roo 2.6.0 → 2.7.0

data/lib/roo/excelx.rb

@@ -4,12 +4,11 @@ require 'roo/link'
 require 'roo/tempdir'
 require 'roo/utils'
 require 'forwardable'
+require 'set'
 
 module Roo
   class Excelx < Roo::Base
     extend Roo::Tempdir
-
-    require 'set'
     extend Forwardable
 
     ERROR_VALUES = %w(#N/A #REF! #NAME? #DIV/0! #NULL! #VALUE! #NUM!).to_set
@@ -46,7 +45,13 @@ module Roo
         basename = find_basename(filename_or_stream)
       end
 
+      # NOTE: Create temp directory and allow Ruby to cleanup the temp directory
+      #       when the object is garbage collected. Initially, the finalizer was
+      #       created in the Roo::Tempdir module, but that led to a segfault
+      #       when testing in Ruby 2.4.0.
       @tmpdir = self.class.make_tempdir(self, basename, options[:tmpdir_root])
+      ObjectSpace.define_finalizer(self, self.class.finalize(object_id))
+
       @shared = Shared.new(@tmpdir)
       @filename = local_filename(filename_or_stream, @tmpdir, packed)
       process_zipfile(@filename || filename_or_stream)
@@ -218,7 +223,7 @@ module Roo
       sheet = sheet_for(sheet)
       key = normalize(row, col)
       cell = sheet.cells[key]
-      !cell || cell.empty? || (cell.type == :string && cell.value.empty?) ||
+      !cell || cell.empty? ||
         (row < sheet.first_row || row > sheet.last_row || col < sheet.first_column || col > sheet.last_column)
     end
 

data/lib/roo/excelx/cell/datetime.rb

@@ -32,14 +32,9 @@ module Roo
         #
         # Returns a String representation of a cell's value.
         def formatted_value
-          date_regex = /(?<date>[dmy]+[\-\/][dmy]+([\-\/][dmy]+)?)/
-          time_regex = /(?<time>(\[?[h]\]?+:)?[m]+(:?ss|:?s)?)/
-
           formatter = @format.downcase.split(' ').map do |part|
-            if part[date_regex] == part
-              part.gsub(/#{DATE_FORMATS.keys.join('|')}/, DATE_FORMATS)
-            elsif part[time_regex]
-              part.gsub(/#{TIME_FORMATS.keys.join('|')}/, TIME_FORMATS)
+            if (parsed_format = parse_date_or_time_format(part))
+              parsed_format
             else
               warn 'Unable to parse custom format. Using "YYYY-mm-dd HH:MM:SS" format.'
               return @value.strftime('%F %T')
@@ -51,35 +46,50 @@ module Roo
 
         private
 
+        def parse_date_or_time_format(part)
+          date_regex = /(?<date>[dmy]+[\-\/][dmy]+([\-\/][dmy]+)?)/
+          time_regex = /(?<time>(\[?[h]\]?+:)?[m]+(:?ss|:?s)?)/
+
+          if part[date_regex] == part
+            formats = DATE_FORMATS
+          elsif part[time_regex]
+            formats = TIME_FORMATS
+          else
+            return false
+          end
+
+          part.gsub(/#{formats.keys.join('|')}/, formats)
+        end
+
         DATE_FORMATS = {
-          'yyyy'.freeze => '%Y'.freeze,  # Year: 2000
-          'yy'.freeze => '%y'.freeze,    # Year: 00
+          'yyyy' => '%Y',  # Year: 2000
+          'yy' => '%y',    # Year: 00
           # mmmmm => J-D
-          'mmmm'.freeze => '%B'.freeze,  # Month: January
-          'mmm'.freeze => '%^b'.freeze,   # Month: JAN
-          'mm'.freeze => '%m'.freeze,    # Month: 01
-          'm'.freeze => '%-m'.freeze,    # Month: 1
-          'dddd'.freeze => '%A'.freeze,  # Day of the Week: Sunday
-          'ddd'.freeze => '%^a'.freeze,   # Day of the Week: SUN
-          'dd'.freeze => '%d'.freeze,    # Day of the Month: 01
-          'd'.freeze => '%-d'.freeze,    # Day of the Month: 1
+          'mmmm' => '%B',  # Month: January
+          'mmm' => '%^b',   # Month: JAN
+          'mm' => '%m',    # Month: 01
+          'm' => '%-m',    # Month: 1
+          'dddd' => '%A',  # Day of the Week: Sunday
+          'ddd' => '%^a',   # Day of the Week: SUN
+          'dd' => '%d',    # Day of the Month: 01
+          'd' => '%-d'    # Day of the Month: 1
           # '\\\\'.freeze => ''.freeze,  # NOTE: Fixes a custom format's output.
         }
 
         TIME_FORMATS = {
-          'hh'.freeze => '%H'.freeze,    # Hour (24): 01
-          'h'.freeze => '%-k'.freeze,    # Hour (24): 1
+          'hh' => '%H',    # Hour (24): 01
+          'h' => '%-k'.freeze,    # Hour (24): 1
           # 'hh'.freeze => '%I'.freeze,    # Hour (12): 08
           # 'h'.freeze => '%-l'.freeze,    # Hour (12): 8
-          'mm'.freeze => '%M'.freeze,    # Minute: 01
+          'mm' => '%M',    # Minute: 01
           # FIXME: is this used? Seems like 'm' is used for month, not minute.
-          'm'.freeze => '%-M'.freeze,    # Minute: 1
-          'ss'.freeze => '%S'.freeze,    # Seconds: 01
-          's'.freeze => '%-S'.freeze,    # Seconds: 1
-          'am/pm'.freeze => '%p'.freeze, # Meridian: AM
-          '000'.freeze => '%3N'.freeze,  # Fractional Seconds: thousandth.
-          '00'.freeze => '%2N'.freeze,   # Fractional Seconds: hundredth.
-          '0'.freeze => '%1N'.freeze,    # Fractional Seconds: tenths.
+          'm' => '%-M',    # Minute: 1
+          'ss' => '%S',    # Seconds: 01
+          's' => '%-S',    # Seconds: 1
+          'am/pm' => '%p', # Meridian: AM
+          '000' => '%3N',  # Fractional Seconds: thousandth.
+          '00' => '%2N',   # Fractional Seconds: hundredth.
+          '0' => '%1N'    # Fractional Seconds: tenths.
         }
 
         def create_datetime(base_date, value)
@@ -93,7 +103,7 @@ module Roo
         def round_datetime(datetime_string)
           /(?<yyyy>\d+)-(?<mm>\d+)-(?<dd>\d+) (?<hh>\d+):(?<mi>\d+):(?<ss>\d+.\d+)/ =~ datetime_string
 
-          ::Time.new(yyyy.to_i, mm.to_i, dd.to_i, hh.to_i, mi.to_i, ss.to_r).round(0)
+          ::Time.new(yyyy, mm, dd, hh, mi, ss.to_r).round(0)
         end
       end
     end

data/lib/roo/excelx/cell/number.rb

@@ -32,7 +32,7 @@ module Roo
           if formatter.is_a? Proc
             formatter.call(@cell_value)
           elsif zero_padded_number?
-            "%0#{@format.size}d"% @cell_value
+            "%0#{@format.size}d" % @cell_value
           else
             Kernel.format(formatter, @cell_value)
           end
@@ -45,12 +45,9 @@ module Roo
             'General' => '%.0f',
             '0' => '%.0f',
             '0.00' => '%.2f',
-            '#,##0' => proc do |number|
-              Kernel.format('%.0f', number).reverse.gsub(/(\d{3})(?=\d)/, '\\1,').reverse
-            end,
-            '#,##0.00' => proc do |number|
-              Kernel.format('%.2f', number).reverse.gsub(/(\d{3})(?=\d)/, '\\1,').reverse
-            end,
+            '0.000000' => '%.6f',
+            '#,##0' => number_format('%.0f'),
+            '#,##0.00' => number_format('%.2f'),
             '0%' =>  proc do |number|
               Kernel.format('%d%', number.to_f * 100)
             end,
@@ -58,22 +55,10 @@ module Roo
               Kernel.format('%.2f%', number.to_f * 100)
             end,
             '0.00E+00' => '%.2E',
-            '#,##0 ;(#,##0)' => proc do |number|
-              formatter = number.to_i > 0 ? '%.0f' : '(%.0f)'
-              Kernel.format(formatter, number.to_f.abs).reverse.gsub(/(\d{3})(?=\d)/, '\\1,').reverse
-            end,
-            '#,##0 ;[Red](#,##0)' => proc do |number|
-              formatter = number.to_i > 0 ? '%.0f' : '[Red](%.0f)'
-              Kernel.format(formatter, number.to_f.abs).reverse.gsub(/(\d{3})(?=\d)/, '\\1,').reverse
-            end,
-            '#,##0.00;(#,##0.00)' => proc do |number|
-              formatter = number.to_i > 0 ? '%.2f' : '(%.2f)'
-              Kernel.format(formatter, number.to_f.abs).reverse.gsub(/(\d{3})(?=\d)/, '\\1,').reverse
-            end,
-            '#,##0.00;[Red](#,##0.00)' => proc do |number|
-              formatter = number.to_i > 0 ? '%.2f' : '[Red](%.2f)'
-              Kernel.format(formatter, number.to_f.abs).reverse.gsub(/(\d{3})(?=\d)/, '\\1,').reverse
-            end,
+            '#,##0 ;(#,##0)' => number_format('%.0f', '(%.0f)'),
+            '#,##0 ;[Red](#,##0)' => number_format('%.0f', '[Red](%.0f)'),
+            '#,##0.00;(#,##0.00)' => number_format('%.2f', '(%.2f)'),
+            '#,##0.00;[Red](#,##0.00)' => number_format('%.2f', '[Red](%.2f)'),
             # FIXME: not quite sure what the format should look like in this case.
             '##0.0E+0' => '%.1E',
             '@' => proc { |number| number }
@@ -82,6 +67,17 @@ module Roo
 
         private
 
+        def number_format(formatter, negative_formatter = nil)
+          proc do |number|
+            if negative_formatter
+              formatter = number.to_i > 0 ? formatter : negative_formatter
+              number = number.to_f.abs
+            end
+
+            Kernel.format(formatter, number).reverse.gsub(/(\d{3})(?=\d)/, '\\1,').reverse
+          end
+        end
+
         def zero_padded_number?
           @format[/0+/] == @format
         end

data/lib/roo/excelx/cell/time.rb

@@ -24,19 +24,19 @@ module Roo
 
         private
 
-        def create_datetime(base_date, value)
-          date = base_date + value.to_f.round(6)
-          datetime_string = date.strftime('%Y-%m-%d %H:%M:%S.%N')
-          t = round_datetime(datetime_string)
-
-          ::DateTime.civil(t.year, t.month, t.day, t.hour, t.min, t.sec)
-        end
-
-        def round_datetime(datetime_string)
-          /(?<yyyy>\d+)-(?<mm>\d+)-(?<dd>\d+) (?<hh>\d+):(?<mi>\d+):(?<ss>\d+.\d+)/ =~ datetime_string
-
-          ::Time.new(yyyy.to_i, mm.to_i, dd.to_i, hh.to_i, mi.to_i, ss.to_r).round(0)
-        end
+        # def create_datetime(base_date, value)
+        #   date = base_date + value.to_f.round(6)
+        #   datetime_string = date.strftime('%Y-%m-%d %H:%M:%S.%N')
+        #   t = round_datetime(datetime_string)
+        #
+        #   ::DateTime.civil(t.year, t.month, t.day, t.hour, t.min, t.sec)
+        # end
+
+        # def round_datetime(datetime_string)
+        #   /(?<yyyy>\d+)-(?<mm>\d+)-(?<dd>\d+) (?<hh>\d+):(?<mi>\d+):(?<ss>\d+.\d+)/ =~ datetime_string
+        #
+        #   ::Time.new(yyyy.to_i, mm.to_i, dd.to_i, hh.to_i, mi.to_i, ss.to_r).round(0)
+        # end
       end
     end
   end

data/lib/roo/formatters/base.rb

@@ -0,0 +1,15 @@
+module Roo
+  module Formatters
+    module Base
+      # converts an integer value to a time string like '02:05:06'
+      def integer_to_timestring(content)
+        h = (content / 3600.0).floor
+        content -= h * 3600
+        m = (content / 60.0).floor
+        content -= m * 60
+        s = content
+        Kernel.format("%02d:%02d:%02d", h, m, s)
+      end
+    end
+  end
+end

data/lib/roo/formatters/csv.rb

@@ -0,0 +1,84 @@
+module Roo
+  module Formatters
+    module CSV
+      def to_csv(filename = nil, separator = ",", sheet = default_sheet)
+        if filename
+          File.open(filename, "w") do |file|
+            write_csv_content(file, sheet, separator)
+          end
+          true
+        else
+          sio = ::StringIO.new
+          write_csv_content(sio, sheet, separator)
+          sio.rewind
+          sio.read
+        end
+      end
+
+      private
+
+      # Write all cells to the csv file. File can be a filename or nil. If the
+      # file argument is nil the output goes to STDOUT
+      def write_csv_content(file = nil, sheet = nil, separator = ",")
+        file ||= STDOUT
+        return unless first_row(sheet) # The sheet is empty
+
+        1.upto(last_row(sheet)) do |row|
+          1.upto(last_column(sheet)) do |col|
+            # TODO: use CSV.generate_line
+            file.print(separator) if col > 1
+            file.print cell_to_csv(row, col, sheet)
+          end
+          file.print("\n")
+        end
+      end
+
+      # The content of a cell in the csv output
+      def cell_to_csv(row, col, sheet)
+        return "" if empty?(row, col, sheet)
+
+        onecell = cell(row, col, sheet)
+
+        case celltype(row, col, sheet)
+        when :string
+          %("#{onecell.gsub('"', '""')}") unless onecell.empty?
+        when :boolean
+          # TODO: this only works for excelx
+          onecell = self.sheet_for(sheet).cells[[row, col]].formatted_value
+          %("#{onecell.gsub('"', '""').downcase}")
+        when :float, :percentage
+          if onecell == onecell.to_i
+            onecell.to_i.to_s
+          else
+            onecell.to_s
+          end
+        when :formula
+          case onecell
+          when String
+            %("#{onecell.gsub('"', '""')}") unless onecell.empty?
+          when Integer
+            onecell.to_s
+          when Float
+            if onecell == onecell.to_i
+              onecell.to_i.to_s
+            else
+              onecell.to_s
+            end
+          when Date, DateTime, TrueClass, FalseClass
+            onecell.to_s
+          else
+            fail "unhandled onecell-class #{onecell.class}"
+          end
+        when :date, :datetime
+          onecell.to_s
+        when :time
+          integer_to_timestring(onecell)
+        when :link
+          %("#{onecell.url.gsub('"', '""')}")
+        else
+          fail "unhandled celltype #{celltype(row, col, sheet)}"
+        end || ""
+      end
+    end
+  end
+end

data/lib/roo/formatters/matrix.rb

@@ -0,0 +1,23 @@
+module Roo
+  module Formatters
+    module Matrix
+      # returns a matrix object from the whole sheet or a rectangular area of a sheet
+      def to_matrix(from_row = nil, from_column = nil, to_row = nil, to_column = nil, sheet = default_sheet)
+        require 'matrix'
+
+        return ::Matrix.empty unless first_row
+
+        from_row ||= first_row(sheet)
+        to_row ||= last_row(sheet)
+        from_column ||= first_column(sheet)
+        to_column ||= last_column(sheet)
+
+        ::Matrix.rows(from_row.upto(to_row).map do |row|
+          from_column.upto(to_column).map do |col|
+            cell(row, col, sheet)
+          end
+        end)
+      end
+    end
+  end
+end

data/lib/roo/formatters/xml.rb

@@ -0,0 +1,31 @@
+# returns an XML representation of all sheets of a spreadsheet file
+module Roo
+  module Formatters
+    module XML
+      def to_xml
+        Nokogiri::XML::Builder.new do |xml|
+          xml.spreadsheet do
+            sheets.each do |sheet|
+              self.default_sheet = sheet
+              xml.sheet(name: sheet) do |x|
+                if first_row && last_row && first_column && last_column
+                  # sonst gibt es Fehler bei leeren Blaettern
+                  first_row.upto(last_row) do |row|
+                    first_column.upto(last_column) do |col|
+                      next if empty?(row, col)
+
+                      x.cell(cell(row, col),
+                      row: row,
+                      column: col,
+                      type: celltype(row, col))
+                    end
+                  end
+                end
+              end
+            end
+          end
+        end.to_xml
+      end
+    end
+  end
+end

data/lib/roo/formatters/yaml.rb

@@ -0,0 +1,40 @@
+module Roo
+  module Formatters
+    module YAML
+      # returns a rectangular area (default: all cells) as yaml-output
+      # you can add additional attributes with the prefix parameter like:
+      # oo.to_yaml({"file"=>"flightdata_2007-06-26", "sheet" => "1"})
+      def to_yaml(prefix = {}, from_row = nil, from_column = nil, to_row = nil, to_column = nil, sheet = default_sheet)
+        # return an empty string if there is no first_row, i.e. the sheet is empty
+        return "" unless first_row
+
+        from_row ||= first_row(sheet)
+        to_row ||= last_row(sheet)
+        from_column ||= first_column(sheet)
+        to_column ||= last_column(sheet)
+
+        result = "--- \n"
+        from_row.upto(to_row) do |row|
+          from_column.upto(to_column) do |col|
+            next if empty?(row, col, sheet)
+
+            result << "cell_#{row}_#{col}: \n"
+            prefix.each do|k, v|
+              result << "  #{k}: #{v} \n"
+            end
+            result << "  row: #{row} \n"
+            result << "  col: #{col} \n"
+            result << "  celltype: #{celltype(row, col, sheet)} \n"
+            value = cell(row, col, sheet)
+            if celltype(row, col, sheet) == :time
+              value = integer_to_timestring(value)
+            end
+            result << "  value: #{value} \n"
+          end
+        end
+
+        result
+      end
+    end
+  end
+end

data/lib/roo/open_office.rb

@@ -5,6 +5,7 @@ require 'zip/filesystem'
 require 'roo/font'
 require 'roo/tempdir'
 require 'base64'
+require 'openssl'
 
 module Roo
   class OpenOffice < Roo::Base
@@ -22,7 +23,12 @@ module Roo
 
       @only_visible_sheets = options[:only_visible_sheets]
       file_type_check(filename, '.ods', 'an Roo::OpenOffice', file_warning, packed)
-      @tmpdir   = self.class.make_tempdir(self, find_basename(filename), options[:tmpdir_root])
+      # NOTE: Create temp directory and allow Ruby to cleanup the temp directory
+      #       when the object is garbage collected. Initially, the finalizer was
+      #       created in the Roo::Tempdir module, but that led to a segfault
+      #       when testing in Ruby 2.4.0.
+      @tmpdir = self.class.make_tempdir(self, find_basename(filename), options[:tmpdir_root])
+      ObjectSpace.define_finalizer(self, self.class.finalize(object_id))
       @filename = local_filename(filename, @tmpdir, packed)
       # TODO: @cells_read[:default] = false
       open_oo_file(options)
@@ -340,7 +346,7 @@ module Roo
     def find_cipher(*args)
       fail ArgumentError, 'Unknown algorithm ' + algorithm unless args[0] == 'http://www.w3.org/2001/04/xmlenc#aes256-cbc'
 
-      cipher = OpenSSL::Cipher.new('AES-256-CBC')
+      cipher = ::OpenSSL::Cipher.new('AES-256-CBC')
       cipher.decrypt
       cipher.padding = 0
       cipher.key     = find_cipher_key(cipher, *args[1..4])
@@ -353,7 +359,7 @@ module Roo
     def find_cipher_key(*args)
       fail ArgumentError, 'Unknown key derivation name ', args[1] unless args[1] == 'PBKDF2'
 
-      OpenSSL::PKCS5.pbkdf2_hmac_sha1(args[2], args[3], args[4], args[0].key_len)
+      ::OpenSSL::PKCS5.pbkdf2_hmac_sha1(args[2], args[3], args[4], args[0].key_len)
     end
 
     # Block decrypt raw bytes from the zip file based on the cipher