libis-metadata 0.2.0

data/lib/libis/metadata/mappers/scope.rb

@@ -0,0 +1,46 @@
+# encoding: utf-8
+
+require 'libis/tools/metadata/dublin_core_record'
+require 'libis/tools/assert'
+
+module Libis
+  module Tools
+    module Metadata
+      module Mappers
+        # noinspection RubyResolve
+
+        # Mixin for {::Libis::Tools::Metadata::DublinCoreRecord} to enable conversion of the Scope exported DC record.
+        module Scope
+
+          # Main conversion method.
+          # @return [::Libis::Tools::Metadata::DublinCoreRecord]
+          def to_dc
+            assert(self.is_a? Libis::Tools::Metadata::DublinCoreRecord)
+
+            doc = Libis::Tools::Metadata::DublinCoreRecord.new(self.to_xml)
+
+            if doc.isPartOf
+
+              # create new node for isReferencedBy
+              new_node = doc.add_node(
+                  'isReferencedBy',
+                  doc.isPartOf.content,
+                  nil,
+                  'xsi:type' => 'dcterms:URI'
+              )
+
+              # Replace isPartOf with isReferencedBy
+              doc.isPartOf.replace new_node
+
+            end
+
+            doc
+
+          end
+
+        end
+
+      end
+    end
+  end
+end

data/lib/libis/metadata/marc21_record.rb

@@ -0,0 +1,49 @@
+# coding: utf-8
+
+require 'cgi'
+
+require_relative 'marc_record'
+
+module Libis
+  module Metadata
+
+    # This class implements the missing private method 'get_all_records' to accomodate for the MARC-XML format.
+    class Marc21Record < Libis::Metadata::MarcRecord
+
+      private
+
+      def get_all_records
+
+        @all_records.clear
+
+        @node.xpath('.//leader').each {|f|
+          @all_records['LDR'] << FixField.new('LDR', f.content)
+        }
+
+        @node.xpath('.//controlfield').each {|f|
+          tag = f['tag']
+          tag = '%03d' % tag.to_i if tag.size < 3
+          @all_records[tag] << FixField.new(tag, f.content)
+        }
+
+        @node.xpath('.//datafield').each {|v|
+
+          tag = v['tag']
+          tag = '%03d' % tag.to_i if tag.size < 3
+
+          varfield = VarField.new(tag, v['ind1'].to_s, v['ind2'].to_s)
+
+          v.xpath('.//subfield').each {|s| varfield.add_subfield(s['code'], s.content)}
+
+          @all_records[tag] << varfield
+
+        }
+
+        @all_records
+
+      end
+
+    end
+
+  end
+end

data/lib/libis/metadata/marc_record.rb

@@ -0,0 +1,285 @@
+# coding: utf-8
+
+require 'set'
+require 'cgi'
+
+require 'libis/tools/xml_document'
+require 'libis/tools/assert'
+
+require_relative 'fix_field'
+require_relative 'var_field'
+require_relative 'field_format'
+
+module Libis
+  module Metadata
+
+    # noinspection RubyTooManyMethodsInspection
+
+    # Base class for reading MARC based records.
+    #
+    # For indicator selection: '#' or '' (empty) is wildcard; '_' or ' ' (space) is blank.
+    class MarcRecord
+
+      # Create a new MarcRecord object
+      #
+      # @param [XML node] xml_node XML node from Nokogiri or XmlDocument that contains child nodes with the data for
+      #         one MARC record.
+      def initialize(xml_node)
+        @node = xml_node
+        @node.document.remove_namespaces!
+        @all_records = Hash.new {|h, k| h[k] = Array.new}
+      end
+
+      # Access to the XML node that was supplied to the constructor
+      # @return [XML node]
+      def to_raw
+        @node
+      end
+
+      # Returns the internal data structure (a Hash) with all the MARC data.
+      #
+      # The internal structure is a Hash with the tag as key and as value an Array of either FixField or VarField
+      # instances.
+      #
+      # @return [Hash] internal data structure
+      def all
+        return @all_records unless @all_records.empty?
+        @all_records = get_all_records
+      end
+
+      # Iterates over all the MARC fields.
+      #
+      # If a block is supplied it will be called for each field in the MARC record. The supplied argument will be the
+      # FixField or VarField instance for each field.
+      #
+      # @return [Array] The list of the field data or return values for each block call.
+      def each
+        all.map {|_, field_array| field_array}.flatten.map do |field|
+          block_given? ? yield(field) : field
+        end
+      end
+
+      # Get all fields matching search criteria.
+      #
+      # A block with one parameter can be supplied when calling this method. Each time a match is found, the block
+      # will be called with the field data as argument and the return value of the block will be added to the method's
+      # return value. This could for example be used to narrow the selection of the fields:
+      #
+      #     # Only select 700 tags where $4 subfield contains 'abc', 'def' or 'xyz'
+      #     record.all_tags('700') { |v| v.subfield['4'] =~ /^(abc|def|xyz)$/ ? v : nil }.compact
+      #
+      # @param [String] tag Tag selection string. Tag name with indicators, '#' for wildcard, '_' for blank. If an
+      #     extra subfield name is added, a result will be created for each instance found of that subfield.
+      # @param [String] subfields Subfield specification. See FieldFormat class for more info; ignored for controlfields.
+      # @param [Proc] select_block block that will be executed once for each field found. The block takes one argument
+      #     (the field) and should return true or false. True selects the field, false rejects it.
+      # @return [Array] If a block was supplied to the method call, the array will contain the result of the block
+      #     for each tag found. Otherwise the array will just contain the data for each matching tag.
+      def all_tags(tag, subfields = '', select_block = Proc.new {|_| true})
+        t, ind1, ind2, subfield = tag =~ /^\d{3}/ ? [tag[0..2], tag[3], tag[4], tag[5]] : [tag, nil, nil, nil]
+        result = get_records(t, ind1, ind2, subfield, subfields, &select_block)
+        return result unless block_given?
+        result.map {|record| yield record}
+      end
+
+      alias_method :each_tag, :all_tags
+
+      # Get all fields matching search criteria.
+      # As {#all_tags} but without subfield criteria.
+      # @param [String] tag Tag selection string. Tag name with indicators, '#' for wildcard, '_' for blank. If an
+      #     extra subfield name is added, a result will be created for each instance found of that subfield.
+      # @param [Proc] select_block block that will be executed once for each field found. The block takes one argument
+      #     (the field) and should return true or false. True selects the field, false rejects it.
+      # @return [Array] If a block was supplied to the method call, the array will contain the result of the block
+      #     for each tag found. Otherwise the array will just contain the data for each matching tag.
+      def select_fields(tag, select_block = nil, &block)
+        all_tags(tag, nil, select_block, &block)
+      end
+
+      # Find the first tag matching the criteria.
+      #
+      # If a block is supplied, it will be called with the found field data. The return value will be whatever the
+      # block returns. If no block is supplied, the field data will be returned. If nothing was found, the return
+      # value is nil.
+      #
+      # @param [String] tag Tag selection string. Tag name with indicators, '#' for wildcard, '_' for blank.
+      # @param [String] subfields Subfield specification. See FieldFormat class for more info; ignored for controlfields.
+      # @return [Object] nil if nothing found; field data or whatever block returns.
+      def first_tag(tag, subfields = '')
+        result = all_tags(tag, subfields).first
+        return nil unless result
+        return result unless block_given?
+        yield result
+      end
+
+      # Find all fields matching the criteria.
+      # (see #first_tag)
+      # @param (see #first_tag)
+      def all_fields(tag, subfields)
+        r = all_tags(tag, subfields).collect {|t| t.subfields_array(subfields)}.flatten.compact
+        return r unless block_given?
+        r.map {|field| yield field}
+        r.size > 0
+      end
+
+      # Find the first field matching the criteria
+      # (see #all_fields)
+      # @param (see #all_fields)
+      def first_field(tag, subfields)
+        result = all_fields(tag, subfields).first
+        return result unless block_given?
+        return false unless result
+        yield result
+        true
+      end
+
+      # Perform action on each field found. Code block required.
+      # @param (see #all_fields)
+      def each_field(tag, subfields)
+        all_fields(tag, subfields).each do |field|
+          yield field
+        end
+      end
+
+      # Dump content to string.
+      def marc_dump
+        all.values.flatten.each_with_object([]) {|record, m| m << record.dump}.join
+      end
+
+      # Save the current MARC record to file.
+      # @param [String] filename name of the file
+      def save(filename)
+        doc = ::Libis::Tools::XmlDocument.new
+        doc.root = @node
+
+        return doc unless filename
+
+        doc.save filename, save_with: (::Nokogiri::XML::Node::SaveOptions::NO_EMPTY_TAGS |
+            ::Nokogiri::XML::Node::SaveOptions::AS_XML |
+            ::Nokogiri::XML::Node::SaveOptions::FORMAT
+        )
+      end
+
+      # Load XML document from file and create a new {MarcRecord} for it.
+      # @param [String] filename name of XML Marc file
+      def self.load(filename)
+        doc = ::Libis::Tools::XmlDocument.open(filename)
+        self.new(doc.root)
+      end
+
+      # Load XML document from stream and create a new {MarcRecord} for it.
+      # @param [IO,String] io input stream
+      def self.read(io)
+        io = StringIO.new(io) if io.is_a? String
+        doc = ::Libis::Tools::XmlDocument.parse(io)
+        self.new(doc.root)
+      end
+
+      # Dump Marc record in Aleph Sequential format
+      # @return [String] Aleph sequential output
+      def to_aseq
+        record = ''
+        doc_number = tag('001').datas
+
+        all.select {|t| t.is_a? Libis::Metadata::FixField}.each {|t| record += "#{format('%09s', doc_number)} #{t.tag}   L #{t.datas}\n"}
+        all.select {|t| t.is_a? Libis::Metadata::VarField}.each {|t|
+          record += "#{format('%09s', doc_number)} #{t.tag}#{t.ind1}#{t.ind2} L "
+          t.keys.each {|k|
+            t.subfield_array(k).each {|f|
+              record += "$$#{k}#{CGI::unescapeHTML(f)}"
+            }
+          }
+          record += "\n"
+        }
+
+        record
+      end
+
+      protected
+
+      def element(*parts)
+        opts = options parts
+        field_format(opts, *parts)
+      end
+
+      def list_s(*parts)
+        opts = options parts, join: ' '
+        field_format(opts, *parts)
+      end
+
+      def list_c(*parts)
+        opts = options parts, join: ', '
+        field_format(opts, *parts)
+      end
+
+      def list_d(*parts)
+        opts = options parts, join: ' - '
+        field_format(opts, *parts)
+      end
+
+      def repeat(*parts)
+        opts = options parts, join: '; '
+        field_format(opts, *parts)
+      end
+
+      def opt_r(*parts)
+        opts = options parts, fix: '()'
+        field_format(opts, *parts)
+      end
+
+      def opt_s(*parts)
+        opts = options parts, fix: '[]'
+        field_format(opts, *parts)
+      end
+
+      def odis_link(group, id, label)
+        "http://www.odis.be/lnk/#{group.downcase[0, 2]}_#{id}\##{label}"
+      end
+
+      private
+
+      def options(args, default = {})
+        default.merge(args.last.is_a?(::Hash) ? args.pop : {})
+      end
+
+      def field_format(default_options, *parts)
+        Libis::Metadata::FieldFormat.new(*parts).add_default_options(default_options).to_s
+      end
+
+      def get_records(tag, ind1 = '', ind2 = '', subfield = nil, subfields = '', &block)
+
+        ind1 ||= ''
+        ind2 ||= ''
+        subfields ||= ''
+
+        ind1.tr!('_', ' ')
+        ind1.tr!('#', '')
+
+        ind2.tr!('_', ' ')
+        ind2.tr!('#', '')
+
+        found = all[tag].select do |v|
+          result = v.is_a?(Libis::Metadata::FixField) ||
+              ((ind1.empty? or v.ind1 == ind1) &&
+                  (ind2.empty? or v.ind2 == ind2) &&
+                  v.match(subfields)
+              )
+          result &&= block.call(v) if block
+          result
+        end
+
+        return found unless subfield
+
+        # duplicate tags for subfield instances
+        found.map do |field|
+          next unless field.is_a? Libis::Metadata::FixField
+          field.subfield_data[subfield].map do |sfield|
+            field.dup.subfield_data[subfield] = [sfield]
+          end
+        end.compact.flatten
+
+      end
+
+    end
+  end
+end

data/lib/libis/metadata/parser/basic_parser.rb

@@ -0,0 +1,116 @@
+require 'parslet'
+require 'parslet/convenience'
+
+module Libis
+  module Metadata
+    module Parser
+      # noinspection RubyResolve
+
+      # New style parsers and converters for metadata. New, not finished and untested.
+      class BasicParser < Parslet::Parser
+        # space
+        rule(:space) {match('\s')}
+        rule(:space?) {space.maybe}
+        rule(:spaces) {space.repeat(1)}
+        rule(:spaces?) {space.repeat}
+
+        # numbers
+        rule(:number) {match('[0-9]')}
+        rule(:number?) {number.maybe}
+        rule(:integer) {number.repeat(1)}
+
+        # chars
+        rule(:character) {match(/[a-z]/i)}
+        rule(:character?) {character.maybe}
+        rule(:characters) {character.repeat(1)}
+
+        # word
+        rule(:wordchar) {match('\w')}
+
+        # name
+        rule(:name_string) {((character | underscore) >> wordchar.repeat).repeat(1)}
+
+        # text
+        rule(:other) {not_paren}
+        rule(:text) {other.repeat(1)}
+        rule(:text?) {text.maybe}
+
+        # special chars
+        rule(:minus) {str('-')}
+        rule(:colon) {str(':')}
+        rule(:semicolon) {str(';')}
+        rule(:underscore) {str('_')}
+        rule(:hashtag) {str('#')}
+        rule(:dollar) {str('$')}
+        rule(:star) {str('*')}
+
+        # grouping
+        rule(:paren) {lparen | rparen}
+        rule(:lparen) {lrparen | lsparen | lcparen | squote | dquote}
+        rule(:rparen) {rrparen | rsparen | rcparen | squote | dquote}
+
+        rule(:not_paren) {paren.absent? >> any}
+        rule(:not_lparen) {lrparen.absent? >> lsparen.absent? >> lcparen.absent? >> squote.absent? >> dquote.absent? >> any}
+        rule(:not_rparen) {rrparen.absent? >> rsparen.absent? >> rcparen.absent? >> squote.absent? >> dquote.absent? >> any}
+
+        rule(:lrparen) {str('(')}
+        rule(:lsparen) {str('[')}
+        rule(:lcparen) {str('{')}
+        rule(:rrparen) {str(')')}
+        rule(:rsparen) {str(']')}
+        rule(:rcparen) {str('}')}
+
+        rule(:squote) {str("'")}
+        rule(:dquote) {str('"')}
+        rule(:quote) {squote | dquote}
+
+        rule(:not_squote) {squote.absent? >> any}
+        rule(:not_dquote) {dquote.absent? >> any}
+        rule(:not_quote) {quote.absent? >> any}
+
+        def complement(char)
+          case char
+          when '('
+            ')'
+          when '{'
+            '}'
+          when '['
+            ']'
+          else
+            char
+          end
+        end
+
+        def grouped(foo, left_paren = lparen)
+          scope {
+            left_paren.capture(:paren).as(:lparen) >>
+                foo >>
+                dynamic {|_, c| str(complement(c.captures[:paren]))}.as(:rparen)
+          }
+        end
+
+        def grouped_anonymous(foo, left_paren = lparen)
+          scope {
+            left_paren.capture(:paren) >>
+                foo >>
+                dynamic {|_, c| str(complement(c.captures[:paren]))}
+          }
+        end
+
+        def any_quoted(key = :text)
+          scope {
+            quote.capture(:quote) >>
+                dynamic {|_, c| (str(c.captures[:quote]).absent? >> any).repeat(1)}.maybe.as(key) >>
+                dynamic {|_, c| str(c.captures[:quote])}
+          }
+        end
+
+        def transformer
+          self.class::Transformer.new rescue nil
+        end
+
+      end
+
+    end
+  end
+end