tabular 0.2.7 → 0.3.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: dff94cc74c5cfbb23e9dfb56d2c4fb7091bab946
-  data.tar.gz: 9e0c5f281e4d5d967881692d06fac660d6b3ba00
+  metadata.gz: a72f7305b34731c1fb2e479cfbb0a13039dfd840
+  data.tar.gz: 11cca1a77c8687abbe9f026edee0422239ff2937
 SHA512:
-  metadata.gz: 32feb486e35a7be09e36188c8e052f4345bdd00dcbc74e32aca381c4447c5a3f59eba3ca287ff0f81bd3a7e4764b8ea81400039a43a01fbb22ae07ee70e33bc5
-  data.tar.gz: a1a5fc2bcc055a00478027c63f35c35498159139b9367591706bcdf319e0d8ad902771eb093df54238c54eaf8fe174946370899474d90db722d5ad8c0b5bdd14
+  metadata.gz: 298c71843ce790a2f6db8ff682c6fed2665b598234e0703384805f40beb2273193dd280b90e979287635d61ce7e2c95b305fc4842ed1b9befbfc13da5ddbbc7e
+  data.tar.gz: c522ca24515089828157fffc266450d7b45fd5993f8439ee1df67d60e7e7706b5c51b837956652edbaddb7397313262f236456cab97a4eb7e24f83069638107e

data/Gemfile

@@ -1,5 +1,4 @@
 source "https://rubygems.org"
 
 gem "rake"
-gem "ruby-ole", :git => "git://github.com/scottwillson/ruby-ole.git"
-gem "spreadsheet", :git => "git://github.com/scottwillson/spreadsheet.git"
+gem "roo", github: "Empact/roo"

data/Gemfile.lock

@@ -1,25 +1,27 @@
 GIT
-  remote: git://github.com/scottwillson/ruby-ole.git
-  revision: 6b9d2530b12259ca21af75ed2dcb26ba745a6115
+  remote: git://github.com/Empact/roo.git
+  revision: 4a1353de22e46880e656d9af062a1436eb5e1a9c
   specs:
-    ruby-ole (1.2.11.3)
-
-GIT
-  remote: git://github.com/scottwillson/spreadsheet.git
-  revision: e5d97bc43f3db66600a70ace5de638f41a4a700b
-  specs:
-    spreadsheet (0.6.6)
-      ruby-ole
+    roo (1.13.2)
+      nokogiri
+      rubyzip
+      spreadsheet (> 0.6.4)
 
 GEM
   remote: https://rubygems.org/
   specs:
+    mini_portile (0.6.0)
+    nokogiri (1.6.3.1)
+      mini_portile (= 0.6.0)
     rake (10.1.0)
+    ruby-ole (1.2.11.7)
+    rubyzip (1.1.6)
+    spreadsheet (1.0.0)
+      ruby-ole (>= 1.0)
 
 PLATFORMS
   ruby
 
 DEPENDENCIES
   rake
-  ruby-ole!
-  spreadsheet!
+  roo!

data/README

@@ -14,17 +14,15 @@ Install
 -------
 sudo gem install tabular
 
-(The gem is hosted on Gemcutter, not RubyForge)
+Or, Gemfile:
+gem "tabular"
 
 
 Dependencies
 ------------
 For tab-delimited data: Ruby standard lib
 
-For CSV: FasterCSV (http://fastercsv.rubyforge.org/)
-sudo gem install fastercsv
-
-For Excel: Spreadsheet gem (http://spreadsheet.rubyforge.org/)
+For Excel: Roo gem (https://github.com/Empact/roo)
 sudo gem install spreadsheet
 
 
@@ -64,6 +62,8 @@ There's basic test coverage. More comprehensive test coverage needs to be extrac
 
 Changes
 -------
+0.3.0 Revise Table creation methods to something sensible. Use Roo to read
+      spreadsheets. Support for xlsx.
 0.2.7 Add Table#to_space_delimited for space-padded fixed layout
 0.2.6 Add :except option for delete_blank_columns!
 0.2.5 Use modern gemspec with no runtime dependencies. Make spreadsheet gem optional.

data/lib/tabular.rb

@@ -6,8 +6,10 @@ require "tabular/blank"
 require "tabular/keys"
 require "tabular/zero"
 
+require "tabular/column_mapper"
 require "tabular/column"
 require "tabular/columns"
 require "tabular/renderer"
 require "tabular/row"
+require "tabular/tables/file_reading"
 require "tabular/table"

data/lib/tabular/column.rb

@@ -1,33 +1,21 @@
 module Tabular
   class Column
-    include Tabular::Blank
-
     attr_reader :key, :column_type
 
     # +table+ -- parent Table
     # +column+ -- parent Columns
-    # +key+ is normalized to a downcase, underscored symbol
-    def initialize(table, columns, key = nil, columns_map = {})
+    # +key+ should be a normalized, downcase, underscored symbol
+    def initialize(table, columns, key = nil)
       @columns = columns
       @table = table
+      @key = self.columns.column_mapper.map(key)
 
-      key = symbolize(key)
-      columns_map = columns_map || {}
-      map_for_key = columns_map[key]
-
-      @column_type = :string
-      case map_for_key
-      when nil
-        @key = key
-        @column_type = :date if key == :date
-      when Symbol
-        @key = map_for_key
-        @column_type = :date if key == :date
-      when Hash
-        @key = key
-        @column_type = map_for_key[:column_type]
+      if @key && @key.to_s["date"]
+        @column_type = :date
+      elsif @key && @key.to_s[/\?\z/]
+        @column_type = :boolean
       else
-        raise "Expected Symbol or Hash, but was #{map_for_key.class}"
+        @column_type = :string
       end
     end
 
@@ -77,22 +65,11 @@ module Tabular
       key.to_s
     end
 
-    private
 
-    def symbolize(key)
-      return nil if is_blank?(key)
+    protected
 
-      begin
-        key.to_s.strip.gsub(/::/, '/').
-          gsub(/([A-Z]+)([A-Z][a-z])/,'\1_\2').
-          gsub(/([a-z\d])([A-Z])/,'\1_\2').
-          tr("-", "_").
-          gsub(/ +/, "_").
-          downcase.
-          to_sym
-      rescue
-        nil
-      end
+    def columns
+      @columns ||= Columns.new
     end
   end
 end

data/lib/tabular/columns.rb

@@ -5,21 +5,35 @@ module Tabular
     include Tabular::Blank
     include Tabular::Keys
 
+    attr_accessor :column_mapper
     attr_accessor :renderer
 
     # +table+ -- Table
     # +data+ -- array of header names
-    # +columns_map+ -- see Table. Maps column names and type conversion.
-    def initialize(table, names, columns_map = {})
+
+    def column_mapper
+      @column_mapper ||= Tabular::ColumnMapper.new
+    end
+
+    def initialize(table = Table.new, names = [], column_mapper = nil)
       @table = table
-      columns_map ||= {}
-      @columns_map = normalize_columns_map(columns_map)
+      self.column_mapper = column_mapper
+
       @column_indexes = {}
       @columns_by_key = {}
+
+      set_columns table, names
+    end
+
+    def set_columns(table = Table.new, names = [])
       index = 0
-      @columns = nil
-      @columns = names.map do |column|
-        new_column = Tabular::Column.new(table, self, column, @columns_map)
+
+      if names.respond_to?(:keys)
+        names = names.keys
+      end
+
+      @columns = names.map do |name|
+        new_column = Tabular::Column.new(table, self, name)
         unless is_blank?(new_column.key)
           @column_indexes[new_column.key] = index
           @columns_by_key[new_column.key] = new_column
@@ -52,7 +66,7 @@ module Tabular
 
     # Add a new Column with +key+
     def <<(key)
-      column = Column.new(@table, self, key, @columns_map)
+      column = Column.new(@table, self, key)
       unless is_blank?(column.key) || has_key?(key)
         @column_indexes[column.key] = @columns.size
         @column_indexes[@columns.size] = column
@@ -89,20 +103,5 @@ module Tabular
     def to_space_delimited
       map(&:to_space_delimited).join "   "
     end
-
-    private
-
-    def normalize_columns_map(columns_map)
-      normalized_columns_map = {}
-      columns_map.each do |key, value|
-        case value
-        when Hash, Symbol
-          normalized_columns_map[key_to_sym(key)] = value
-        else
-          normalized_columns_map[key_to_sym(key)] = value.to_sym
-        end
-      end
-      normalized_columns_map
-    end
   end
 end

data/lib/tabular/row.rb

@@ -78,6 +78,16 @@ module Tabular
       end
     end
 
+    # Next Row
+    def next
+      @table.rows[index + 1]
+    end
+
+    # Is this the last row?
+    def last?
+      index == @table.rows.size - 1
+    end
+
     # Tabluar::Columns
     def columns
       @table.columns
@@ -133,7 +143,7 @@ module Tabular
               else
                 begin
                   if @array[index]
-                    @hash[column.key] = Date.parse(@array[index], true)
+                    @hash[column.key] = Date.parse(@array[index].to_s, true)
                   else
                     @hash[column.key] = nil
                   end
@@ -162,7 +172,7 @@ module Tabular
     def parse_invalid_date(value)
       return unless value
 
-      parts = value.split("/")
+      parts = value.to_s.split("/")
       return unless parts.size == 3
 
       month = parts[0].to_i

data/lib/tabular/table.rb

@@ -4,72 +4,22 @@ module Tabular
   class Table
     include Tabular::Blank
     include Tabular::Keys
+    include Tabular::Tables::FileReading
     include Tabular::Zero
 
-    attr_reader :options, :rows
+    attr_accessor :column_mapper
     attr_accessor :row_mapper
+    attr_reader   :rows
 
-    # +file+ : file path as String or File
-    # Assumes .txt = tab-delimited, .csv = CSV, .xls = Excel. Assumes first row is the header.
-    # Normalizes column names to lower-case with underscores.
-    def self.read(file, *options)
-      file_path = case file
-      when File
-         file.path
-      else
-        file
-      end
-
-      raise "Could not find '#{file_path}'" unless File.exists?(file_path)
-      options = extract_options(options)
-
-      format = self.format_from(options.delete(:as), file_path)
-      data = read_file(file_path, format)
-
-      Table.new data, options
-    end
-
-    # +format+ : :csv, :txt, or :xls
-    # Returns Array of Arrays
-    def self.read_file(file_path, format)
-      case format
-      when :xls
-        require "spreadsheet"
-        # Row#to_a coerces Excel data to Strings, but we want Dates and Numbers
-        data = []
-        Spreadsheet.open(file_path).worksheets.first.each do |excel_row|
-          data << excel_row.inject([]) { |row, cell| row << cell; row }
-        end
-        data
-      when :txt
-        require "csv"
-        if RUBY_VERSION < "1.9"
-          ::CSV.open(file_path, "r", "\t").collect { |row| row }
-        else
-          CSV.read(file_path, :col_sep => "\t")
-        end
-      when :csv
-        if RUBY_VERSION < "1.9"
-          require "fastercsv"
-          FasterCSV.read(file_path)
-        else
-          require "csv"
-          CSV.read(file_path)
-        end
-      else
-        raise "Cannot read '#{format}' format. Expected :xls, :xlsx, :txt, or :csv"
-      end
+    def self.read(file, options = {})
+      table = Table.new
+      table.read file, options[:as]
+      table
     end
 
     # Pass data in as +rows+. Expects rows to be an Enumerable of Enumerables.
     # Maps rows to Hash-like Tabular::Rows.
-    #
-    # Options:
-    # :columns => { :original_name => :preferred_name, :column_name => { :column_type => :boolean } }
-    #
-    # The :columns option will likely be deprecated and options for mappers and renderers added
-    def initialize(rows = [], *options)
-      @options = Table.extract_options(options)
+    def initialize(rows = [])
       self.rows = rows
     end
 
@@ -103,9 +53,11 @@ module Tabular
         cells = row
       end
 
-      if @columns.nil? && !cells.respond_to?(:keys)
-        @columns = Tabular::Columns.new(self, cells, options[:columns])
-        return columns
+      if @columns.nil?
+        @columns = Tabular::Columns.new(self, cells, column_mapper)
+        if !cells.respond_to?(:keys)
+          return columns
+        end
       end
 
       _row = Tabular::Row.new(self, cells, row)
@@ -122,7 +74,7 @@ module Tabular
 
     # Instance of Tabular::Columns
     def columns
-      @columns ||= Tabular::Columns.new(self, [])
+      @columns ||= Tabular::Columns.new(self, [], column_mapper)
     end
 
     # Remove all columns that only contain a blank string, zero, or nil
@@ -180,6 +132,13 @@ module Tabular
       columns.renderers
     end
 
+    def column_mapper=(mapper)
+      if rows.nil? || rows.size == 0
+        @columns = nil
+      end
+      @column_mapper = mapper
+    end
+
     # Last-resort storage for client code data
     def metadata
       @metadata ||= {}
@@ -196,29 +155,6 @@ module Tabular
 
     private
 
-    def self.extract_options(options)
-      if options
-        options.flatten.first || {}
-      else
-        {}
-      end
-    end
-
-    def self.format_from(as_option, file_path)
-      if as_option
-        as_option
-      else
-        case File.extname(file_path)
-        when ".xls", ".xlsx"
-          :xls
-        when ".txt"
-          :txt
-        when ".csv"
-          :csv
-        end
-      end
-    end
-
     def extract_exceptions(options)
       if options.first && options.first[:except]
         options.first[:except]

data/lib/tabular/version.rb

@@ -1,3 +1,3 @@
 module Tabular
-  VERSION = "0.2.7"
+  VERSION = "0.3.0"
 end

data/tabular.gemspec

@@ -35,7 +35,5 @@ Gem::Specification.new do |s|
   s.require_paths = ["lib"]
   s.summary = "Read, write, and manipulate CSV, tab-delimited and Excel data"
 
-  s.add_development_dependency "ruby-ole", "~> 1.2.11.3"
-  s.add_development_dependency "spreadsheet", "~> 0.6.6"
+  s.add_development_dependency "roo", "~> 1.3"
 end
-

metadata

@@ -1,43 +1,29 @@
 --- !ruby/object:Gem::Specification
 name: tabular
 version: !ruby/object:Gem::Version
-  version: 0.2.7
+  version: 0.3.0
 platform: ruby
 authors:
 - Scott Willson
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2014-09-30 00:00:00.000000000 Z
+date: 2014-10-29 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
-  name: ruby-ole
+  name: roo
   requirement: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: 1.2.11.3
+        version: '1.3'
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: 1.2.11.3
-- !ruby/object:Gem::Dependency
-  name: spreadsheet
-  requirement: !ruby/object:Gem::Requirement
-    requirements:
-    - - "~>"
-      - !ruby/object:Gem::Version
-        version: 0.6.6
-  type: :development
-  prerelease: false
-  version_requirements: !ruby/object:Gem::Requirement
-    requirements:
-    - - "~>"
-      - !ruby/object:Gem::Version
-        version: 0.6.6
+        version: '1.3'
 description: Tabular is a Ruby library for reading, writing, and manipulating CSV,
   tab-delimited and Excel data.
 email: scott.willson@gmail.c0m
@@ -82,7 +68,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
       version: '0'
 requirements: []
 rubyforge_project: 
-rubygems_version: 2.2.2
+rubygems_version: 2.4.2
 signing_key: 
 specification_version: 4
 summary: Read, write, and manipulate CSV, tab-delimited and Excel data