Merge pull request #367 from roo-rb/feature/roo_formatters

Extracted testing classes for formatters

README.md

@@ -262,6 +262,16 @@ Roo's public methods have stayed relatively consistent between 1.13.x and 2.0.0,
 5. Push to the branch (`git push origin my-new-feature`)
 6. Create a new Pull Request
 
+### Testing
+Roo uses Minitest and RSpec. The best of both worlds! Run `bundle exec rake` to
+run the tests/examples.
+
+Roo also has a few tests that take a long time (5+ seconds). To run these, use
+`LONG_RUN=true bundle exec rake`
+
+When testing using Ruby 2.0 or 2.1, use this command:
+`BUNDLE_GEMFILE=Gemfile_ruby2 bundle exec rake`
+
 ### Issues
 
 If you find an issue, please create a gist and refer to it in an issue ([sample gist](https://gist.github.com/stevendaniels/98a05849036e99bb8b3c)). Here are some instructions for creating such a gist.

lib/roo/base.rb

@@ -221,32 +221,6 @@ def info
     end
   end
 
-  # # returns an XML representation of all sheets of a spreadsheet file
-  # 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
-
   # when a method like spreadsheet.a42 is called
   # convert it to a call of spreadsheet.cell('a',42)
   def method_missing(m, *args)
@@ -622,14 +596,4 @@ def process_zipfile_packed(zip, tmpdir, path = '')
       ret
     end
   end
-
-  # # 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
-  #   sprintf('%02d:%02d:%02d', h, m, s)
-  # end
 end

lib/roo/formatters/csv.rb

@@ -33,7 +33,6 @@ def write_csv_content(file = nil, sheet = nil, separator = ",")
         end
       end
 
-
       # The content of a cell in the csv output
       def cell_to_csv(row, col, sheet)
         return "" if empty?(row, col, sheet)
@@ -65,7 +64,7 @@ def cell_to_csv(row, col, sheet)
             else
               onecell.to_s
             end
-          when DateTime
+          when Date, DateTime, TrueClass, FalseClass
             onecell.to_s
           else
             fail "unhandled onecell-class #{onecell.class}"

lib/roo/formatters/yaml.rb

@@ -5,7 +5,8 @@ module YAML
       # 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 '' unless first_row # empty result if there is no first_row in a 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)

test/files/expected_results/time-test.yml

@@ -0,0 +1,26 @@
+--- 
+cell_1_1: 
+  row: 1 
+  col: 1 
+  celltype: string 
+  value: Mittags: 
+cell_1_2: 
+  row: 1 
+  col: 2 
+  celltype: time 
+  value: 12:13:14 
+cell_1_3: 
+  row: 1 
+  col: 3 
+  celltype: time 
+  value: 15:16:00 
+cell_1_4: 
+  row: 1 
+  col: 4 
+  celltype: time 
+  value: 23:00:00 
+cell_2_1: 
+  row: 2 
+  col: 1 
+  celltype: date 
+  value: 2007-11-21 

test/formatters/test_csv.rb

@@ -0,0 +1,119 @@
+require "test_helper"
+
+class TestRooFormatterCSV < Minitest::Test
+  def test_date_time_to_csv
+    with_each_spreadsheet(name: "time-test") do |workbook|
+      Dir.mktmpdir do |tempdir|
+        csv_output = File.join(tempdir, "time_test.csv")
+        assert workbook.to_csv(csv_output)
+        assert File.exist?(csv_output)
+        assert_equal "", `diff --strip-trailing-cr #{TESTDIR}/time-test.csv #{csv_output}`
+        # --strip-trailing-cr is needed because the test-file use 0A and
+        # the test on an windows box generates 0D 0A as line endings
+      end
+    end
+  end
+
+  def test_boolean_to_csv
+    with_each_spreadsheet(name: "boolean") do |workbook|
+      Dir.mktmpdir do |tempdir|
+        csv_output = File.join(tempdir,"boolean.csv")
+        assert workbook.to_csv(csv_output)
+        assert File.exist?(csv_output)
+        assert_equal "", `diff --strip-trailing-cr #{TESTDIR}/boolean.csv #{csv_output}`
+        # --strip-trailing-cr is needed because the test-file use 0A and
+        # the test on an windows box generates 0D 0A as line endings
+      end
+    end
+  end
+
+  def test_link_to_csv
+    with_each_spreadsheet(name: "link", format: :excelx) do |workbook|
+      Dir.mktmpdir do |tempdir|
+        csv_output = File.join(tempdir, "link.csv")
+        assert workbook.to_csv(csv_output)
+        assert File.exist?(csv_output)
+        assert_equal "", `diff --strip-trailing-cr #{TESTDIR}/link.csv #{csv_output}`
+        # --strip-trailing-cr is needed because the test-file use 0A and
+        # the test on an windows box generates 0D 0A as line endings
+      end
+    end
+  end
+
+  # "/tmp/xxxx" darf man unter Windows nicht verwenden, weil das nicht erkannt
+  # wird.
+  # Besser: Methode um temporaeres Dir. portabel zu bestimmen
+  def test_huge_document_to_csv
+    skip unless ENV["LONG_RUN"]
+
+    original_csv_path = File.join(TESTDIR, "Bibelbund.csv")
+    with_each_spreadsheet(name: "Bibelbund", format: [:openoffice, :excelx]) do |workbook|
+      Dir.mktmpdir do |tempdir|
+        new_csv_path = File.join(tempdir, "Bibelbund.csv")
+        assert_equal "Tagebuch des Sekret\303\244rs.    Letzte Tagung 15./16.11.75 Schweiz", workbook.cell(45, "A")
+        assert_equal "Tagebuch des Sekret\303\244rs.  Nachrichten aus Chile", workbook.cell(46, "A")
+        assert_equal "Tagebuch aus Chile  Juli 1977", workbook.cell(55, "A")
+        assert workbook.to_csv(new_csv_path)
+        assert File.exist?(new_csv_path)
+        assert FileUtils.identical?(original_csv_path, new_csv_path), "error in class #{workbook.class}"
+      end
+    end
+  end
+
+  def test_bug_empty_sheet
+    with_each_spreadsheet(name: "formula", format: [:openoffice, :excelx]) do |workbook|
+      workbook.default_sheet = "Sheet3" # is an empty sheet
+      Dir.mktmpdir do |tempdir|
+        workbook.to_csv(File.join(tempdir, "emptysheet.csv"))
+        assert_equal "", `cat #{File.join(tempdir, "emptysheet.csv")}`
+      end
+    end
+  end
+
+  def test_bug_quotes_excelx
+    skip unless ENV["LONG_RUN"]
+    # TODO: run this test with a much smaller document
+    with_each_spreadsheet(name: "Bibelbund", format: [:openoffice, :excelx]) do |workbook|
+      workbook.default_sheet = workbook.sheets.first
+      assert_equal(
+        'Einflüsse der neuen Theologie in "de gereformeerde Kerken van Nederland"',
+        workbook.cell("A", 76)
+      )
+      workbook.to_csv("csv#{$$}")
+      assert_equal(
+        'Einflüsse der neuen Theologie in "de gereformeerde Kerken van Nederland"',
+        workbook.cell("A", 78)
+      )
+      File.delete_if_exist("csv#{$$}")
+    end
+  end
+
+  def test_bug_datetime_to_csv
+    with_each_spreadsheet(name: "datetime") do |workbook|
+      Dir.mktmpdir do |tempdir|
+        datetime_csv_file = File.join(tempdir, "datetime.csv")
+
+        assert workbook.to_csv(datetime_csv_file)
+        assert File.exist?(datetime_csv_file)
+        assert_equal "", file_diff("#{TESTDIR}/so_datetime.csv", datetime_csv_file)
+      end
+    end
+  end
+
+  def test_true_class
+    assert_equal "true", cell_to_csv(1, 1)
+  end
+
+  def test_false_class
+    assert_equal "false", cell_to_csv(2, 1)
+  end
+
+  def test_date_class
+    assert_equal "2017-01-01", cell_to_csv(3, 1)
+  end
+
+  def cell_to_csv(row, col)
+    filename = File.join(TESTDIR, "formula_cell_types.xlsx")
+    Roo::Spreadsheet.open(filename).send("cell_to_csv", row, col, "Sheet1")
+  end
+end

test/formatters/test_matrix.rb

@@ -0,0 +1,76 @@
+require "test_helper"
+require "matrix"
+
+class TestRooFormatterMatrix < Minitest::Test
+  def test_matrix
+    expected_result = Matrix[
+      [1.0, 2.0, 3.0],
+      [4.0, 5.0, 6.0],
+      [7.0, 8.0, 9.0]
+    ]
+    with_each_spreadsheet(name: "matrix", format: :openoffice) do |workbook|
+      workbook.default_sheet = workbook.sheets.first
+      assert_equal expected_result, workbook.to_matrix
+    end
+  end
+
+  def test_matrix_selected_range
+    expected_result = Matrix[
+      [1.0, 2.0, 3.0],
+      [4.0, 5.0, 6.0],
+      [7.0, 8.0, 9.0]
+    ]
+    with_each_spreadsheet(name: "matrix", format: :openoffice) do |workbook|
+      workbook.default_sheet = "Sheet2"
+      assert_equal expected_result, workbook.to_matrix(3, 4, 5, 6)
+    end
+  end
+
+  def test_matrix_all_nil
+    expected_result = Matrix[
+      [nil, nil, nil],
+      [nil, nil, nil],
+      [nil, nil, nil]
+    ]
+    with_each_spreadsheet(name: "matrix", format: :openoffice) do |workbook|
+      workbook.default_sheet = "Sheet2"
+      assert_equal expected_result, workbook.to_matrix(10, 10, 12, 12)
+    end
+  end
+
+  def test_matrix_values_and_nil
+    expected_result = Matrix[
+      [1.0, nil, 3.0],
+      [4.0, 5.0, 6.0],
+      [7.0, 8.0, nil]
+    ]
+    with_each_spreadsheet(name: "matrix", format: :openoffice) do |workbook|
+      workbook.default_sheet = "Sheet3"
+      assert_equal expected_result, workbook.to_matrix(1, 1, 3, 3)
+    end
+  end
+
+  def test_matrix_specifying_sheet
+    expected_result = Matrix[
+      [1.0, nil, 3.0],
+      [4.0, 5.0, 6.0],
+      [7.0, 8.0, nil]
+    ]
+    with_each_spreadsheet(name: "matrix", format: :openoffice) do |workbook|
+      workbook.default_sheet = workbook.sheets.first
+      assert_equal expected_result, workbook.to_matrix(nil, nil, nil, nil, "Sheet3")
+    end
+  end
+
+  # #to_matrix of an empty sheet should return an empty matrix and not result in
+  # an error message
+  # 2011-06-25
+  def test_bug_to_matrix_empty_sheet
+    options = { name: "emptysheets", format: [:openoffice, :excelx] }
+    with_each_spreadsheet(options) do |workbook|
+      workbook.default_sheet = workbook.sheets.first
+      workbook.to_matrix
+      assert_equal(Matrix.empty(0, 0), workbook.to_matrix)
+    end
+  end
+end

test/formatters/test_xml.rb

@@ -0,0 +1,74 @@
+require "test_helper"
+
+class TestRooFormatterXML < Minitest::Test
+  def test_to_xml
+    expected_sheet_count = 5
+    with_each_spreadsheet(name: "numbers1", encoding: "utf8") do |workbook|
+      skip if defined? JRUBY_VERSION
+      workbook.to_xml
+      sheetname = workbook.sheets.first
+      doc = Nokogiri::XML(workbook.to_xml)
+      all_cells = init_all_cells(workbook, sheetname)
+
+      assert_equal expected_sheet_count, doc.xpath("//spreadsheet/sheet").count
+
+      doc.xpath("//spreadsheet/sheet").each do |xml_sheet|
+        assert_equal sheetname, xml_sheet.attributes["name"].value
+        xml_sheet.children.each_with_index do |cell, i|
+          next unless cell.attributes["name"]
+
+          expected = [
+            all_cells[i][:row],
+            all_cells[i][:column],
+            all_cells[i][:content],
+            all_cells[i][:type],
+          ]
+          result = [
+            cell.attributes["row"],
+            cell.attributes["column"],
+            cell.content,
+            cell.attributes["type"],
+          ]
+          assert_equal expected, result
+        end # end of sheet
+        sheetname = workbook.sheets[workbook.sheets.index(sheetname) + 1]
+      end
+    end
+  end
+
+  def test_bug_to_xml_with_empty_sheets
+    with_each_spreadsheet(name: "emptysheets", format: [:openoffice, :excelx]) do |workbook|
+      workbook.sheets.each do |sheet|
+        assert_nil workbook.first_row, "first_row not nil in sheet #{sheet}"
+        assert_nil workbook.last_row, "last_row not nil in sheet #{sheet}"
+        assert_nil workbook.first_column, "first_column not nil in sheet #{sheet}"
+        assert_nil workbook.last_column, "last_column not nil in sheet #{sheet}"
+        assert_nil workbook.first_row(sheet), "first_row not nil in sheet #{sheet}"
+        assert_nil workbook.last_row(sheet), "last_row not nil in sheet #{sheet}"
+        assert_nil workbook.first_column(sheet), "first_column not nil in sheet #{sheet}"
+        assert_nil workbook.last_column(sheet), "last_column not nil in sheet #{sheet}"
+      end
+      workbook.to_xml
+    end
+  end
+
+  # Erstellt eine Liste aller Zellen im Spreadsheet. Dies ist nötig, weil ein einfacher
+  # Textvergleich des XML-Outputs nicht funktioniert, da xml-builder die Attribute
+  # nicht immer in der gleichen Reihenfolge erzeugt.
+  def init_all_cells(workbook, sheet)
+    all = []
+    workbook.first_row(sheet).upto(workbook.last_row(sheet)) do |row|
+      workbook.first_column(sheet).upto(workbook.last_column(sheet)) do |col|
+        next if workbook.empty?(row, col, sheet)
+
+        all << {
+          row: row.to_s,
+          column: col.to_s,
+          content: workbook.cell(row, col, sheet).to_s,
+          type: workbook.celltype(row, col, sheet).to_s,
+        }
+      end
+    end
+    all
+  end
+end