Class: Bulkrax::CsvParser

Inherits:
ApplicationParser show all
Includes:
CsvTemplateGeneration, CsvValidation, ErroredEntries, ExportBehavior
Defined in:
app/parsers/bulkrax/csv_parser.rb,
app/parsers/concerns/bulkrax/csv_parser/csv_validation.rb,
app/parsers/concerns/bulkrax/csv_parser/csv_validation_helpers.rb,
app/parsers/concerns/bulkrax/csv_parser/csv_template_generation.rb,
app/parsers/concerns/bulkrax/csv_parser/csv_validation_hierarchy.rb

Overview

rubocop:disable Metrics/ClassLength

Direct Known Subclasses

BagitParser

Defined Under Namespace

Modules: CsvTemplateGeneration, CsvValidation, CsvValidationHelpers, CsvValidationHierarchy

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, #file_extension, #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, #calculate_type_delay, #copy_file, #create_collections, #create_entry_and_job, #create_file_sets, #create_objects, #create_relationships, #create_works, #exporter?, #extract_zip_entry, #find_or_create_entry, #generated_metadata_mapping, #get_field_mapping_hash_for, #import_file_path, import_supported?, #importer?, #initialize, #invalid_record, #limit_reached?, #macos_junk_entry?, #model_field_mappings, #new_entry, parser_fields, #path_for_import, #perform_method, #rebuild_entries, #rebuild_entry_query, #record, #record_deleted?, #record_has_source_identifier, #record_raw_metadata, #record_remove_and_rerun?, #reject_unsafe_entry!, #related_children_parsed_mapping, #related_children_raw_mapping, #related_parents_parsed_mapping, #related_parents_raw_mapping, #required_elements, #safe_extract_path, #source_identifier, #untar, #unzip, #visibility, #work_entry_class, #work_identifier, #work_identifier_search_field, #write, #write_import_file, #zip

Constructor Details

This class inherits a constructor from Bulkrax::ApplicationParser

Instance Attribute Details

#collections ⇒ Object

rubocop:enabled Metrics/AbcSize



74
75
76
77
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 74

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

#file_sets ⇒ Object 



84
85
86
87
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 84

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

#validation_mode ⇒ Object

Returns the value of attribute validation_mode.



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

def validation_mode
  @validation_mode
end

#works ⇒ Object 



79
80
81
82
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 79

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

Class Method Details

.export_supported? ⇒ Boolean

Returns:

  • (Boolean) 


12
13
14
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 12

def self.export_supported?
  true
end

.file_split_pattern ⇒ Regexp

Returns the pattern String#split should use on a ‘file` cell. Honours the `file` mapping’s ‘split:` when set, otherwise falls back to Bulkrax#multi_value_element_split_on.

Returns:

  • (Regexp) —

    the pattern String#split should use on a ‘file` cell. Honours the `file` mapping’s ‘split:` when set, otherwise falls back to Bulkrax#multi_value_element_split_on.



19
20
21
22
23
24
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 19

def self.file_split_pattern
  file_mapping = Bulkrax.field_mappings.dig(to_s, 'file') ||
                 Bulkrax.field_mappings.dig(to_s, :file) || {}
  split_value  = file_mapping['split'] || file_mapping[:split]
  Bulkrax::SplitPatternCoercion.coerce(split_value) || Bulkrax.multi_value_element_split_on
end

Instance Method Details

#build_records ⇒ Object

rubocop:disable Metrics/AbcSize



41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 41

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
        # TODO: Eventually this should be refactored to us Hyrax.config.collection_model
        #       We aren't right now because so many Bulkrax users are in between Fedora and Valkyrie
        if model.casecmp('collection').zero? || model.casecmp('collectionresource').zero?
          @collections << r
        elsif model.casecmp('fileset').zero? || model.casecmp('hyrax::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 



175
176
177
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 175

def collection_entry_class
  CsvCollectionEntry
end

#collections_total ⇒ Object 



89
90
91
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 89

def collections_total
  collections.size
end

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



153
154
155
156
157
158
159
160
161
162
163
164
165
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 153

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

#current_records_for_export ⇒ Object 



146
147
148
149
150
151
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 146

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

#entry_class ⇒ Object 



171
172
173
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 171

def entry_class
  CsvEntry
end

#export_headers ⇒ Object

All possible column names



299
300
301
302
303
304
305
306
307
308
309
310
311
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 299

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 



293
294
295
296
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 293

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) 


361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 361

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

  # Compute once — these don't vary per record.
  file_mapping  = Bulkrax.field_mappings.dig(self.class.to_s, 'file', :from)&.first&.to_sym || :file
  split_pattern = self.class.file_split_pattern
  files_dir     = path_to_files

  @file_paths ||= records.map do |r|
    next if r[file_mapping].blank?
    raise StandardError, "Record references local files but no files directory could be resolved from the import path" if files_dir.nil?

    r[file_mapping].split(split_pattern).map do |f|
      file = File.join(files_dir, f.strip.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 



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

def file_set_entry_class
  CsvFileSetEntry
end

#file_sets_total ⇒ Object 



97
98
99
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 97

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



102
103
104
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 102

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

#missing_elements(record) ⇒ Object 



110
111
112
113
114
115
116
117
118
119
120
121
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 110

def missing_elements(record)
  keys_from_record = keys_without_numbers(record.reject { |_, v| v.blank? }.keys.compact.uniq.map(&:to_s))
  keys = []
  mapping_values = importerexporter.mapping.stringify_keys
  mapping_values.each do |k, v|
    from_values = Array.wrap(v.is_a?(Hash) ? (v['from'] || v[:from]) : nil)
    from_values.each do |vf|
      keys << k if vf.present? && keys_from_record.include?(vf.to_s.strip)
    end
  end
  required_elements.map(&:to_s) - keys.uniq.map(&:to_s)
end

#object_names ⇒ Object 



313
314
315
316
317
318
319
320
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 313

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 for this import. After ImporterJob#unzip_imported_file runs (zip cases), attachments live under ‘importer_unzip_path/files/`. For a server-path-style import (the user specified a CSV file path with a sibling `files/` directory on disk), resolve relative to the CSV’s directory instead.

When called with ‘filename:`, returns the full path to that file if it exists on disk, or `nil` otherwise — callers like `Bulkrax::FileSetEntryBehavior#add_path_to_file` rely on the nil sentinel to fall back to the raw filename in their error messages.

When called with no filename, returns the ‘files/` directory itself (only when that directory exists on disk — else `nil` so callers can raise a clear “no files directory” error).



399
400
401
402
403
404
405
406
407
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 399

def path_to_files(**args)
  filename = args.fetch(:filename, '')
  base_dir = files_dir
  return base_dir if filename.blank? && Dir.exist?(base_dir)
  return nil if filename.blank?

  candidate = File.join(base_dir, filename)
  candidate if File.exist?(candidate)
end

#records(_opts = {}) ⇒ Object 



26
27
28
29
30
31
32
33
34
35
36
37
38
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 26

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

  file_for_import = only_updates ? parser_fields['partial_import_file_path'] : import_file_path
  csv_data = entry_class.read_data(file_for_import)
  unless validation_mode
    importer.parser_fields['total'] = csv_data.count
    importer.save
  end

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

#records_split_count ⇒ Object 



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

def records_split_count
  1000
end

#remove_spaces_from_filenames ⇒ Object

File names referenced in CSVs have spaces replaced with underscores.

See Also:



462
463
464
465
466
467
468
469
470
471
472
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 462

def remove_spaces_from_filenames
  files = Dir.glob(File.join(importer_unzip_path, 'files', '*'))
  files_with_spaces = files.select { |f| f.split('/').last.include?(' ') }
  return if files_with_spaces.blank?

  files_with_spaces.map! { |path| Pathname.new(path) }
  files_with_spaces.each do |path|
    filename_without_spaces = path.basename.to_s.tr(' ', '_')
    path.rename(File.join(path.dirname, filename_without_spaces))
  end
end

#required_elements?(record) ⇒ Boolean

Returns:

  • (Boolean) 


106
107
108
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 106

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

#retrieve_cloud_files(files, importer) ⇒ Object

TODO:

  • investigate getting directory structure

TODO:

  • investigate using perform_later, and having the importer check for

DownloadCloudFileJob before it starts



211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 211

def retrieve_cloud_files(files, importer)
  files_path = File.join(path_for_import, 'files')
  FileUtils.mkdir_p(files_path) unless File.exist?(files_path)
  target_files = []
  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(' ', '_'))
    target_files << target_file
    # Now because we want the files in place before the importer runs
    # Problematic for a large upload
    Bulkrax::DownloadCloudFileJob.perform_later(file, target_file)
  end
  importer[:parser_fields]['original_file_paths'] = target_files
  return nil
end

#setup_export_file(folder_count) ⇒ Object

in the parser as it is specific to the format



352
353
354
355
356
357
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 352

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 



322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 322

def sort_entries(entries)
  # always export models in the same order: work, collection, file set
  #
  # TODO: This is a problem in that only these classes are compared.  Instead
  #       We should add a comparison operator to the classes.
  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 



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

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 



258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 258

def store_files(identifier, folder_count)
  record = Bulkrax.object_factory.find(identifier)
  return unless record

  file_sets = Array.wrap(record) if record.file_set?
  if file_sets.nil? # for valkyrie
    file_sets = record.respond_to?(:file_sets) ? record.file_sets : record.members&.select(&:file_set?)
  end

  if importerexporter.include_thumbnails?
    thumbnail = Bulkrax.object_factory.thumbnail_for(resource: record)
    file_sets << thumbnail if thumbnail.present?
  end

  file_sets.each do |fs|
    path = File.join(exporter_export_path, folder_count, 'files')
    FileUtils.mkdir_p(path) unless File.exist? path

    original_file = Bulkrax.object_factory.original_file(fileset: fs)
    next if original_file.blank?
    file = filename(fs)

    io = original_file.respond_to?(:uri) ? open(original_file.uri) : original_file.file.io

    File.open(File.join(path, file), 'wb') do |f|
      f.write(io.read)
      f.close
    end
  end
rescue Ldp::Gone
  return
rescue StandardError => e
  raise StandardError, "Unable to retrieve files for identifier #{identifier} - #{e.message}"
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



189
190
191
192
193
194
195
196
197
198
199
200
201
202
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 189

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

#unzip_attachments_only(file_to_unzip) ⇒ Object

Extracts a zip that accompanies a separately-uploaded CSV. Every entry lands under #importer_unzip_path/files/ — including any CSVs inside the zip, which are treated as attachments since the primary CSV was uploaded outside the zip. Strips a single top-level wrapper directory if present, so users can zip either the contents or the enclosing folder.

Parameters:

  • file_to_unzip (String) —

    absolute path to a .zip



446
447
448
449
450
451
452
453
454
455
456
457
458
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 446

def unzip_attachments_only(file_to_unzip)
  dest_dir = importer_unzip_path(mkdir: true)
  Zip::File.open(file_to_unzip) do |zip_file|
    entries = real_zip_entries(zip_file)
    wrapper = single_top_level_wrapper(entries)

    entries.each do |entry|
      relative = wrapper ? entry.name.delete_prefix("#{wrapper}/") : entry.name
      next if relative.empty?
      extract_to(zip_file, entry, dest_dir, File.join('files', relative))
    end
  end
end

#unzip_with_primary_csv(file_to_unzip) ⇒ Object

Extracts a zip that contains a primary CSV. The primary CSV lands at the root of #importer_unzip_path; every other entry lands under #importer_unzip_path/files/, preserving its path relative to the primary CSV’s directory.

Primary-CSV selection matches the guided-import validator’s rule (see ImporterFileHandler#locate_csv_entry_in_zip): the CSV entry at the shallowest directory level. Visible errors are raised on zero CSVs or multiple CSVs at the shallowest level.

Parameters:

  • file_to_unzip (String) —

    absolute path to a .zip

Raises:



421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 421

def unzip_with_primary_csv(file_to_unzip)
  dest_dir = importer_unzip_path(mkdir: true)
  Zip::File.open(file_to_unzip) do |zip_file|
    entries = real_zip_entries(zip_file)
    primary = select_primary_csv!(entries)
    primary_dir = File.dirname(primary.name)

    entries.each do |entry|
      if entry == primary
        extract_to(zip_file, entry, dest_dir, File.basename(entry.name))
      else
        extract_to(zip_file, entry, dest_dir, File.join('files', relative_to(primary_dir, entry.name)))
      end
    end
  end
end

#valid_entry_types ⇒ Object 



183
184
185
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 183

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) 


123
124
125
126
127
128
129
130
131
132
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 123

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 



93
94
95
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 93

def works_total
  works.size
end

#write_files ⇒ Object

export methods



234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 234

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
        # TODO: This is precarious when we have descendents of Bulkrax::CsvCollectionEntry
        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 



134
135
136
137
138
139
140
141
142
143
144
 
# File 'app/parsers/bulkrax/csv_parser.rb', line 134

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