Class: Aspera::Cli::Plugins::Preview

Inherits:
BasicAuth show all
Defined in:
lib/aspera/cli/plugins/preview.rb

Constant Summary

Constants inherited from Base

Base::FILTER_ARGS

Instance Attribute Summary collapse

Attributes inherited from Base

#context, #help_path

Instance Method Summary collapse

Methods inherited from BasicAuth

#basic_auth_api, #basic_auth_params

Methods inherited from Base

#action_for, #add_manual_header, application_name, #bulk_result, command, command_registry, commands_under, #config, crud_commands, declare_options, define_action_method, #dispatch_child, #dispatch_from_registry, #dispatch_leaf, #entity_create, #entity_delete, entity_display_name, #entity_list, #entity_modify, #entity_res_path, #entity_show, #execute_action, #execute_leaf, file_matcher, #formatter, #generate_help, #http_config, #invoke_action, option, #options, #persistency, #presets, #progress_bar, #query_read_delete, #resolve_argument, root_setup, #transfer, use_options, used_option_sources

Constructor Details

#initialize(**_) ⇒ Preview

Returns a new instance of Preview.



73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
# File 'lib/aspera/cli/plugins/preview.rb', line 73

def initialize(**_)
  super
  # Generator configuration populated from CLI options.
  @gen_options = Aspera::Preview::Options.new
  # Used to rate-limit periodic progress logging and checkpoint persistence.
  @periodic = TimerLimiter.new(LOG_LIMITER_SEC)
  # Optional callback used to filter entries before generation.
  @filter_block = nil
  @access_remote = true
  # Bind generator-specific options to @gen_options (Category C: set_handler after object creation)
  Aspera::Preview::Options::DESCRIPTIONS.each do |opt|
    options.set_handler(opt[:name], object: @gen_options, method: opt[:name])
  end

  options.parse_options!
  Api::Node.api_options[:cache] = !@option_folder_reset_cache.eql?(:header)
  # Start from the full supported format list, then remove any skipped format.
  @preview_formats_to_generate = Aspera::Preview::Generator::PREVIEW_FORMATS.clone
  skip = options.get_option(:skip_format)
  @preview_formats_to_generate.delete(skip) if skip
  @tmp_folder = File.join(TempFileManager.instance.global_temp, "#{TMP_DIR_PREFIX}.#{SecureRandom.uuid}")
  FileUtils.mkdir_p(@tmp_folder)
  Log.log.debug { "tmpdir: #{@tmp_folder}" }
end

Instance Attribute Details

#option_folder_reset_cacheObject

Returns the value of attribute option_folder_reset_cache.



49
50
51
# File 'lib/aspera/cli/plugins/preview.rb', line 49

def option_folder_reset_cache
  @option_folder_reset_cache
end

#option_overwriteObject

Returns the value of attribute option_overwrite.



49
50
51
# File 'lib/aspera/cli/plugins/preview.rb', line 49

def option_overwrite
  @option_overwrite
end

#option_previews_folderObject

Returns the value of attribute option_previews_folder.



49
50
51
# File 'lib/aspera/cli/plugins/preview.rb', line 49

def option_previews_folder
  @option_previews_folder
end

#option_skip_foldersObject

Returns the value of attribute option_skip_folders.



49
50
51
# File 'lib/aspera/cli/plugins/preview.rb', line 49

def option_skip_folders
  @option_skip_folders
end

#option_skip_typesObject

Returns the value of attribute option_skip_types.



49
50
51
# File 'lib/aspera/cli/plugins/preview.rb', line 49

def option_skip_types
  @option_skip_types
end

Instance Method Details

#action_checkObject



493
494
495
496
497
498
# File 'lib/aspera/cli/plugins/preview.rb', line 493

def action_check
  check_tools_and_mimemagic
  Result::Status.new('Tools validated')
ensure
  cleanup_tmp_folder
end

#action_eventsObject



479
480
481
482
483
484
# File 'lib/aspera/cli/plugins/preview.rb', line 479

def action_events(**)
  check_tools_and_mimemagic
  run_event_loop(:events, filter: options.get_option(:filter))
ensure
  cleanup_tmp_folder
end

#action_scan(path:) ⇒ Object

--- handlers ---



463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
# File 'lib/aspera/cli/plugins/preview.rb', line 463

def action_scan(path:, **)
  check_tools_and_mimemagic
  apifid =
    if (selector = Parser.percent_selector(path))
      Aspera.assert_values(selector[:field], ['id'], type: BadArgument) { 'file id' }
      Api::NodeFileId.new(@api_node, selector[:value].to_s.empty? ? @access_key_self['root_file_id'] : selector[:value])
    else
      @api_node.resolve_api_fid(@access_key_self['root_file_id'], path)
    end
  @filter_block = Base.file_matcher(options.get_option(:filter))
  scan_folder_files(@api_node.read("files/#{apifid.file_id}"))
  Result::Status.new('scan finished')
ensure
  cleanup_tmp_folder
end

#action_show(source_file:) ⇒ Object



509
510
511
512
513
514
515
516
517
# File 'lib/aspera/cli/plugins/preview.rb', line 509

def action_show(source_file:, **)
  check_tools_and_mimemagic
  generated_file_path = preview_filename(:png, options.get_option(:base))
  Aspera::Preview::Generator.new(source_file, generated_file_path, @gen_options, @tmp_folder).generate
  formatter.display_status("generated: #{generated_file_path}")
  Result::Image.new(UriReader.file_url(generated_file_path))
ensure
  cleanup_tmp_folder
end

#action_test(source_file:, format:) ⇒ Object



500
501
502
503
504
505
506
507
# File 'lib/aspera/cli/plugins/preview.rb', line 500

def action_test(source_file:, format:, **)
  check_tools_and_mimemagic
  generated_file_path = preview_filename(format, options.get_option(:base))
  Aspera::Preview::Generator.new(source_file, generated_file_path, @gen_options, @tmp_folder).generate
  Result::Status.new("generated: #{generated_file_path}")
ensure
  cleanup_tmp_folder
end

#action_treventsObject



486
487
488
489
490
491
# File 'lib/aspera/cli/plugins/preview.rb', line 486

def action_trevents(**)
  check_tools_and_mimemagic
  run_event_loop(:trevents, filter: options.get_option(:filter))
ensure
  cleanup_tmp_folder
end

#do_transfer(direction, folder_id, source_filename, destination = '/') ⇒ Object

Transfer a file to or from the configured Node storage using a tagged transfer spec.

Parameters:

  • transfer direction, typically from Transfer::Spec

  • Node API identifier of the reference folder

  • relative source path inside the transfer root

  • (defaults to: '/')

    local destination root for receive operations

Returns:



188
189
190
191
192
193
194
195
196
197
198
# File 'lib/aspera/cli/plugins/preview.rb', line 188

def do_transfer(direction, folder_id, source_filename, destination = '/')
  Aspera.assert(!(destination.nil? && direction.eql?(Transfer::Spec::DIRECTION_RECEIVE)), 'destination must be set for receive direction')
  t_spec = @api_node.transfer_spec_gen4(folder_id, direction, {
    'paths' => [{'source' => source_filename}],
    'tags'  => {Transfer::Spec::TAG_RESERVED => {PREV_GEN_TAG => true}}
  })
  # Force the destination on the transfer agent object.
  # Setting `t_spec['destination_root']` directly would later be overwritten.
  transfer.user_transfer_spec['destination_root'] = destination
  Runner.result_transfer(transfer.start(t_spec))
end

#entry_preview_folder_name(entry) ⇒ String

Build the preview folder name for a file entry using the Node API convention.

Parameters:

  • file entry containing an id

Returns:

  • preview folder name for the entry



254
255
256
# File 'lib/aspera/cli/plugins/preview.rb', line 254

def entry_preview_folder_name(entry)
  "#{entry['id']}#{PREVIEW_FOLDER_SUFFIX}"
end

#generate_preview(entry) ⇒ nil

Generate all required previews for a single file entry when regeneration is needed.

Remote entries must include parent_file_id.

Parameters:

  • local or remote file entry to preview

Returns:



274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
# File 'lib/aspera/cli/plugins/preview.rb', line 274

def generate_preview(entry)
  # prepare generic information
  gen_infos = @preview_formats_to_generate.map do |preview_format|
    {
      preview_format: preview_format,
      base_dest:      preview_filename(preview_format)
    }
  end
  # lets gather some infos on possibly existing previews
  # it depends if files access locally or remotely
  # folder where previews will be generated for this particular entry
  local_entry_preview_dir = @access_remote ? get_infos_remote(gen_infos, entry) : get_infos_local(gen_infos, entry)
  # here we have the status on preview files
  # let's find if they need generation
  gen_infos.select! do |gen_info|
    # if it exists, what about overwrite policy ?
    if gen_info[:preview_exist]
      case @option_overwrite
      when :always
        # continue: generate
      when :never
        # never overwrite
        next false
      when :mtime
        # skip if preview is newer than original
        next false if gen_info[:preview_newer_than_original]
      end
    end
    begin
      # need generator for further checks
      gen_info[:generator] = Aspera::Preview::Generator.new(gen_info[:src], gen_info[:dst], @gen_options, @tmp_folder, mime: entry['content_type'])
    rescue
      # no conversion supported
      next false
    end
    # shall we skip it ?
    next false if @option_skip_types.include?(gen_info[:generator].conversion_type)
    # ok we need to generate
    true
  end
  return if gen_infos.empty?
  # create folder if needed
  FileUtils.mkdir_p(local_entry_preview_dir)
  if @access_remote
    Aspera.assert(!entry['parent_file_id'].nil?, 'missing parent_file_id in entry')
    #  download original file to temp folder
    do_transfer(Transfer::Spec::DIRECTION_RECEIVE, entry['parent_file_id'], entry['name'], @tmp_folder)
  end
  Log.log.debug { "source: #{entry['id']}: #{entry['path']}" }
  gen_infos.each do |gen_info|
    gen_info[:generator].generate
  rescue => e
    Log.log.error { "Ignoring: #{e.class} #{e.message}" }
    Log.log.debug(e.backtrace.join("\n").red)
    # in case of any error, place a standard error image
    FileUtils.cp(gen_info[:generator].error_asset, gen_info[:dst])
  end
  if @access_remote
    # upload
    do_transfer(Transfer::Spec::DIRECTION_SEND, @previews_folder_entry['id'], local_entry_preview_dir)
    # cleanup after upload
    FileUtils.rm_rf(local_entry_preview_dir)
    File.delete(File.join(@tmp_folder, entry['name']))
  end
  # force read file updated previews
  @api_node.read("files/#{entry['id']}") if @option_folder_reset_cache.eql?(:read)
rescue StandardError => e
  Log.log.error { "Ignore: #{e.message}" }
  Log.log.debug(e.backtrace.join("\n").red)
end

#get_infos_local(gen_infos, entry) ⇒ String

Populate generation metadata for a source file available on the local filesystem.

Parameters:

  • preview generation descriptors to enrich

  • file entry containing at least the relative path

Returns:

  • local directory where previews for this entry are stored



205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
# File 'lib/aspera/cli/plugins/preview.rb', line 205

def get_infos_local(gen_infos, entry)
  # Resolve to a real path and ensure it stays within the storage root (prevents path traversal via API response)
  candidate = File.expand_path(File.join(@local_storage_root, entry['path']))
  resolved_root = File.realpath(@local_storage_root)
  Aspera.assert(candidate.start_with?("#{resolved_root}/")) { 'Preview entry path traversal attempt detected' }
  local_original_filepath = candidate
  original_mtime = File.mtime(local_original_filepath)
  # Output directory for previews generated from the local source file.
  local_entry_preview_dir = File.join(@local_preview_folder, entry_preview_folder_name(entry))
  gen_infos.each do |gen_info|
    gen_info[:src] = local_original_filepath
    gen_info[:dst] = File.join(local_entry_preview_dir, gen_info[:base_dest])
    gen_info[:preview_exist] = File.exist?(gen_info[:dst])
    gen_info[:preview_newer_than_original] = (gen_info[:preview_exist] && (File.mtime(gen_info[:dst]) > original_mtime))
  end
  return local_entry_preview_dir
end

#get_infos_remote(gen_infos, entry) ⇒ String

Populate generation metadata for a source file stored remotely on Node.

Parameters:

  • preview generation descriptors to enrich

  • remote file entry returned by Node API

Returns:

  • temporary local directory where previews are generated



228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
# File 'lib/aspera/cli/plugins/preview.rb', line 228

def get_infos_remote(gen_infos, entry)
  # Download the source file into the temporary workspace before generating previews.
  local_original_filepath = File.join(@tmp_folder, entry['name'])
  # require 'date'
  # original_mtime=DateTime.parse(entry['modified_time'])
  # Local directory where previews are generated before being uploaded back.
  local_entry_preview_dir = File.join(@tmp_folder, entry_preview_folder_name(entry))
  file_info = @api_node.read("files/#{entry['id']}")
  # TODO: This does not work with Gen4 because preview folders are hidden by the API.
  # this_preview_folder_entries=@api_node.read_folder_content(@previews_folder_entry['id'],{name: @entry_preview_folder_name})
  # TODO: Query Gen3 APIs to list preview files and retrieve timestamps.
  gen_infos.each do |gen_info|
    gen_info[:src] = local_original_filepath
    gen_info[:dst] = File.join(local_entry_preview_dir, gen_info[:base_dest])
    # TODO: Reuse `this_preview_folder_entries` once preview folders become visible.
    gen_info[:preview_exist] = file_info.key?('preview')
    # TODO: Compare source and preview modification times when remote timestamps are available.
    gen_info[:preview_newer_than_original] = gen_info[:preview_exist]
  end
  return local_entry_preview_dir
end

#preview_filename(preview_format, base_name = nil) ⇒ String

Build a preview filename from a basename and target format.

Parameters:

  • preview format used as filename extension

  • (defaults to: nil)

    basename to use before the extension

Returns:

  • preview filename



263
264
265
266
# File 'lib/aspera/cli/plugins/preview.rb', line 263

def preview_filename(preview_format, base_name = nil)
  base_name ||= PREVIEW_BASENAME
  return "#{base_name}.#{preview_format}"
end

#process_events(iteration_persistency) ⇒ nil

Process recent Node API file events since the last persisted checkpoint.

Parameters:

  • stores the last processed event id

Returns:



148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
# File 'lib/aspera/cli/plugins/preview.rb', line 148

def process_events(iteration_persistency)
  # Restrict the event stream to file-related changes for the current access key.
  events_filter = {
    'access_key' => @access_key_self['id'],
    'type'       => 'file.*'
  }
  # Resume from the last persisted event id when available.
  events_filter['iteration_token'] = iteration_persistency.data.first unless iteration_persistency.nil?
  events = @api_node.read('events', events_filter)
  return if events.empty?
  events.each do |event|
    # Ignore non-file events such as folder notifications.
    if event.dig('data', 'type').eql?('file')
      file_entry = @api_node.read("files/#{event['data']['id']}") rescue nil
      if !file_entry.nil? &&
          @option_skip_folders.none? { |d| file_entry['path'].start_with?(d) }
        file_entry['parent_file_id'] = event['data']['parent_file_id']
        Log.log.error('TODO'.red) if event['types'].include?('file.deleted')
        generate_preview(file_entry) if event['types'].include?('file.deleted')
      end
    end
    # Periodically log progress and persist the latest processed event.
    next unless @periodic.trigger? || event.equal?(events.last)
    Log.log.debug { "Processing event #{event['id']}" }
    # Save a checkpoint to avoid replaying the full batch after a failure.
    if !iteration_persistency.nil?
      iteration_persistency.data[0] = event['id'].to_s
      iteration_persistency.save
    end
  end
  nil
end

#process_trevents(iteration_persistency) ⇒ nil

Process legacy transfer events and trigger preview generation for completed downloads.

Parameters:

  • stores the last processed event id

Returns:



102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
# File 'lib/aspera/cli/plugins/preview.rb', line 102

def process_trevents(iteration_persistency)
  events_filter = {
    'access_key' => @access_key_self['id'],
    'type'       => 'download.ended'
  }
  # Resume from the last persisted event id when available.
  events_filter['iteration_token'] = iteration_persistency.data.first unless iteration_persistency.nil?
  begin
    events = @api_node.read('events', events_filter)
  rescue RestCallError => e
    if e.message.include?('Invalid iteration_token')
      Log.log.warn { "Retrying without iteration token: #{e}" }
      events_filter.delete('iteration_token')
      retry
    end
    raise e
  end
  return if events.empty?
  events.each do |event|
    if event['data']['direction'].eql?(Transfer::Spec::DIRECTION_RECEIVE) &&
        event['data']['status'].eql?('completed') &&
        event['data']['error_code'].eql?(0) &&
        event['data'].dig('tags', Transfer::Spec::TAG_RESERVED, PREV_GEN_TAG).nil?
      folder_id = event.dig('data', 'tags', Transfer::Spec::TAG_RESERVED, 'node', 'file_id')
      folder_id ||= event.dig('data', 'file_id')
      if !folder_id.nil?
        folder_entry = @api_node.read("files/#{folder_id}") rescue nil
        scan_folder_files(folder_entry) unless folder_entry.nil?
      end
    end
    # Periodically log progress and persist the latest processed event.
    next unless @periodic.trigger? || event.equal?(events.last)
    Log.log.debug { "Processed event #{event['id']}" }
    # Save a checkpoint to avoid replaying the full batch after a failure.
    if !iteration_persistency.nil?
      iteration_persistency.data[0] = event['id'].to_s
      iteration_persistency.save
    end
  end
  nil
end

#scan_folder_files(top_entry, top_path = nil) ⇒ Object

Scan all files in provided folder entry

Parameters:

  • the top entry to scan

  • (defaults to: nil)

    subpath to start folder scan inside



348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
# File 'lib/aspera/cli/plugins/preview.rb', line 348

def scan_folder_files(top_entry, top_path = nil)
  unless top_path.nil?
    # canonical path: start with / and ends with /
    top_path = "/#{top_path.split('/').reject(&:empty?).join('/')}/"
  end
  Log.log.debug { "scan: #{top_entry} : #{top_path}".green }
  # don't use recursive call, use list instead
  entries_to_process = [top_entry]
  until entries_to_process.empty?
    entry = entries_to_process.shift
    # process this entry only if it is within the top_path
    entry_path_with_slash = entry['path']
    Log.log.debug { "processing entry #{entry_path_with_slash}" } if @periodic.trigger?
    entry_path_with_slash = "#{entry_path_with_slash}/" unless entry_path_with_slash.end_with?('/')
    if !top_path.nil? && !top_path.start_with?(entry_path_with_slash) && !entry_path_with_slash.start_with?(top_path)
      Log.log.debug { "#{entry['path']} folder (skip start)".bg_red }
      next
    end
    Log.log.debug { "item:#{entry}" }
    begin
      case entry['type']
      when 'file'
        if @filter_block.call(entry)
          generate_preview(entry)
        else
          Log.log.debug('skip by filter')
        end
      when 'link'
        Log.log.debug('Ignoring link.')
      when 'folder'
        if @option_skip_folders.include?(entry['path'])
          Log.log.debug { "#{entry['path']} folder (skip list)".bg_red }
        else
          Log.log.debug { "#{entry['path']} folder".green }
          # get folder content
          folder_entries = @api_node.read_folder_content(entry['id'])
          # process all items in current folder
          folder_entries.each do |folder_entry|
            # add path for older versions of ES
            folder_entry['path'] = entry_path_with_slash + folder_entry['name'] if !folder_entry.key?('path')
            folder_entry['parent_file_id'] = entry['id']
            entries_to_process.push(folder_entry)
          end
        end
      else
        Log.log.warn { "unknown entry type: #{entry['type']}" }
      end
    rescue StandardError => e
      Log.log.warn { "An error occurred: #{e}, ignoring" }
    end
  end
end

#setup_node_apiHash

Build the Node API and set up storage references for scan/events/trevents.

Returns:

  • empty ctx (state stored in instance variables)



420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
# File 'lib/aspera/cli/plugins/preview.rb', line 420

def setup_node_api(**)
  @api_node = Api::Node.new(**basic_auth_params)
  @transfer_server_address = URI.parse(@api_node.base_url).host
  # Get current access key information
  @access_key_self = @api_node.read('access_keys/self')
  # TODO: check events is activated here:
  # note that docroot is good to look at as well
  node_info = @api_node.read('info')
  Log.log.debug { "root: #{node_info['docroot']}" }
  # Default storage url to local file if not provided
  option_root_url = options.get_option(:root_url, mandatory: true)
  option_root_url = UriReader.file_url(@access_key_self['storage']['path']) if option_root_url.eql?(REMOTE_ACCESS) && @access_key_self['storage']['type'].eql?('local')
  @access_remote = !UriReader.file?(option_root_url)
  Log.log.debug { "remote: #{@access_remote}" }
  # TODO: can the `previews` folder parameter be read from Node API ?
  @option_skip_folders.push("/#{@option_previews_folder}")
  if @access_remote
    # NOTE: the filter "name", it's why we take the first one
    @previews_folder_entry = @api_node.read_folder_content(@access_key_self['root_file_id'], {name: @option_previews_folder}).first
    raise Cli::Error, "Folder #{@option_previews_folder} does not exist on node. " \
      'Please create it in the storage root, or specify an alternate name.' if @previews_folder_entry.nil?
  else
    @local_storage_root = UriReader.file_path(option_root_url)
    # TODO: Windows could have "C:" ?
    Aspera.assert(@local_storage_root.start_with?('/')) { "not local storage: #{@local_storage_root}" }
    Aspera.assert(File.directory?(@local_storage_root), type: Cli::Error) { "Local storage root folder #{@local_storage_root} does not exist." }
    @local_preview_folder = File.join(@local_storage_root, @option_previews_folder)
    Aspera.assert(File.directory?(@local_preview_folder), type: Cli::Error) { "Folder #{@local_preview_folder} does not exist locally. Please create it, or specify an alternate name." }
    # Protection to avoid clash of file id for two different access keys
    marker_file = File.join(@local_preview_folder, AK_MARKER_FILE)
    Log.log.debug { "marker file: #{marker_file}" }
    if File.exist?(marker_file)
      ak = File.read(marker_file).chomp
      Aspera.assert(@access_key_self['id'].eql?(ak)) { "mismatch access key in #{marker_file}: contains #{ak}, using #{@access_key_self['id']}" }
    else
      File.write(marker_file, @access_key_self['id'])
    end
  end
  {}
end