moab-versioning 4.2.0 → 4.2.1

data/lib/moab/file_instance_difference.rb

@@ -1,7 +1,6 @@
 require 'moab'
 
 module Moab
-
   # A container for recording difference information at the file level
   # * If there was no change, the change type is set to <i>identical</i>
   # * If the signature is unchanged, but the path has moved, the change type is set to <i>renamed</i>
@@ -21,15 +20,14 @@ module Moab
   #
   # @note Copyright (c) 2012 by The Board of Trustees of the Leland Stanford Junior University.
   #   All rights reserved.  See {file:LICENSE.rdoc} for details.
-  class FileInstanceDifference  < Serializer::Serializable
-
+  class FileInstanceDifference < Serializer::Serializable
     include HappyMapper
 
     # The name of the XML element used to serialize this objects data
     tag 'file'
 
     # (see Serializable#initialize)
-    def initialize(opts={})
+    def initialize(opts = {})
       @signatures = Array.new
       super(opts)
     end
@@ -49,7 +47,5 @@ module Moab
     # @attribute
     # @return [Array<FileSignature>] The fixity data of the file manifestation(s) (plural if change was a content modification)
     has_many :signatures, FileSignature, :tag => 'fileSignature'
-
   end
-
 end

data/lib/moab/file_inventory.rb

@@ -1,7 +1,6 @@
 require 'moab'
 
 module Moab
-
   # A structured container for recording information about a collection of related files.
   #
   # The <b>scope</b> of the file collection depends on inventory type:
@@ -29,14 +28,13 @@ module Moab
   # @note Copyright (c) 2012 by The Board of Trustees of the Leland Stanford Junior University.
   #   All rights reserved.  See {file:LICENSE.rdoc} for details.
   class FileInventory < Serializer::Manifest
-
     include HappyMapper
 
     # The name of the XML element used to serialize this object's data
     tag 'fileInventory'
 
     # (see Serializable#initialize)
-    def initialize(opts={})
+    def initialize(opts = {})
       @groups = Array.new
       @inventory_datetime = Time.now
       super(opts)
@@ -106,7 +104,7 @@ module Moab
 
     # @param non_empty [Boolean] if true, return group_id's only for groups having files
     # @return [Array<String>] group identifiers contained in this file inventory
-    def group_ids(non_empty=nil)
+    def group_ids(non_empty = nil)
       my_groups = non_empty ? self.non_empty_groups : groups
       my_groups.map { |g| g.group_id }
     end
@@ -114,7 +112,7 @@ module Moab
     # @param [String] group_id The identifer of the group to be selected
     # @return [FileGroup] The file group in this inventory for the specified group_id
     def group(group_id)
-      groups.find { |group| group.group_id == group_id}
+      groups.find { |group| group.group_id == group_id }
     end
 
     # @param group_id [String] File group identifer (e.g. data, metadata, manifests)
@@ -183,7 +181,7 @@ module Moab
     #   if nil, then the directory is assumed to contain both content and metadata subdirectories
     # @return [FileInventory] Traverse a directory and return an inventory of the files it contains
     # @example {include:file:spec/features/inventory/harvest_inventory_spec.rb}
-    def inventory_from_directory(data_dir, group_id=nil)
+    def inventory_from_directory(data_dir, group_id = nil)
       if group_id
         groups << FileGroup.new(group_id: group_id).group_from_directory(data_dir)
       else
@@ -202,7 +200,7 @@ module Moab
       signatures_from_bag = signatures_from_bagit_manifests(bag_pathname)
       bag_data_subdirs = bag_pathname.join('data').children
       bag_data_subdirs.each do |subdir|
-        groups << FileGroup.new(:group_id=>subdir.basename.to_s).group_from_bagit_subdir(subdir, signatures_from_bag)
+        groups << FileGroup.new(:group_id => subdir.basename.to_s).group_from_bagit_subdir(subdir, signatures_from_bag)
       end
       self
     end
@@ -219,7 +217,7 @@ module Moab
         if manifest_pathname[type].exist?
           manifest_pathname[type].each_line do |line|
             line.chomp!
-            checksum,data_path = line.split(/\s+\**/,2)
+            checksum, data_path = line.split(/\s+\**/, 2)
             if checksum && data_path
               file_pathname = bag_pathname.join(data_path)
               signature = signatures[file_pathname]
@@ -251,18 +249,18 @@ module Moab
     # @api internal
     # @param type [String] Specifies the type of inventory, and thus the filename used for storage
     # @return [String] The standard name for the serialized inventory file of the given type
-    def self.xml_filename(type=nil)
+    def self.xml_filename(type = nil)
       case type
-        when "version"
-          'versionInventory.xml'
-        when "additions"
-          'versionAdditions.xml'
-        when "manifests"
-          'manifestInventory.xml'
-        when "directory"
-          'directoryInventory.xml'
-        else
-          raise ArgumentError, "unknown inventory type: #{type}"
+      when "version"
+        'versionInventory.xml'
+      when "additions"
+        'versionAdditions.xml'
+      when "manifests"
+        'manifestInventory.xml'
+      when "directory"
+        'directoryInventory.xml'
+      else
+        raise ArgumentError, "unknown inventory type: #{type}"
       end
     end
 
@@ -271,11 +269,9 @@ module Moab
     # @param type [String] The inventory type, which governs the filename used for serialization
     # @return [void] write the {FileInventory} instance to a file
     # @example {include:file:spec/features/inventory/write_inventory_xml_spec.rb}
-    def write_xml_file(parent_dir, type=nil)
+    def write_xml_file(parent_dir, type = nil)
       type = @type if type.nil?
       self.class.write_xml_file(self, parent_dir, type)
     end
-
   end
-
 end

data/lib/moab/file_inventory_difference.rb

@@ -1,7 +1,6 @@
 require 'moab'
 
 module Moab
-
   # Compares two {FileInventory} instances based primarily on file signatures and secondarily on file pathnames.
   # Although the usual use will be to compare the content of 2 different temporal versions of the same object,
   # it can also be used to verify an inventory document against an inventory harvested from a directory.
@@ -18,14 +17,13 @@ module Moab
   # @note Copyright (c) 2012 by The Board of Trustees of the Leland Stanford Junior University.
   #   All rights reserved.  See {file:LICENSE.rdoc} for details.
   class FileInventoryDifference < Serializer::Manifest
-
     include HappyMapper
 
     # The name of the XML element used to serialize this objects data
     tag 'fileInventoryDifference'
 
     # (see Serializable#initialize)
-    def initialize(opts={})
+    def initialize(opts = {})
       @group_differences = Array.new
       super(opts)
     end
@@ -36,7 +34,7 @@ module Moab
 
     # @attribute
     # @return [Integer] the number of differences found between the two inventories that were compared (dynamically calculated)
-    attribute :difference_count, Integer, :tag=> 'differenceCount',:on_save => Proc.new {|i| i.to_s}
+    attribute :difference_count, Integer, :tag => 'differenceCount', :on_save => Proc.new { |i| i.to_s }
 
     def difference_count
       @group_differences.inject(0) { |sum, group| sum + group.difference_count }
@@ -74,7 +72,7 @@ module Moab
     # @param [String] group_id The identifer of the group to be selected
     # @return [FileGroupDifference] The subset of this report for the specified group_id (or nil if not found)
     def group_difference(group_id)
-      @group_differences.find{ |group_difference| group_difference.group_id == group_id}
+      @group_differences.find { |group_difference| group_difference.group_id == group_id }
     end
 
     # @api external
@@ -126,7 +124,5 @@ module Moab
       end
       inv_diff
     end
-
   end
-
 end

data/lib/moab/file_manifestation.rb

@@ -1,7 +1,6 @@
 require 'moab'
 
 module Moab
-
   # A container for a file signature and all the physical file instances that have that signature
   # This element has one child {FileSignature} element, and one or more {FileInstance} elements
   # Regarding the class name, see
@@ -24,7 +23,7 @@ module Moab
     tag 'file'
 
     # (see Serializable#initialize)
-    def initialize(opts={})
+    def initialize(opts = {})
       @instances = Array.new
       super(opts)
     end
@@ -48,7 +47,7 @@ module Moab
     # @api internal
     # @return [Array<String>] Create an array from all the file paths of the child {FileInstance} objects
     def paths
-      instances.collect { |i| i.path}
+      instances.collect { |i| i.path }
     end
 
     # @api internal
@@ -69,7 +68,7 @@ module Moab
     #   (estimating du -k result)
     def block_count
       block_size = 1024
-      instance_blocks = (signature.size.to_i + block_size - 1)/block_size
+      instance_blocks = (signature.size.to_i + block_size - 1) / block_size
       file_count * instance_blocks
     end
 
@@ -80,7 +79,5 @@ module Moab
       return false unless (other.respond_to?(:signature) && other.respond_to?(:instances)) # Cannot equal an incomparable type!
       (self.signature == other.signature) && (self.instances == other.instances)
     end
-
   end
-
 end

data/lib/moab/file_signature.rb

@@ -1,7 +1,6 @@
 require 'moab'
 
 module Moab
-
   # The fixity properties of a file, used to determine file content equivalence regardless of filename.
   # Placing this data in a class by itself facilitates using file size together with the MD5 and SHA1 checksums
   # as a single key when doing comparisons against other file instances.  The Moab design assumes that this file signature
@@ -40,17 +39,11 @@ module Moab
   # @note Copyright (c) 2012 by The Board of Trustees of the Leland Stanford Junior University.
   #   All rights reserved.  See {file:LICENSE.rdoc} for details.
   class FileSignature < Serializer::Serializable
-
     include HappyMapper
 
     # The name of the XML element used to serialize this objects data
     tag 'fileSignature'
 
-    # (see Serializable#initialize)
-    def initialize(opts={})
-      super(opts)
-    end
-
     # @attribute
     # @return [Integer] The size of the file in bytes
     attribute :size, Integer, :on_save => Proc.new { |n| n.to_s }
@@ -67,19 +60,47 @@ module Moab
     # @return [String] The SHA256 checksum value of the file
     attribute :sha256, String, :on_save => Proc.new { |n| n.nil? ? "" : n.to_s }
 
+    KNOWN_ALGOS = {
+      md5: proc { Digest::MD5.new },
+      sha1: proc { Digest::SHA1.new },
+      sha256: proc { Digest::SHA2.new(256) }
+    }.freeze
+
+    def self.active_algos
+      Moab::Config.checksum_algos
+    end
+
+    # Reads the file once for ALL (requested) algorithms, not once per.
+    # @param [Pathname] pathname
+    # @param [Array<Symbol>] one or more keys of KNOWN_ALGOS to be computed
+    # @return [Moab::FileSignature] object populated with (requested) checksums
+    def self.from_file(pathname, algos_to_use = active_algos)
+      raise 'Unrecognized algorithm requested' unless algos_to_use.all? { |a| KNOWN_ALGOS.include?(a) }
+
+      signatures = algos_to_use.map { |k| [k, KNOWN_ALGOS[k].call] }.to_h
+
+      pathname.open("r") do |stream|
+        while (buffer = stream.read(8192))
+          signatures.each_value { |digest| digest.update(buffer) }
+        end
+      end
+
+      new(signatures.map { |k, digest| [k, digest.hexdigest] }.to_h.merge(size: pathname.size))
+    end
+
     # @param type [Symbol,String] The type of checksum
     # @param value [String] The checksum value
     # @return [void] Set the value of the specified checksum type
-    def set_checksum(type,value)
+    def set_checksum(type, value)
       case type.to_s.downcase.to_sym
-        when :md5
-          @md5 = value
-        when :sha1
-          @sha1 = value
-        when :sha256
-          @sha256 = value
-        else
-          raise ArgumentError, "Unknown checksum type '#{type}'"
+      when :md5
+        @md5 = value
+      when :sha1
+        @sha1 = value
+      when :sha256
+        @sha256 = value
+      else
+        raise ArgumentError, "Unknown checksum type '#{type}'"
       end
     end
 
@@ -141,24 +162,17 @@ module Moab
       @size.to_i
     end
 
-    # @api internal
+    # @deprecated
+    # this method is a holdover from an earlier version.  use the class method .from_file going forward.
+    # @api external
     # @param pathname [Pathname] The location of the file to be digested
     # @return [FileSignature] Generate a FileSignature instance containing size and checksums for a physical file
     def signature_from_file(pathname)
-      @size = pathname.size
-      md5_digest = Digest::MD5.new
-      sha1_digest = Digest::SHA1.new
-      sha256_digest = Digest::SHA2.new(256)
-      pathname.open("r") do |stream|
-        while buffer = stream.read(8192)
-          md5_digest.update(buffer)
-          sha1_digest.update(buffer)
-          sha256_digest.update(buffer)
-        end
-      end
-      @md5 = md5_digest.hexdigest
-      @sha1 = sha1_digest.hexdigest
-      @sha256 = sha256_digest.hexdigest
+      file_signature = self.class.from_file(pathname)
+      self.size = file_signature.size
+      self.md5 = file_signature.md5
+      self.sha1 = file_signature.sha1
+      self.sha256 = file_signature.sha256
       self
     end
 
@@ -195,7 +209,5 @@ module Moab
       end
       type_for_name
     end
-
   end
-
 end

data/lib/moab/signature_catalog.rb

@@ -1,7 +1,6 @@
 require 'moab'
 
 module Moab
-
   # A digital object's Signature Catalog is derived from an filtered aggregation of the file inventories
   # of a digital object's set of versions.  (see {#update})
   # It has an entry for every file (identified by {FileSignature}) found in any of the versions,
@@ -32,7 +31,7 @@ module Moab
     tag 'signatureCatalog'
 
     # (see Serializable#initialize)
-    def initialize(opts={})
+    def initialize(opts = {})
       @entries = Array.new
       @signature_hash = Hash.new
       super(opts)
@@ -44,7 +43,7 @@ module Moab
 
     # @attribute
     # @return [Integer] The ordinal version number
-    attribute :version_id, Integer, :tag => 'versionId', :key => true, :on_save => Proc.new {|n| n.to_s}
+    attribute :version_id, Integer, :tag => 'versionId', :key => true, :on_save => Proc.new { |n| n.to_s }
 
     # @return [String] The unique identifier concatenating digital object id with version id
     def composite_key
@@ -56,7 +55,7 @@ module Moab
     attribute :catalog_datetime, Time, :tag => 'catalogDatetime'
 
     def catalog_datetime=(datetime)
-      @catalog_datetime=Moab::UtcTime.input(datetime)
+      @catalog_datetime = Moab::UtcTime.input(datetime)
     end
 
     def catalog_datetime
@@ -65,7 +64,7 @@ module Moab
 
     # @attribute
     # @return [Integer] The total number of data files (dynamically calculated)
-    attribute :file_count, Integer, :tag => 'fileCount', :on_save => Proc.new {|t| t.to_s}
+    attribute :file_count, Integer, :tag => 'fileCount', :on_save => Proc.new { |t| t.to_s }
 
     def file_count
       entries.size
@@ -73,7 +72,7 @@ module Moab
 
     # @attribute
     # @return [Integer] The total size (in bytes) of all data files (dynamically calculated)
-    attribute :byte_count, Integer, :tag => 'byteCount', :on_save => Proc.new {|t| t.to_s}
+    attribute :byte_count, Integer, :tag => 'byteCount', :on_save => Proc.new { |t| t.to_s }
 
     def byte_count
       entries.inject(0) { |sum, entry| sum + entry.signature.size.to_i }
@@ -81,11 +80,11 @@ module Moab
 
     # @attribute
     # @return [Integer] The total disk usage (in 1 kB blocks) of all data files (estimating du -k result) (dynamically calculated)
-    attribute :block_count, Integer, :tag => 'blockCount', :on_save => Proc.new {|t| t.to_s}
+    attribute :block_count, Integer, :tag => 'blockCount', :on_save => Proc.new { |t| t.to_s }
 
     def block_count
-      block_size=1024
-      entries.inject(0) { |sum, entry| sum + (entry.signature.size.to_i + block_size - 1)/block_size }
+      block_size = 1024
+      entries.inject(0) { |sum, entry| sum + (entry.signature.size.to_i + block_size - 1) / block_size }
     end
 
     # @return [Array<String>] The data fields to include in summary reports
@@ -128,8 +127,8 @@ module Moab
     # @param group [FileGroup] A group of the files from a file inventory
     # @param group_pathname [Pathname] The location of the directory containing the group's files
     # @return [void] Inspect and upgrade the group's signature data to include all desired checksums
-    def normalize_group_signatures(group, group_pathname=nil)
-      unless  group_pathname.nil?
+    def normalize_group_signatures(group, group_pathname = nil)
+      unless group_pathname.nil?
         group_pathname = Pathname(group_pathname)
         raise "Could not locate #{group_pathname}" unless group_pathname.exist?
       end
@@ -162,7 +161,7 @@ module Moab
             if file.signature.complete?
               entry.signature = file.signature
             else
-              file_pathname = data_pathname.join(group.group_id,entry.path)
+              file_pathname = data_pathname.join(group.group_id, entry.path)
               entry.signature = file.signature.normalized_signature(file_pathname)
             end
             add_entry(entry)
@@ -179,20 +178,18 @@ module Moab
     #   containing only those files that were added in this version
     # @example {include:file:spec/features/catalog/version_additions_spec.rb}
     def version_additions(version_inventory)
-      version_additions = FileInventory.new(:type=>'additions')
+      version_additions = FileInventory.new(:type => 'additions')
       version_additions.copy_ids(version_inventory)
       version_inventory.groups.each do |group|
         group_addtions = FileGroup.new(:group_id => group.group_id)
         group.files.each do |file|
           unless @signature_hash.has_key?(file.signature)
-            group_addtions.add_file_instance(file.signature,file.instances[0])
+            group_addtions.add_file_instance(file.signature, file.instances[0])
           end
         end
         version_additions.groups << group_addtions if group_addtions.files.size > 0
       end
       version_additions
     end
-
   end
-
 end