PDF Ingester

pdf_ingester_examples.rb

path = '/some/path/to/publication/sn99999999'
ingester = NewspaperWorks::Ingest::PDFIngester.new(path)
# if LCCN cannot be determined and validated from path, then raise an error
#   during construction above...
#     - Presumption: no character padding in LCCNs in either path or
#       in any command form.
#     - Normalize LCCN:
#       - Strip whitespace?
#       - Make lower case.
#     - Validate LCCN with regex:
#       - must begin with 1-3 digit alpha record prefix
#       - must contain 8 or 10 subsequent digits.
# def normalized_lccn(v)
#   p = /^[A-Za-z]{0,3}[0-9]{8}([0-9]{2})?$/
#   v = v.gsub(/\s+/, '').downcase.slice(0,13)
#   raise ArgumentError, "LCCN appears invalid: #{v}" unless p.match(v)
# end
ingester.lccn == 'sn99999999' # => true
ingester.issues.class == NewspaperWorks::Ingest::PDFIssues # => true
ingester.ingest

  # => Ensures title exists with proper metadata for LCCN, saved.
  # => Syncrhonously loops through each issue:
  #    1. creates a work for each,
  #    2. saved w/ issue metadata (publication_date, edition_number, lccn, title).
  #    3. Ensures administrative metadata is set to match opts
  #    4. Ensures that issue work is in publication.ordered_members
  #    5. Attaches PDF via:
  #         attachment = WorkFiles.of(work)
  #         attachment.assign(pdf_path)
  #         attachment.commit!
  #    6. queues CreateIssuePagesJob.perform_later(work, [pdf_path], nil, nil)
  #       - user and admin_set args are nil, becuase they are already to be
  #         set in step 3 above.
  #       - Most of this work is logged into sidekiq or similar job runner log.

# alternate construction:
ingester = NewspaperWorks::Ingest::PDFIngester.new(path, lccn='sn88888888')
# construction shall not attempt to care or parse LCCN from directory structure
ingester.ingest
# performs as above.

# The ingester will have an alternate constructor self.from_command,
# used by rake task; will supported rake task arguments via OptionParser:
#  --path: path to directory containing issues
#  --lccn: LCCN of pub
#  --admin_set: name of admin set
#  --depositor: name of user depositing
#  --visibility: visibility of

# Out of scope:
#  Each instance of PDFingester is tied to ONE PUBLICATION.
#  However, there might be a class method that could hanldle a rake task
#    pointed at a parent directory, containing multiple sub-dirs per LCCN.
#  This is TBD later. ^^^

# The interface for traversing issues, the PDFIssues class:
path = '/some/path/to/publication/sn99999999'
ingester = NewspaperWorks::Ingest::PDFIngester.new(path)
ingester.issues
# Get size
ingester.issues.size
# PDFIngester is enumerable of key (path) / value (PDFIssue object) pairs,
#   and presents a hash-like interface.
# Get paths:
ingester.issues.keys
# Enumerate over each issue:
ingester.issues.each do |path, issue|
  issue.lccn == ingester.lccn # => true
  issue.publication_date # => and ISO 8601 date stamp (from file name)
  issue.edition_number # => found edition number from file name or 1
  issue.title # title containing publication title, date, optionally edition #
end