fixed_width_file_validator 0.4.1 → 0.5.1

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: aed7c7903ecc7262d1a138ab2157d150a7bc962d
-  data.tar.gz: 9d43c367c4297c69a67ff94f3e6c1fc905644083
+  metadata.gz: 89bb3fdfe0ee80fb5dd745811c0c0e537ffe4ec4
+  data.tar.gz: 3a38f104702b48cad5877016771ee64c41ff1b2b
 SHA512:
-  metadata.gz: 808a4f076734290a2cb926efb3f9b1ecc1b38ced222e4fe896f62bba85620983f308a655330115908a9266185c92baac84ce5a7228c27b59931d1d9da23257d2
-  data.tar.gz: 17c030f20448d6041885e6ffd099f2c6e402af871e0353d37400cc6a5e464239490fdfbd62157289cba0ce0d793c0f29a135d8014841518b426fa5d5cd04fcfc
+  metadata.gz: 460e4afe1463c190ba3eb0d9a728617150965f0c0f9ad884d1a61b31d35d57a4249a6e2fa22d91ea3235e682873b1599a533bdd13357e25bf545e4b0f5d456e5
+  data.tar.gz: 711d13b90813f99decba451d3a60b0c83420486ebab91900a84876a446913f022d7dc60270a964d821c7a6e78dc0cf95944aa0c4d7705423410b5f30fabf8fc3

data/lib/fixed_width_file_validator/file_format.rb

@@ -1,6 +1,7 @@
 require 'fixed_width_file_validator/file_reader'
 require 'fixed_width_file_validator/validator'
 require 'fixed_width_file_validator/record_parser'
+require 'fixed_width_file_validator/record_formatter'
 
 require 'yaml'
 
@@ -20,12 +21,18 @@ module FixedWidthFileValidator
       @fields = {}
       @unique_fields = []
       @parser_field_list = []
+      @formatter_field_list = []
       @file_settings = {}
       @column = 1
       @config_file = config_file
 
-      File.open(@config_file) do |f|
-        @raw_config = symbolize(YAML.safe_load(f, [], [], true))
+      if @config_file.is_a?(String)
+        File.open(@config_file) do |f|
+          @raw_config = symbolize(YAML.safe_load(f, [], [], true))
+        end
+      else
+        # config_file must be an IO object
+        @raw_config = symbolize(YAML.safe_load(config_file, [], [], true))
       end
 
       load_config(@record_type)
@@ -53,14 +60,16 @@ module FixedWidthFileValidator
       reader.close
     end
 
-    private
+    def record_formatter
+      @record_formatter ||= RecordFormatter.new(@formatter_field_list, file_settings[:encoding])
+    end
 
     def record_parser
-      # rubocop:disable Naming/MemoizedInstanceVariableName
-      @parser ||= RecordParser.new(@parser_field_list, file_settings[:encoding])
-      # rubocop:enable Naming/MemoizedInstanceVariableName
+      @record_parser ||= RecordParser.new(@parser_field_list, file_settings[:encoding])
     end
 
+    private
+
     def load_config(record_type)
       format_config = config_for(record_type)
       format_config[:fields].each do |field_config|
@@ -68,6 +77,7 @@ module FixedWidthFileValidator
         key = field_config[:field_name].to_sym
         @fields[key] = field_config
         @parser_field_list << parser_params(key)
+        @formatter_field_list << formatter_params(key)
         @column = fields[key][:end_column] + 1
       end
       @file_settings = { skip_top_lines: format_config[:skip_top_lines] || 0,
@@ -83,12 +93,14 @@ module FixedWidthFileValidator
       end_column = start_column + width - 1
       validations = field_config[:validate]
       validations = [validations] unless validations.is_a?(Array)
+      format = field_config[:format] || "%-#{width}s" # defaults to left aligned text
 
       {
         field_name: field_name,
         start_column: start_column,
         end_column: end_column,
-        validations: validations
+        validations: validations,
+        output_format: format
       }
     end
 
@@ -105,6 +117,11 @@ module FixedWidthFileValidator
       { name: field_name, position: (f[:start_column] - 1..f[:end_column] - 1) }
     end
 
+    def formatter_params(field_name)
+      f = fields[field_name]
+      { name: field_name, position: f[:start_column], width: f[:end_column] - f[:start_column] + 1, format: f[:output_format] }
+    end
+
     def config_for(record_type)
       format_config = @raw_config[record_type]
       format_fields = format_config[:fields]

data/lib/fixed_width_file_validator/file_reader.rb

@@ -1,19 +1,19 @@
 module FixedWidthFileValidator
   class FileReader
-    def initialize(file_name, parser = nil, settings = {})
+    def initialize(data_file, parser = nil, settings = {})
       @skip_top_lines = settings[:skip_top_lines] || 0
       @skip_bottom_lines = settings[:skip_bottom_lines] || 0
-      @data_file_path = file_name
       @line_num = 0
       @buffer = []
       @parser = parser
       @skip_top_done = false
+      @data_file = data_file
     end
 
     def next_record
       line_num, content = readline_with_skip
       return unless line_num
-      @parser ? @parser.parse(content, line_num, content) : content.strip
+      @parser ? @parser.parse(content, line_num, content) : content.chop
     end
 
     def each_record
@@ -47,8 +47,17 @@ module FixedWidthFileValidator
 
     private
 
+    def open_data_file
+      @file = if @data_file.is_a?(String)
+                File.open(@data_file, 'r')
+              else
+                # assume data_file is an IO object
+                @data_file
+              end
+    end
+
     def readline_with_skip
-      @file ||= File.open(@data_file_path, 'r')
+      @file ||= open_data_file
       skip_top
       skip_bottom
       readline

data/lib/fixed_width_file_validator/record_formatter.rb

@@ -0,0 +1,44 @@
+module FixedWidthFileValidator
+  class RecordFormatter
+    attr_reader :field_list, :encoding
+
+    def initialize(field_list, encoding = nil)
+      @field_list = field_list.sort_by { |a| a[:position] }
+      @encoding = encoding || 'ISO-8859-1'
+    end
+
+    def record_width
+      last_field = @field_list.last
+      last_field[:position] + last_field[:width] - 1
+    end
+
+    def string_for(record)
+      out = ''
+      last_pos = 1
+      @field_list.each do |field|
+        current_pos = field[:position]
+        out << ' ' * (current_pos - last_pos)
+        out << formatted_value(record[field[:name]], field[:format], field[:width])
+        last_pos = current_pos + field[:width]
+      end
+      out
+    end
+
+    private
+
+    # format the string using given format
+    # if the result is shorter than width, pad space to the left
+    # if the result is longer than width, truncate the last characters
+    def formatted_value(value, format, width)
+      value_str = value ? format(format, value) : ' ' * width # all space for nil value
+      length = value_str.length
+      if length > width
+        value_str.slice(0..width - 1)
+      elsif length < width
+        ' ' * (width - length) + value_str
+      else
+        value_str
+      end
+    end
+  end
+end

data/lib/fixed_width_file_validator/record_parser.rb

@@ -7,7 +7,7 @@ module FixedWidthFileValidator
       @encoding = encoding || 'ISO-8859-1'
     end
 
-    def parse(line, line_num, raw_line)
+    def parse(line, line_num = 0, raw_line = '_NA_')
       record = { _line_num: line_num, _raw: raw_line }
       encoded = line.encode(@encoding, 'binary', invalid: :replace, undef: :replace)
       field_list.each do |field|

data/lib/fixed_width_file_validator/version.rb

@@ -1,3 +1,3 @@
 module FixedWidthFileValidator
-  VERSION = '0.4.1'.freeze
+  VERSION = '0.5.1'.freeze
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: fixed_width_file_validator
 version: !ruby/object:Gem::Version
-  version: 0.4.1
+  version: 0.5.1
 platform: ruby
 authors:
 - Li Lin
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2018-04-06 00:00:00.000000000 Z
+date: 2018-04-19 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: bundler
@@ -63,6 +63,7 @@ files:
 - lib/fixed_width_file_validator.rb
 - lib/fixed_width_file_validator/file_format.rb
 - lib/fixed_width_file_validator/file_reader.rb
+- lib/fixed_width_file_validator/record_formatter.rb
 - lib/fixed_width_file_validator/record_parser.rb
 - lib/fixed_width_file_validator/report_formatter.rb
 - lib/fixed_width_file_validator/string_helper.rb