axlsx 1.0.16 → 1.0.18

data/lib/axlsx/util/validators.rb

@@ -1,3 +1,4 @@
+# encoding: UTF-8
 module Axlsx
   # Validate a value against a specific list of allowed values.
   class RestrictionValidator
@@ -52,6 +53,14 @@ module Axlsx
     DataTypeValidator.validate(:unsigned_int, [Fixnum, Integer], v, lambda { |arg| arg.respond_to?(:>=) && arg >= 0 })
   end
 
+  # Requires that the value is a Fixnum Integer or Float and is greater or equal to 0
+  # @param [Any] v The value validated
+  # @raise [ArgumentError] raised if the value is not a Fixnum or Integer value greater or equal to 0
+  # @return [Boolean] true if the data is valid
+  def self.validate_unsigned_numeric(v)
+    DataTypeValidator.validate("Invalid column width", [Fixnum, Integer, Float], v, lambda { |arg| arg.respond_to?(:>=) && arg >= 0 })
+  end
+
   # Requires that the value is a Fixnum or Integer
   # @param [Any] v The value validated
   def self.validate_int(v)
@@ -116,10 +125,10 @@ module Axlsx
   end
 
   # Requires that the value is a valid relationship_type
-  # XML_NS_R, TABLE_R, WORKBOOK_R, WORKSHEET_R, APP_R, RELS_R, CORE_R, STYLES_R, CHART_R, DRAWING_R are allowed
+  # XML_NS_R, TABLE_R, WORKBOOK_R, WORKSHEET_R, APP_R, RELS_R, CORE_R, STYLES_R, CHART_R, DRAWING_R, IMAGE_R, HYPERLINK_R, SHARED_STRINGS_R are allowed
   # @param [Any] v The value validated
   def self.validate_relationship_type(v)
-    RestrictionValidator.validate :relationship_type, [XML_NS_R, TABLE_R, WORKBOOK_R, WORKSHEET_R, APP_R, RELS_R, CORE_R, STYLES_R, CHART_R, DRAWING_R, IMAGE_R, HYPERLINK_R], v
+    RestrictionValidator.validate :relationship_type, [XML_NS_R, TABLE_R, WORKBOOK_R, WORKSHEET_R, APP_R, RELS_R, CORE_R, STYLES_R, CHART_R, DRAWING_R, IMAGE_R, HYPERLINK_R, SHARED_STRINGS_R], v
   end
 
   # Requires that the value is a valid table element type

data/lib/axlsx/version.rb

@@ -1,9 +1,10 @@
+# encoding: UTF-8
 module Axlsx
 
   # The version of the gem
   # When using bunle exec rake and referencing the gem on github or locally
   # it will use the gemspec, which preloads this constant for the gem's version.
   # We check to make sure that it has not already been loaded
-  VERSION="1.0.16" unless Axlsx.const_defined? :VERSION
+  VERSION="1.0.18" unless Axlsx.const_defined? :VERSION
 
 end

data/lib/axlsx/workbook/shared_strings_table.rb

@@ -0,0 +1,71 @@
+# encoding: UTF-8
+module Axlsx
+
+  # The Shared String Table class is responsible for managing and serializing common strings in a workbook.
+  # While the ECMA-376 spec allows for both inline and shared strings it seems that at least some applications like iWorks Numbers
+  # and Google Docs require that the shared string table is populated in order to interoperate properly. 
+  # As a developer, you should never need to directly work against this class. Simply set 'use_shared_strings'
+  # on the package or workbook to generate a package that uses the shared strings table instead of inline strings.
+  # @note Serialization performance is affected by using this serialization method so if you do not need interoperability
+  # it is recomended that you use the default inline string method of serialization.
+  class SharedStringsTable
+
+    # The total number of strings in the workbook including duplicates
+    # Empty cells are treated as blank strings
+    # @return [Integer]
+    attr_reader :count
+
+    # The total number of unique strings in the workbook. 
+    # @return [Integer]
+    def unique_count
+      @unique_cells.size
+    end
+
+    # An array of unique cells. Multiple attributes of the cell are used in comparison
+    # each of these unique cells is parsed into the shared string table.     
+    # @see Cell#sharable
+    attr_reader :unique_cells
+
+    # Creates a new Shared Strings Table agains an array of cells
+    # @param [Array] cells This is an array of all of the cells in the workbook
+    def initialize(cells)
+      cells = cells.flatten.reject { |c| c.type != :string || c.value.start_with?('=') }
+      @count = cells.size
+      @unique_cells = []
+      resolve(cells)
+    end
+
+
+    # Generate the xml document for the Shared Strings Table
+    # @return [String]
+    def to_xml
+      builder = Nokogiri::XML::Builder.new(:encoding => ENCODING) do |xml|
+        xml.sst(:xmlns => Axlsx::XML_NS, :count => count, :uniqueCount => unique_count) {
+          @unique_cells.each do |cell|
+            xml.si { cell.run_xml(xml) }
+          end
+        }
+      end
+      builder.to_xml(:save_with => 0)
+    end
+
+    private
+    
+    # Interate over all of the cells in the array. 
+    # if our unique cells array does not contain a sharable cell, 
+    # add the cell to our unique cells array and set the ssti attribute on the index of this cell in the shared strings table
+    # if a sharable cell already exists in our unique_cells array, set the ssti attribute of the cell and move on.
+    # @return [Array] unique cells
+    def resolve(cells)
+      cells.each do |cell|
+        index = @unique_cells.index { |item| item.shareable(cell) }
+        if index == nil
+          cell.send :ssti=, @unique_cells.size
+          @unique_cells << cell
+        else
+          cell.send :ssti=, index
+        end
+      end
+    end    
+  end
+end

data/lib/axlsx/workbook/shared_strings_table.rb~

@@ -0,0 +1,69 @@
+# encoding: UTF--8
+module Axlsx
+
+  # The Shared String Table class is responsible for managing and serializing common strings in a workbook.
+  # While the ECMA-376 spec allows for both inline and shared strings it seems that at least some applications like Numbers (Mac)
+  # and Google Docs require that the shared string table is populated in order to interoperate properly. 
+  # As a developer, you should never need to directly work against this class. Simply set 'use_shared_strings'
+  # on the package or workbook to generate a package that uses the shared strings table instead of inline strings.
+  # @note Serialization performance is affected by using this serialization method so if you do not need interoperability
+  # it is recomended that you use the default inline string method of serialization.
+  class SharedStringsTable
+
+    # The total number of strings in the workbook including duplicates
+    # Empty cells are treated as blank strings
+    # @return [Integer]
+    attr_reader :count
+
+    # The total number of unique strings in the workbook. 
+    # @return [Integer]
+    def unique_count
+      @unique_cells.size
+    end
+
+    # An array of unique cells. Multiple attributes of the cell are used in comparison
+    # each of these unique cells is parsed into the shared string table.     
+    # @see Cell#sharable
+    attr_reader :unique_cells
+
+    # Creates a new Shared Strings Table agains an array of cells
+    # @param [Array] cells This is an array of all of the cells in the workbook
+    def initialize(cells)
+      cells = cells.reject { |c| c.type != :string }
+      @count = cells.size
+      @unique_cells = []
+      resolve cells
+    end
+
+    # Interate over all of the cells in the array. 
+    # if our unique cells array does not contain a sharable cell, 
+    # add the cell to our unique cells array and set the ssti attribute on the index of this cell in the shared strings table
+    # if a sharable cell already exists in our unique_cells array, set the ssti attribute of the cell and move on.
+    # @return [Array] unique cells
+    def resolve(cells)
+      cells.each do |cell|
+        index = @unique_cells.index { |item| item.sharable(cell) }
+        if index == nil
+          cell.ssti = @unique_cells.size
+          @unique_cells << cell
+        else
+          cell.ssti = index
+        end
+      end
+    end
+
+    # Generate the xml document for the Shared Strings Table
+    # @return [String]
+    def to_xml
+      builder = Nokogiri::XML::Builder.new(:encoding => ENCODING) do |xml|
+        xml.sst(:xmlns => Axlsx::XML_NS, :count => count, :uniqueCount => unique_count) {
+          @unique_cells.each do |cell|
+            xml.si { xml.t cell.value.to_s }
+          end
+        }
+      end
+      builder.to_xml(:save_with => 0)
+    end
+  end
+
+end

data/lib/axlsx/workbook/workbook.rb

@@ -1,13 +1,16 @@
 # -*- coding: utf-8 -*-
-module Axlsx 
+module Axlsx
 
+require 'axlsx/workbook/worksheet/date_time_converter.rb'
 require 'axlsx/workbook/worksheet/cell.rb'
+require 'axlsx/workbook/worksheet/page_margins.rb'
 require 'axlsx/workbook/worksheet/row.rb'
 require 'axlsx/workbook/worksheet/worksheet.rb'
+require 'axlsx/workbook/shared_strings_table.rb'
 
   # The Workbook class is an xlsx workbook that manages worksheets, charts, drawings and styles.
   # The following parts of the Office Open XML spreadsheet specification are not implimented in this version.
-  # 
+  #
   #   bookViews
   #   calcPr
   #   customWorkbookViews
@@ -30,7 +33,20 @@ require 'axlsx/workbook/worksheet/worksheet.rb'
   #   *workbookPr is only supported to the extend of date1904
   class Workbook
 
-    # A collection of worksheets associated with this workbook. 
+    # When true, the Package will be generated with a shared string table. This may be required by some OOXML processors that do not
+    # adhere to the ECMA specification that dictates string may be inline in the sheet.
+    # Using this option will increase the time required to serialize the document as every string in every cell must be analzed and referenced.
+    # @return [Boolean]
+    attr_reader :use_shared_strings
+
+    # @see use_shared_strings
+    def use_shared_strings=(v)
+      Axlsx::validate_boolean(v)
+      @use_shared_strings = v
+    end
+
+
+    # A collection of worksheets associated with this workbook.
     # @note The recommended way to manage worksheets is add_worksheet
     # @see Workbook#add_worksheet
     # @see Worksheet
@@ -72,40 +88,31 @@ require 'axlsx/workbook/worksheet/worksheet.rb'
     # Indicates if the epoc date for serialization should be 1904. If false, 1900 is used.
     @@date1904 = false
 
-    # lets come back to this later when we are ready for parsing. 
+    # lets come back to this later when we are ready for parsing.
     #def self.parse entry
     #  io = entry.get_input_stream
     #  w = self.new
     #  w.parser_xml = Nokogiri::XML(io.read)
-    #  w.parse_string :date1904, "//xmlns:workbookPr/@date1904"      
+    #  w.parse_string :date1904, "//xmlns:workbookPr/@date1904"
     #  w
     #end
 
     # Creates a new Workbook
-    #
-    # @option options [Boolean] date1904. If this is not specified, we try to determine if the platform is bsd/darwin and set date1904 to true automatically.
-    #
+    # The recomended way to work with workbooks is via Package#workbook
+    # @option options [Boolean] date1904. If this is not specified, date1904 is set to false. Office 2011 for Mac defaults to false.
     def initialize(options={})
       @styles = Styles.new
       @worksheets = SimpleTypedList.new Worksheet
       @drawings = SimpleTypedList.new Drawing
       @charts = SimpleTypedList.new Chart
       @images = SimpleTypedList.new Pic
-      self.date1904= options[:date1904].nil? ? is_bsd? : options[:date1904]
-      yield self if block_given?      
-    end
-
-    # Uses RUBY_PLATFORM constant to determine if the OS is freebsd or darwin
-    # based on this value we attempt to set date1904.
-    # @return [Boolean]
-    def is_bsd?
-      platform = RUBY_PLATFORM.downcase
-      platform.include?('freebsd') || platform.include?('darwin')
+      self.date1904= !options[:date1904].nil? && options[:date1904]
+      yield self if block_given?
     end
 
     # Instance level access to the class variable 1904
     # @return [Boolean]
-    def date1904() @@date1904; end    
+    def date1904() @@date1904; end
 
     # see @date1904
     def date1904=(v) Axlsx::validate_boolean v; @@date1904 = v; end
@@ -121,6 +128,7 @@ require 'axlsx/workbook/worksheet/worksheet.rb'
     # Adds a worksheet to this workbook
     # @return [Worksheet]
     # @option options [String] name The name of the worksheet.
+    # @option options [Hash] page_margins The page margins for the worksheet.
     # @see Worksheet#initialize
     def add_worksheet(options={})
       worksheet = Worksheet.new(self, options)
@@ -134,19 +142,28 @@ require 'axlsx/workbook/worksheet/worksheet.rb'
       r = Relationships.new
       @worksheets.each do |sheet|
         r << Relationship.new(WORKSHEET_R, WORKSHEET_PN % (r.size+1))
-      end 
+      end
       r << Relationship.new(STYLES_R,  STYLES_PN)
+      if use_shared_strings
+          r << Relationship.new(SHARED_STRINGS_R,  SHARED_STRINGS_PN)
+      end
       r
     end
 
+    # generates a shared string object against all cells in all worksheets.
+    # @return [SharedStringTable]
+    def shared_strings
+      SharedStringsTable.new(worksheets.collect { |ws| ws.cells })
+    end
+
     # returns a range of cells in a worksheet
-    # @param [String] cell_def The excel style reference defining the worksheet and cells. The range must specify the sheet to 
+    # @param [String] cell_def The excel style reference defining the worksheet and cells. The range must specify the sheet to
     # retrieve the cells from. e.g. range('Sheet1!A1:B2') will return an array of four cells [A1, A2, B1, B2] while range('Sheet1!A1') will return a single Cell.
     # @return [Cell, Array]
     def [](cell_def)
       sheet_name = cell_def.split('!')[0] if cell_def.match('!')
       worksheet =  self.worksheets.select { |s| s.name == sheet_name }.first
-      raise ArgumentError, 'Unknown Sheet' unless sheet_name && worksheet.is_a?(Worksheet) 
+      raise ArgumentError, 'Unknown Sheet' unless sheet_name && worksheet.is_a?(Worksheet)
       worksheet[cell_def.gsub(/.+!/,"")]
     end
 
@@ -154,17 +171,22 @@ require 'axlsx/workbook/worksheet/worksheet.rb'
     # @return [String]
     def to_xml()
       add_worksheet unless worksheets.size > 0
-      builder = Nokogiri::XML::Builder.new(:encoding => ENCODING) do |xml|        
+      builder = Nokogiri::XML::Builder.new(:encoding => ENCODING) do |xml|
         xml.workbook(:xmlns => XML_NS, :'xmlns:r' => XML_NS_R) {
           xml.workbookPr(:date1904=>@@date1904)
           #<x:workbookProtection workbookPassword="xsd:hexBinary data" lockStructure="1" lockWindows="1" />
+          # Required to support rubyXL parsing as it requires sheetView, which requires this.
+          # and removed because it seems to cause some odd [Grouped] behaviour in excel.
+          # xml.bookViews {
+          #    xml.workbookView :activeTab=>0
+          # }
           xml.sheets {
-            @worksheets.each_with_index do |sheet, index|              
+            @worksheets.each_with_index do |sheet, index|
               xml.sheet(:name=>sheet.name, :sheetId=>index+1, :"r:id"=>sheet.rId)
             end
           }
         }
-      end      
+      end
       builder.to_xml(:save_with => 0)
     end
   end