csv_plus_plus 0.0.2

data/lib/csv_plus_plus/language/code_section.tab.rb

@@ -0,0 +1,443 @@
+#
+# DO NOT MODIFY!!!!
+# This file is automatically generated by Racc 1.6.2
+# from Racc grammar file "".
+#
+
+require 'racc/parser.rb'
+
+require_relative '../lexer'
+require_relative '../code_section'
+
+module CSVPlusPlus
+  module Language
+    class CodeSectionParser < Racc::Parser
+
+module_eval(<<'...end code_section.y/module_eval...', 'code_section.y', 67)
+  include ::CSVPlusPlus::Lexer
+
+  def initialize
+    super
+    @code_section = CodeSection.new
+  end
+
+  protected
+
+  def anything_to_parse?(input)
+    @rest = input.strip
+
+    return !@rest.index(::CSVPlusPlus::Lexer::END_OF_CODE_SECTION).nil?
+  end
+
+  def parse_subject
+    'code section'
+  end
+
+  def tokenizer(input)
+    ::CSVPlusPlus::Lexer::Tokenizer.new(
+      catchall: /[\(\)\{\}\/\*\+\-,=&]/,
+      ignore: /\s+|\#[^\n]+\n/,
+      input:,
+      stop_fn: lambda do |scanner|
+        return false unless scanner.scan(/#{::CSVPlusPlus::Lexer::END_OF_CODE_SECTION}/)
+
+        @tokens << [:END_OF_CODE, scanner.matched]
+        @rest = scanner.rest.strip
+        true
+      end,
+      tokens: [
+        [/\n/, :EOL],
+        [/:=/, :ASSIGN],
+        [/\bdef\b/, :FN_DEF],
+        [/\bTRUE\b/i, :TRUE],
+        [/\bFALSE\b/i, :FALSE],
+        [/"(?:[^"\\]|\\(?:["\\\/bfnrt]|u[0-9a-fA-F]{4}))*"/, :STRING],
+        [/-?[\d.]+/, :NUMBER],
+        [/\$\$/, :VAR_REF],
+        [/[!:\w_]+/, :ID],
+      ],
+    )
+  end
+
+  def return_value
+    [@code_section, @rest]
+  end
+
+  private
+
+  def e(type, *entity_args)
+    ::CSVPlusPlus::Language::TYPES[type].new(*entity_args)
+  end
+
+  def def_function(id, arguments, body)
+    fn_def = e(:function, id, arguments, body)
+    @code_section.def_function(fn_def.id, fn_def)
+  end
+
+  def def_variable(id, ast)
+    @code_section.def_variable(id, ast)
+  end
+...end code_section.y/module_eval...
+##### State transition tables begin ###
+
+racc_action_table = [
+    45,    27,     3,    37,    51,     7,    38,    52,    28,     9,
+    19,     8,    29,    30,    31,    32,    33,    24,    25,    22,
+    21,    23,    20,    19,    29,    30,    31,    32,    33,    19,
+    24,    25,    22,    21,    23,    20,    24,    25,    22,    21,
+    23,    20,    19,    29,    30,    31,    32,    33,    19,    24,
+    25,    22,    21,    23,    20,    24,    25,    22,    21,    23,
+    20,    19,    29,    30,    31,    32,    33,    19,    24,    25,
+    22,    21,    23,    20,    24,    25,    22,    21,    23,    20,
+    19,    29,    30,    31,    32,    33,    12,    24,    25,    22,
+    21,    23,    20,    19,    48,    13,    14,    15,    35,    19,
+    24,    25,    22,    21,    23,    20,    24,    25,    22,    21,
+    23,    20,    19,    29,    30,    31,    32,    33,    36,    24,
+    25,    22,    21,    23,    20,    10,    50,   nil,     7,    29,
+    30,    31,    32,    33,     8,    29,    30,    31,    32,    33,
+    29,    30,    31,    32,    33,    29,    30,    31,    32,    33,
+    29,    30,    31,    32,    33 ]
+
+racc_action_check = [
+    34,    15,     0,    26,    47,     0,    26,    47,    15,     1,
+    13,     0,    34,    34,    34,    34,    34,    13,    13,    13,
+    13,    13,    13,    19,    16,    16,    16,    16,    16,    27,
+    19,    19,    19,    19,    19,    19,    27,    27,    27,    27,
+    27,    27,    29,    39,    39,    39,    39,    39,    30,    29,
+    29,    29,    29,    29,    29,    30,    30,    30,    30,    30,
+    30,    31,    40,    40,    40,    40,    40,    32,    31,    31,
+    31,    31,    31,    31,    32,    32,    32,    32,    32,    32,
+    33,    41,    41,    41,    41,    41,     7,    33,    33,    33,
+    33,    33,    33,    36,    36,     8,     9,    12,    20,    37,
+    36,    36,    36,    36,    36,    36,    37,    37,    37,    37,
+    37,    37,    52,    42,    42,    42,    42,    42,    25,    52,
+    52,    52,    52,    52,    52,     2,    38,   nil,     2,    43,
+    43,    43,    43,    43,     2,    44,    44,    44,    44,    44,
+    46,    46,    46,    46,    46,    49,    49,    49,    49,    49,
+    53,    53,    53,    53,    53 ]
+
+racc_action_pointer = [
+     0,     9,   123,   nil,   nil,   nil,   nil,    75,    89,    96,
+   nil,   nil,    94,     7,   nil,    -3,     8,   nil,   nil,    20,
+    87,   nil,   nil,   nil,   nil,   115,    -1,    26,   nil,    39,
+    45,    58,    64,    77,    -4,   nil,    90,    96,   115,    27,
+    46,    65,    97,   113,   119,   nil,   124,     0,   nil,   129,
+   nil,   nil,   109,   134 ]
+
+racc_action_default = [
+   -30,   -30,   -30,    -2,    -4,    -5,    -6,   -30,   -30,   -30,
+    -1,    -3,   -30,   -30,    54,   -30,   -11,   -12,   -13,   -30,
+   -30,   -16,   -17,   -18,   -19,   -20,   -30,   -30,   -10,   -30,
+   -30,   -30,   -30,   -30,   -30,   -15,   -30,   -30,   -30,    -8,
+   -21,   -22,   -23,   -24,   -25,   -14,   -29,   -30,   -27,    -7,
+    -9,   -26,   -30,   -28 ]
+
+racc_goto_table = [
+    16,     4,     1,    11,     2,    26,    34,    47,   nil,   nil,
+   nil,   nil,   nil,   nil,    39,   nil,    40,    41,    42,    43,
+    44,   nil,   nil,    46,    49,   nil,   nil,   nil,   nil,   nil,
+   nil,   nil,   nil,   nil,   nil,   nil,   nil,   nil,   nil,    53 ]
+
+racc_goto_check = [
+     7,     3,     1,     3,     2,     6,     7,    10,   nil,   nil,
+   nil,   nil,   nil,   nil,     7,   nil,     7,     7,     7,     7,
+     7,   nil,   nil,     7,     7,   nil,   nil,   nil,   nil,   nil,
+   nil,   nil,   nil,   nil,   nil,   nil,   nil,   nil,   nil,     7 ]
+
+racc_goto_pointer = [
+   nil,     2,     4,     1,   nil,   nil,   -10,   -13,   nil,   nil,
+   -29 ]
+
+racc_goto_default = [
+   nil,   nil,   nil,   nil,     5,     6,   nil,   nil,    17,    18,
+   nil ]
+
+racc_reduce_table = [
+  0, 0, :racc_error,
+  2, 22, :_reduce_none,
+  1, 22, :_reduce_none,
+  2, 23, :_reduce_none,
+  1, 23, :_reduce_none,
+  1, 24, :_reduce_none,
+  1, 24, :_reduce_none,
+  6, 25, :_reduce_7,
+  5, 25, :_reduce_8,
+  3, 27, :_reduce_9,
+  1, 27, :_reduce_10,
+  3, 26, :_reduce_11,
+  1, 28, :_reduce_none,
+  1, 28, :_reduce_none,
+  3, 28, :_reduce_14,
+  2, 28, :_reduce_15,
+  1, 28, :_reduce_16,
+  1, 28, :_reduce_17,
+  1, 28, :_reduce_18,
+  1, 28, :_reduce_19,
+  1, 28, :_reduce_20,
+  3, 30, :_reduce_21,
+  3, 30, :_reduce_22,
+  3, 30, :_reduce_23,
+  3, 30, :_reduce_24,
+  3, 30, :_reduce_25,
+  4, 29, :_reduce_26,
+  3, 29, :_reduce_27,
+  3, 31, :_reduce_28,
+  1, 31, :_reduce_29 ]
+
+racc_reduce_n = 30
+
+racc_shift_n = 54
+
+racc_token_table = {
+  false => 0,
+  :error => 1,
+  :END_OF_CODE => 2,
+  "(" => 3,
+  ")" => 4,
+  :FN_DEF => 5,
+  :ASSIGN => 6,
+  "," => 7,
+  :CELL_REF => 8,
+  :EOL => 9,
+  :FALSE => 10,
+  :ID => 11,
+  :NUMBER => 12,
+  :STRING => 13,
+  :TRUE => 14,
+  :VAR_REF => 15,
+  "&" => 16,
+  "*" => 17,
+  "+" => 18,
+  "-" => 19,
+  "/" => 20 }
+
+racc_nt_base = 21
+
+racc_use_result_var = true
+
+Racc_arg = [
+  racc_action_table,
+  racc_action_check,
+  racc_action_default,
+  racc_action_pointer,
+  racc_goto_table,
+  racc_goto_check,
+  racc_goto_default,
+  racc_goto_pointer,
+  racc_nt_base,
+  racc_reduce_table,
+  racc_token_table,
+  racc_shift_n,
+  racc_reduce_n,
+  racc_use_result_var ]
+
+Racc_token_to_s_table = [
+  "$end",
+  "error",
+  "END_OF_CODE",
+  "\"(\"",
+  "\")\"",
+  "FN_DEF",
+  "ASSIGN",
+  "\",\"",
+  "CELL_REF",
+  "EOL",
+  "FALSE",
+  "ID",
+  "NUMBER",
+  "STRING",
+  "TRUE",
+  "VAR_REF",
+  "\"&\"",
+  "\"*\"",
+  "\"+\"",
+  "\"-\"",
+  "\"/\"",
+  "$start",
+  "code_section",
+  "code",
+  "def",
+  "fn_def",
+  "var_def",
+  "fn_def_args",
+  "exp",
+  "fn_call",
+  "infix_fn_call",
+  "fn_call_args" ]
+
+Racc_debug_parser = false
+
+##### State transition tables end #####
+
+# reduce 0 omitted
+
+# reduce 1 omitted
+
+# reduce 2 omitted
+
+# reduce 3 omitted
+
+# reduce 4 omitted
+
+# reduce 5 omitted
+
+# reduce 6 omitted
+
+module_eval(<<'.,.,', 'code_section.y', 29)
+  def _reduce_7(val, _values, result)
+     def_function(val[1], val[3], val[5])
+    result
+  end
+.,.,
+
+module_eval(<<'.,.,', 'code_section.y', 30)
+  def _reduce_8(val, _values, result)
+     def_function(val[1], [], val[4])
+    result
+  end
+.,.,
+
+module_eval(<<'.,.,', 'code_section.y', 32)
+  def _reduce_9(val, _values, result)
+     result = val[0] << val[2]
+    result
+  end
+.,.,
+
+module_eval(<<'.,.,', 'code_section.y', 33)
+  def _reduce_10(val, _values, result)
+     result = [val[0]]
+    result
+  end
+.,.,
+
+module_eval(<<'.,.,', 'code_section.y', 35)
+  def _reduce_11(val, _values, result)
+     def_variable(val[0], val[2])
+    result
+  end
+.,.,
+
+# reduce 12 omitted
+
+# reduce 13 omitted
+
+module_eval(<<'.,.,', 'code_section.y', 39)
+  def _reduce_14(val, _values, result)
+     result = val[1]
+    result
+  end
+.,.,
+
+module_eval(<<'.,.,', 'code_section.y', 40)
+  def _reduce_15(val, _values, result)
+     result = e(:variable, val[1])
+    result
+  end
+.,.,
+
+module_eval(<<'.,.,', 'code_section.y', 41)
+  def _reduce_16(val, _values, result)
+     result = e(:string, val[0])
+    result
+  end
+.,.,
+
+module_eval(<<'.,.,', 'code_section.y', 42)
+  def _reduce_17(val, _values, result)
+     result = e(:number, val[0])
+    result
+  end
+.,.,
+
+module_eval(<<'.,.,', 'code_section.y', 43)
+  def _reduce_18(val, _values, result)
+     result = e(:boolean, true)
+    result
+  end
+.,.,
+
+module_eval(<<'.,.,', 'code_section.y', 44)
+  def _reduce_19(val, _values, result)
+     result = e(:boolean, false)
+    result
+  end
+.,.,
+
+module_eval(<<'.,.,', 'code_section.y', 45)
+  def _reduce_20(val, _values, result)
+     result = e(:cell_reference, val[0])
+    result
+  end
+.,.,
+
+module_eval(<<'.,.,', 'code_section.y', 47)
+  def _reduce_21(val, _values, result)
+     result = e(:function_call, :concat, [val[0], val[2]])
+    result
+  end
+.,.,
+
+module_eval(<<'.,.,', 'code_section.y', 48)
+  def _reduce_22(val, _values, result)
+     result = e(:function_call, :multiply, [val[0], val[2]])
+    result
+  end
+.,.,
+
+module_eval(<<'.,.,', 'code_section.y', 49)
+  def _reduce_23(val, _values, result)
+     result = e(:function_call, :add, [val[0], val[2]])
+    result
+  end
+.,.,
+
+module_eval(<<'.,.,', 'code_section.y', 50)
+  def _reduce_24(val, _values, result)
+     result = e(:function_call, :minus, [val[0], val[2]])
+    result
+  end
+.,.,
+
+module_eval(<<'.,.,', 'code_section.y', 51)
+  def _reduce_25(val, _values, result)
+     result = e(:function_call, :divide, [val[0], val[2]])
+    result
+  end
+.,.,
+
+module_eval(<<'.,.,', 'code_section.y', 53)
+  def _reduce_26(val, _values, result)
+     result = e(:function_call, val[0], val[2])
+    result
+  end
+.,.,
+
+module_eval(<<'.,.,', 'code_section.y', 54)
+  def _reduce_27(val, _values, result)
+     result = e(:function_call, val[0], [])
+    result
+  end
+.,.,
+
+module_eval(<<'.,.,', 'code_section.y', 56)
+  def _reduce_28(val, _values, result)
+     result = val[0] << val[2]
+    result
+  end
+.,.,
+
+module_eval(<<'.,.,', 'code_section.y', 57)
+  def _reduce_29(val, _values, result)
+     result = [val[0]]
+    result
+  end
+.,.,
+
+def _reduce_none(val, _values, result)
+  val[0]
+end
+
+    end   # class CodeSectionParser
+  end   # module Language
+end   # module CSVPlusPlus

data/lib/csv_plus_plus/language/compiler.rb

@@ -0,0 +1,170 @@
+# frozen_string_literal: true
+
+require 'benchmark'
+require 'csv'
+require_relative '../cell'
+require_relative '../modifier'
+require_relative '../modifier.tab'
+require_relative '../row'
+require_relative '../template'
+require_relative 'code_section.tab'
+require_relative 'entities'
+require_relative 'runtime'
+require_relative 'scope'
+
+module CSVPlusPlus
+  module Language
+    ##
+    # Encapsulates the parsing and building of objects (+Template+ -> +Row+ -> +Cell+).
+    # Variable resolution is delegated to the +Scope+
+    # rubocop:disable Metrics/ClassLength
+    class Compiler
+      attr_reader :timings, :benchmark, :options, :runtime, :scope
+
+      # Create a compiler and make sure it gets cleaned up
+      def self.with_compiler(input:, filename:, options:, &block)
+        runtime = ::CSVPlusPlus::Language::Runtime.new(filename:, input:)
+
+        if options.verbose
+          compiler_with_timings(runtime:, options:) do |c|
+            block.call(c)
+          end
+        else
+          yield(new(runtime:, options:))
+        end
+      ensure
+        runtime.cleanup!
+      end
+
+      # Create a compiler that can time each of it's stages
+      def self.compiler_with_timings(options:, runtime:, &block)
+        ::Benchmark.benchmark(::Benchmark::CAPTION, 25, ::Benchmark::FORMAT, '> Total') do |x|
+          compiler = new(options:, runtime:, benchmark: x)
+          block.call(compiler)
+          [compiler.timings.reduce(:+)]
+        end
+      end
+
+      # initialize
+      def initialize(runtime:, options:, scope: nil, benchmark: nil)
+        @options = options
+        @runtime = runtime
+        @scope = scope || ::CSVPlusPlus::Language::Scope.new(runtime:)
+        @benchmark = benchmark
+        @timings = [] if benchmark
+      end
+
+      # Parse an entire template and return a +::CSVPlusPlus::Template+ instance
+      def parse_template
+        parse_code_section!
+        rows = parse_csv_section!
+
+        ::CSVPlusPlus::Template.new(rows:, scope: @scope).tap do |t|
+          t.validate_infinite_expands(@runtime)
+          expanding { t.expand_rows! }
+          resolve_all_cells!(t)
+        end
+      end
+
+      # parses the input file and returns a +CodeSection+
+      def parse_code_section!
+        parsing_code_section do |input|
+          code_section, csv_section = ::CSVPlusPlus::Language::CodeSectionParser.new.parse(input, self)
+          # TODO: infer a type
+          # allow user-supplied key/values to override anything global or from the code section
+          code_section.def_variables(
+            options.key_values.transform_values { |v| ::CSVPlusPlus::Language::Entities::String.new(v.to_s) }
+          )
+          @scope.code_section = code_section
+
+          # return the csv_section to the caller because they're gonna re-write input with it
+          next csv_section
+        end
+        @scope.code_section
+      end
+
+      # workflow when parsing csv
+      def parse_csv_section!
+        workflow(stage: 'Parsing CSV section') do
+          @runtime.map_rows(::CSV.new(runtime.input)) do |csv_row|
+            parse_row(csv_row)
+          end
+        end
+      ensure
+        # we're done with the file and everything is in memory
+        @runtime.cleanup!
+      end
+
+      # Using the current +@runtime+ and the given +csv_row+ parse it into a +Row+ of +Cell+s
+      # +csv_row+ should have already been run through a CSV parser and is an array of strings
+      def parse_row(csv_row)
+        row_modifier = ::CSVPlusPlus::Modifier.new(row_level: true)
+
+        cells =
+          @runtime.map_row(csv_row) do |value, _cell_index|
+            cell_modifier = ::CSVPlusPlus::Modifier.new
+            parsed_value = ::CSVPlusPlus::ModifierParser.new(row_modifier:, cell_modifier:).parse(value, @runtime)
+
+            ::CSVPlusPlus::Cell.parse(parsed_value, runtime:, modifier: cell_modifier)
+          end
+
+        ::CSVPlusPlus::Row.new(@runtime.row_index, cells, row_modifier)
+      end
+
+      # workflow when resolving the values of all cells
+      def resolve_all_cells!(template)
+        workflow(stage: 'Resolving each cell') do
+          @runtime.map_rows(template.rows, cells_too: true) do |cell|
+            cell.ast = @scope.resolve_cell_value if cell.ast
+          end
+        end
+      end
+
+      # workflow when writing results
+      def outputting!(&block)
+        workflow(stage: 'Writing the spreadsheet') do
+          block.call
+        end
+      end
+
+      # to_s
+      def to_s
+        "Compiler(options: #{@options}, runtime: #{@runtime}, scope: #{@scope})"
+      end
+
+      private
+
+      # workflow when parsing the code section
+      def parsing_code_section(&block)
+        workflow(
+          stage: 'Parsing code section',
+          processing_code_section: true
+        ) do
+          csv_section = block.call(@runtime.input.read)
+          @runtime.rewrite_input!(csv_section)
+        end
+      end
+
+      # workflow when expanding rows
+      def expanding(&block)
+        workflow(stage: 'Expanding rows') do
+          block.call
+        end
+      end
+
+      def workflow(stage:, processing_code_section: false, &block)
+        @runtime.init!(processing_code_section ? 1 : (@runtime.length_of_code_section || 1))
+
+        ret = nil
+        if @benchmark
+          @timings << @benchmark.report(stage) { ret = block.call }
+        else
+          ret = block.call
+        end
+
+        ret
+      end
+    end
+    # rubocop:enable Metrics/ClassLength
+  end
+end

data/lib/csv_plus_plus/language/entities/boolean.rb

@@ -0,0 +1,32 @@
+# frozen_string_literal: true
+
+require_relative './entity'
+
+module CSVPlusPlus
+  module Language
+    module Entities
+      ##
+      # A boolean value
+      class Boolean < Entity
+        attr_reader :value
+
+        # initialize
+        def initialize(value)
+          super(:boolean)
+          # TODO: probably can do a lot better in general on type validation
+          @value = value.is_a?(::String) ? (value.downcase == 'true') : value
+        end
+
+        # to_s
+        def to_s
+          @value.to_s.upcase
+        end
+
+        # ==
+        def ==(other)
+          super && value == other.value
+        end
+      end
+    end
+  end
+end

data/lib/csv_plus_plus/language/entities/cell_reference.rb

@@ -0,0 +1,26 @@
+# frozen_string_literal: true
+
+require_relative './entity'
+
+module CSVPlusPlus
+  module Language
+    module Entities
+      ##
+      # A reference to a cell
+      class CellReference < Entity
+        attr_reader :cell_reference
+
+        # initialize
+        def initialize(cell_reference)
+          super(:cell_reference)
+          @cell_reference = cell_reference
+        end
+
+        # to_s
+        def to_s
+          @cell_reference
+        end
+      end
+    end
+  end
+end