sdr-replication 0.1.0

data/lib/replication/operating_system.rb

@@ -0,0 +1,33 @@
+require File.join(File.dirname(__FILE__),'../libdir')
+require 'sdr_replication'
+
+module Replication
+
+  # A wrapper class around the systemu gem that is used for shelling out to the operating system
+  # and executing a command
+  #
+  # @note Copyright (c) 2014 by The Board of Trustees of the Leland Stanford Junior University.
+  #   All rights reserved.  See {file:LICENSE.rdoc} for details.
+  class OperatingSystem
+
+    # Executes a system command in a subprocess.
+    # The method will return stdout from the command if execution was successful.
+    # The method will raise an exception if if execution fails.
+    # The exception's message will contain the explaination of the failure.
+    # @param [String] command the command to be executed
+    # @return [String] stdout from the command if execution was successful
+    def OperatingSystem.execute(command)
+      status, stdout, stderr = systemu(command)
+      if (status.exitstatus != 0)
+        raise stderr
+      end
+      return stdout
+    rescue
+      msg = "Command failed to execute: [#{command}] caused by <STDERR = #{stderr.split($/).join('; ')}>"
+      msg << " STDOUT = #{stdout.split($/).join('; ')}" if (stdout && (stdout.length > 0))
+      raise msg
+    end
+
+  end
+
+end

data/lib/replication/replica.rb

@@ -0,0 +1,62 @@
+require File.join(File.dirname(__FILE__),'../libdir')
+require 'sdr_replication'
+
+module Replication
+
+  # The metadata concerning the digital object/version that is the subject of replication.
+  #
+  # @note Copyright (c) 2014 by The Board of Trustees of the Leland Stanford Junior University.
+  #   All rights reserved.  See {file:LICENSE.rdoc} for details.
+  class Replica
+
+    @@replica_cache_pathname = nil
+
+    # @return [Pathname] The base location of the replica cache
+    def Replica.replica_cache_pathname
+      @@replica_cache_pathname
+    end
+
+    # @return [Pathname] Set the base location of the replica cache
+    # @param [Pathname,String] replica_cache_pathname The base location of the replica cache
+    def Replica.replica_cache_pathname=(replica_cache_pathname)
+      @@replica_cache_pathname = Pathname(replica_cache_pathname)
+    end
+
+    # @return [String] The unique identifier for the digital object replica
+    attr_accessor :replica_id
+
+    # @return [String] The original home location of the replica (sdr or dpn)
+    attr_accessor :home_repository
+
+    # @return [Time] The timestamp of the datetime at which the replica was created
+    attr_accessor :create_date
+
+    # @return [BagitBag] A bag containing a copy of the replica
+    attr_accessor :bagit_bag
+
+    # @return [Integer] The size (in bytes) of the replic bag's payload
+    attr_accessor :payload_size
+
+    # @return [String] The type of checksum/digest type (:sha1, :sha256)
+    attr_accessor :payload_fixity_type
+
+    # @return [String] The value of the checksum/digest
+    attr_accessor :payload_fixity
+
+    # @param [String] replica_id  The unique identifier for the digital object replica
+    # @param [String] home_repository The original home location of the replica (sdr or dpn)
+    # @return [Replica] Initialize a new Replica object
+    def initialize(replica_id, home_repository)
+      @replica_id = replica_id
+      @home_repository = home_repository
+    end
+
+    # @return [Pathname] The location of the replica bag
+    def replica_pathname
+      @replica_pathname ||= @@replica_cache_pathname.join(home_repository,replica_id)
+    end
+
+  end
+end
+
+

data/lib/replication/sdr_object_version.rb

@@ -0,0 +1,63 @@
+require File.join(File.dirname(__FILE__),'../libdir')
+require 'sdr_replication'
+
+module Replication
+
+  #
+  # @note Copyright (c) 2014 by The Board of Trustees of the Leland Stanford Junior University.
+  #   All rights reserved.  See {file:LICENSE.rdoc} for details.
+  class SdrObjectVersion
+
+    # @return [Moab::StorageObjectVersion] Represents the object version's storage location
+    attr_accessor :moab_object_version
+
+    # @param [Moab::StorageObjectVersion] object_version Represents the object version's storage location
+    # @return [SdrObjectVersion] Initialize a new SdrObjectVersion object
+    def initialize(object_version)
+      @moab_object_version = object_version
+    end
+
+    # @return [String] The digital object identifier (druid)
+    def sdr_object_id
+      @sdr_object_id ||=  moab_object_version.storage_object.digital_object_id
+    end
+
+    # @return [Integer] The digital object version number
+    def sdr_version_id
+      @sdr_version_id ||= moab_object_version.version_id
+    end
+
+    # @return [Moab::FileInventory] The moab version manifest for the version
+    def version_inventory
+      @version_inventory ||= moab_object_version.file_inventory('version')
+    end
+
+    # @return [Moab::FileInventory] The moab version manifest for the version
+    def version_additions
+      @version_additions ||= moab_object_version.file_inventory('additions')
+    end
+
+    # @return [String] The unique identifier for the digital object replica
+    def replica_id
+      @replica_id ||= "#{sdr_object_id.split(':').last}-#{moab_object_version.version_name}"
+    end
+
+    # @return [Replica] The Replica of the object version that is archived to tape, etc
+    def replica
+      @replica ||= Replica.new(replica_id, 'sdr')
+    end
+
+    # @return [BagitBag] Copy the object version into a BagIt Bag in tarfile format
+    def moab_to_replica_bag
+      bag_dir = replica.replica_pathname
+      bag = BagitBag.create_bag(bag_dir)
+      bag.bag_checksum_types = [:sha256]
+      bag.add_payload_tarfile("#{replica_id}.tar",moab_object_version.version_pathname, moab_object_version.storage_object.object_pathname.parent)
+      bag.write_bag_info_txt
+      bag.write_manifest_checksums('tagmanifest', bag.generate_tagfile_checksums)
+      bag
+    end
+
+  end
+
+end

data/lib/replication/tarfile.rb

@@ -0,0 +1,160 @@
+require File.join(File.dirname(__FILE__),'../libdir')
+require 'sdr_replication'
+
+module Replication
+
+  # A tar archive file containing a set of digital object files
+  #
+  # @note Copyright (c) 2014 by The Board of Trustees of the Leland Stanford Junior University.
+  #   All rights reserved.  See {file:LICENSE.rdoc} for details.
+  class Tarfile
+
+    # @return [String] create archive of the specified format
+    # * gnu = GNU tar 1.13.x format
+    # * posix = POSIX 1003.1-2001 (pax) format
+    attr_accessor :format
+
+    # @return [Boolean] Follow symlinks and archive the files they point to
+    attr_accessor :dereference
+
+    # @return [Boolean] Verify that files were copied faithfully
+    attr_accessor :verify
+
+    # @return [Boolean] Create/list/extract multi-volume archive (not yet implemented)
+    attr_accessor :multi_volume
+
+    # @param [Hash<Symbol,Object>] options Key,Value pairs specifying initial values of attributes
+    # @return [Tarfile] Initialize a new Tarfile object
+    def initialize(options=nil)
+      options={} if options.nil?
+      # set defaults
+      @format=:posix
+      @dereference = true
+      @verify = false
+      @multi_volume = false
+      # override defaults
+      options.each do |key,value|
+        #instance_variable_set("@#{key}", value)
+        send "#{key}=", value
+      end
+    end
+
+    # @return [Pathname] The full path of the ancestor dir in which the tar file resides
+    def tarfile_basepath
+      raise "Tarfile basepath is nil" unless @tarfile_basepath
+      @tarfile_basepath
+    end
+
+    # @param [Pathname,String] basepath The full path of the ancestor dir in which the tar file resides
+    # @return [Void] Set the full path of the ancestor dir in which the tar file resides
+    def tarfile_basepath=(basepath)
+      raise "No pathname specified" unless basepath
+      @tarfile_basepath = Pathname(basepath).expand_path
+    end
+
+    # @return [Pathname] the full path of the tar archive file to be created or extracted from
+    def tarfile_fullpath
+      @tarfile_fullpath
+    end
+
+    # @param [Pathname,String] fullpath The full path of tar file
+    # @return [Void] Sets the full path of tar file
+    def tarfile_fullpath=(fullpath)
+      @tarfile_fullpath = Pathname(fullpath).expand_path
+    end
+
+    # @return [String] The id (path relative to basepath) of the tar file
+    def tarfile_relative_path
+      @tarfile_fullpath.relative_path_from(@tarfile_basepath).to_s
+    end
+
+    # @return [Pathname] The full path of the source file or directory being archived
+    def source_fullpath
+      raise "Source pathname is nil" unless @source_pathname
+      @source_pathname
+    end
+
+    # @param [Pathname,String] source The full path of the source file or directory being archived
+    # @return [Void] Set the full path of the source file or directory being archived
+    def source_fullpath=(source)
+      raise "No pathname specified" unless source
+      @source_pathname = Pathname(source).expand_path
+    end
+
+    # @return [Pathname] The directory that is the basis of relative paths
+    def source_basepath
+      @source_basepath
+    end
+
+    # @param [Pathname,String] base The directory that is the basis of relative paths
+    # @return [Void] Set the base path of the source file or directory being archived
+    def source_basepath=(base)
+      raise "No pathname specified" unless base
+      @source_basepath = Pathname(base).expand_path
+    end
+
+    # @return [Pathname] The relative path from the source base directory to the source directory
+    def source_relative_path
+        source_fullpath.relative_path_from(source_basepath)
+    end
+
+    # @return [String] The shell command string to be used to create the tarfile
+    def create_cmd
+      command = "tar --create --file=#{tarfile_fullpath} --format=#{@format} "
+      command << "--dereference " if @dereference
+      command << "--verify " if @verify
+      command << "--directory='#{source_basepath}' " if source_basepath
+      command << source_relative_path.to_s
+      command
+    end
+
+    # @return [Tarfile] Shell out to the operating system and create the tar archive file
+    def create_tarfile
+      command = create_cmd
+      OperatingSystem.execute(command)
+      self
+    end
+
+    # @return [String] The shell command that will list the tarfile's contents
+    def list_cmd
+      command = "tar --list --file=#{tarfile_fullpath} "
+      command
+    end
+
+    # @return [String] The list of the tarfile's contents
+    def list_tarfile
+      command = list_cmd
+      list = OperatingSystem.execute(command)
+      list
+    end
+
+    # @return [Pathname] The location of the directory into which the tarfile should be extracted
+    def target_pathname
+      raise "Target pathname is nil" unless @target_pathname
+      @target_pathname
+    end
+
+    # @param [Pathname,String] source The location of the directory into which the tarfile should be extracted
+    # @return [Void] Set the location of the directory into which the tarfile should be extracted
+    def target_pathname=(target)
+      raise "No target pathname specified" unless target
+      @target_pathname = Pathname(target).expand_path
+    end
+
+    # @return [String] The shell command that will extract the tarfile's contents    # @return [Void]
+    def extract_cmd
+      command = "tar --extract --file=#{tarfile_fullpath} "
+      command << "--directory='#{target_pathname}' " if target_pathname
+      command
+    end
+
+    # @return [String] Shell out to the operating system and extract the tar archive file
+    def extract_tarfile
+      command = extract_cmd
+      stdout = OperatingSystem.execute(command)
+      stdout
+    end
+
+  end
+
+end

data/lib/sdr_replication.rb

@@ -0,0 +1,26 @@
+require 'rubygems'
+require 'bundler/setup'
+Bundler.setup
+require 'digest'
+require 'find'
+require 'json/pure'
+require 'moab_stanford'
+require 'pathname'
+require 'rest-client'
+require 'systemu'
+
+
+# The classes used for SDR Replication workflows
+module Replication
+end
+
+require 'replication/archive_catalog'
+require 'replication/bagit_bag'
+require 'replication/file_fixity'
+require 'replication/fixity'
+require 'replication/operating_system'
+require 'replication/replica'
+require 'replication/sdr_object_version'
+require 'replication/tarfile'
+include Replication
+

metadata

@@ -0,0 +1,198 @@
+--- !ruby/object:Gem::Specification
+name: sdr-replication
+version: !ruby/object:Gem::Version
+  version: 0.1.0
+platform: ruby
+authors:
+- Richard Anderson
+autorequire: 
+bindir: bin
+cert_chain: []
+date: 2014-06-24 00:00:00.000000000 Z
+dependencies:
+- !ruby/object:Gem::Dependency
+  name: json_pure
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
+- !ruby/object:Gem::Dependency
+  name: moab-versioning
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
+- !ruby/object:Gem::Dependency
+  name: systemu
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
+- !ruby/object:Gem::Dependency
+  name: rest-client
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
+- !ruby/object:Gem::Dependency
+  name: awesome_print
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
+- !ruby/object:Gem::Dependency
+  name: equivalent-xml
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
+- !ruby/object:Gem::Dependency
+  name: fakeweb
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
+- !ruby/object:Gem::Dependency
+  name: rspec
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: 2.14.1
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: 2.14.1
+- !ruby/object:Gem::Dependency
+  name: simplecov
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: 0.7.1
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: 0.7.1
+- !ruby/object:Gem::Dependency
+  name: yard
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
+description: Contains classes to archive and retrieve digital object version content
+  and metadata
+email:
+- rnanders@stanford.edu
+executables: []
+extensions: []
+extra_rdoc_files: []
+files:
+- lib/libdir.rb
+- lib/replication/archive_catalog.rb
+- lib/replication/bagit_bag.rb
+- lib/replication/command_consumer.rb
+- lib/replication/command_producer.rb
+- lib/replication/dpn_check_rep.rb
+- lib/replication/file_fixity.rb
+- lib/replication/fixity.rb
+- lib/replication/operating_system.rb
+- lib/replication/replica.rb
+- lib/replication/sdr_object_version.rb
+- lib/replication/tarfile.rb
+- lib/sdr_replication.rb
+homepage: 
+licenses: []
+metadata: {}
+post_install_message: 
+rdoc_options: []
+require_paths:
+- lib
+required_ruby_version: !ruby/object:Gem::Requirement
+  requirements:
+  - - ">="
+    - !ruby/object:Gem::Version
+      version: '0'
+required_rubygems_version: !ruby/object:Gem::Requirement
+  requirements:
+  - - ">="
+    - !ruby/object:Gem::Version
+      version: 2.2.1
+requirements: []
+rubyforge_project: 
+rubygems_version: 2.2.1
+signing_key: 
+specification_version: 4
+summary: Core methods for support of SDR Preservation Core replication
+test_files: []
+has_rdoc: