object_prefix}_#{sort_number}_#{remainder}"
  end
end

#store_files(identifier, folder_count) ⇒ Object 



288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 288

def store_files(identifier, folder_count)
  record = ActiveFedora::Base.find(identifier)
  return unless record

  file_sets = record.file_set? ? Array.wrap(record) : record.file_sets
  file_sets << record.thumbnail if exporter.include_thumbnails && record.thumbnail.present? && record.work?
  file_sets.each do |fs|
    path = File.join(exporter_export_path, folder_count, 'files')
    FileUtils.mkdir_p(path) unless File.exist? path
    file = filename(fs)
    next if file.blank? || fs.original_file.blank?

    io = open(fs.original_file.uri)
    File.open(File.join(path, file), 'wb') do |f|
      f.write(io.read)
      f.close
    end
  end
rescue Ldp::Gone
  return
end

#total ⇒ Object

TODO: figure out why using the version of this method that’s in the bagit parser breaks specs for the “if importer?” line



223
224
225
226
227
228
229
230
231
232
233
234
235
236
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 223

def total
  @total =
    if importer?
      importer.parser_fields['total'] || 0
    elsif exporter?
      limit.to_i.zero? ? current_records_for_export.count : limit.to_i
    else
      0
    end

  return @total
rescue StandardError
  @total = 0
end

#valid_entry_types ⇒ Object 



217
218
219
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 217

def valid_entry_types
  [collection_entry_class.to_s, file_set_entry_class.to_s, entry_class.to_s]
end

#valid_import? ⇒ Boolean

Returns:

  • (Boolean) 


105
106
107
108
109
110
111
112
113
114
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 105

def valid_import?
  compressed_record = records.flat_map(&:to_a).partition { |_, v| !v }.flatten(1).to_h
  error_alert = "Missing at least one required element, missing element(s) are: #{missing_elements(compressed_record).join(', ')}"
  raise StandardError, error_alert unless required_elements?(compressed_record)

  file_paths.is_a?(Array)
rescue StandardError => e
  set_status_info(e)
  false
end

#works_total ⇒ Object 



75
76
77
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 75

def works_total
  works.size
end

#write_files ⇒ Object

export methods



265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 265

def write_files
  require 'open-uri'
  folder_count = 0
  # TODO: This is not performant as well; unclear how to address, but lower priority as of
  #       <2023-02-21 Tue>.
  sorted_entries = sort_entries(importerexporter.entries.uniq(&:identifier))
                   .select { |e| valid_entry_types.include?(e.type) }

  group_size = limit.to_i.zero? ? total : limit.to_i
  sorted_entries[0..group_size].in_groups_of(records_split_count, false) do |group|
    folder_count += 1

    CSV.open(setup_export_file(folder_count), "w", headers: export_headers, write_headers: true) do |csv|
      group.each do |entry|
        csv << entry.parsed_metadata
        next if importerexporter.metadata_only? || entry.type == 'Bulkrax::CsvCollectionEntry'

        store_files(entry.identifier, folder_count.to_s)
      end
    end
  end
end

#write_partial_import_file(file) ⇒ Object 



167
168
169
170
171
172
173
174
175
176
177
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 167

def write_partial_import_file(file)
  import_filename = import_file_path.split('/').last
  partial_import_filename = "#{File.basename(import_filename, '.csv')}_corrected_entries.csv"

  path = File.join(path_for_import, partial_import_filename)
  FileUtils.mv(
    file.path,
    path
  )
  path
end