csv_plus_plus 0.1.3 → 0.2.0

data/lib/csv_plus_plus/error.rb

@@ -2,10 +2,13 @@
 # frozen_string_literal: true
 
 require_relative './error/error'
+require_relative './error/positional_error'
+
+require_relative './error/cli_error'
+require_relative './error/compiler_error'
 require_relative './error/formula_syntax_error'
 require_relative './error/modifier_syntax_error'
 require_relative './error/modifier_validation_error'
-require_relative './error/syntax_error'
 require_relative './error/writer_error'
 
 module CSVPlusPlus

data/lib/csv_plus_plus/error_formatter.rb

@@ -0,0 +1,111 @@
+# typed: strict
+# frozen_string_literal: true
+
+module CSVPlusPlus
+  # Handle any errors potentially thrown during compilation.  This could be anything from a user error (for example
+  # calling with invalid csvpp code) to an error calling Google Sheets API or writing to the filesystem.
+  class ErrorFormatter
+    extend ::T::Sig
+
+    sig do
+      params(options: ::CSVPlusPlus::Options::Options, runtime: ::CSVPlusPlus::Runtime::Runtime).void
+    end
+    # @param options [Options]
+    # @param runtime [Runtime::Runtime]
+    def initialize(options:, runtime:)
+      @options = options
+      @runtime = runtime
+    end
+
+    sig { params(error: ::StandardError).void }
+    # Nicely handle a given error.  How it's handled depends on if it's our error and if @options.verbose
+    #
+    # @param error [CSVPlusPlus::Error, Google::Apis::ClientError, StandardError]
+    def handle_error(error)
+      # make sure that we're on a newline (verbose mode will probably be in the middle of printing a benchmark)
+      puts("\n\n") if @options.verbose
+
+      case error
+      when ::CSVPlusPlus::Error::Error
+        handle_internal_error(error)
+      when ::Google::Apis::ClientError
+        handle_google_error(error)
+      else
+        unhandled_error(error)
+      end
+    end
+
+    private
+
+    sig { params(error: ::StandardError).void }
+    # An error was thrown that we weren't planning on
+    def unhandled_error(error)
+      warn(
+        <<~ERROR_MESSAGE)
+          An unexpected error was encountered.  Please try running again with --verbose and
+          report the error at: https://github.com/patrickomatic/csv-plus-plus/issues/new'
+        ERROR_MESSAGE
+
+      return unless @options.verbose
+
+      warn(error.full_message)
+      warn("Cause: #{error.cause}") if error.cause
+    end
+
+    sig { params(error: ::CSVPlusPlus::Error::Error).void }
+    def handle_internal_error(error)
+      warn(with_position(error))
+      handle_wrapped_error(::T.must(error.wrapped_error)) if error.wrapped_error
+    end
+
+    sig { params(wrapped_error: ::StandardError).void }
+    def handle_wrapped_error(wrapped_error)
+      return unless @options.verbose
+
+      warn(wrapped_error.full_message)
+      warn((wrapped_error.backtrace || []).join("\n")) if wrapped_error.backtrace
+    end
+
+    sig { params(error: ::Google::Apis::ClientError).void }
+    def handle_google_error(error)
+      warn("Error making Google Sheets API request: #{error.message}")
+      return unless @options.verbose
+
+      warn("#{error.status_code} Error making Google API request [#{error.message}]: #{error.body}")
+    end
+
+    sig { params(error: ::CSVPlusPlus::Error::Error).returns(::String) }
+    # Output a user-helpful string that references the runtime state
+    #
+    # @param error [Error::Error] The error message to be prefixed with a filename and position
+    #
+    # @return [String]
+    def with_position(error)
+      message = error.error_message
+      case error
+      when ::CSVPlusPlus::Error::PositionalError
+        "#{message_prefix}#{cell_index} #{message}"
+      else
+        message
+      end
+    end
+
+    sig { returns(::String) }
+    def cell_index
+      if @runtime.parsing_csv_section?
+        "[#{@runtime.position.row_index},#{@runtime.position.cell_index}]"
+      else
+        ''
+      end
+    end
+
+    sig { returns(::String) }
+    def message_prefix
+      line_number = @runtime.position.line_number
+      filename = @runtime.source_code.filename
+
+      line_str = ":#{line_number}"
+      "#{filename}#{line_str}"
+    end
+  end
+end

data/lib/csv_plus_plus/google_api_client.rb

@@ -10,20 +10,30 @@ module CSVPlusPlus
     # Get a +Google::Apis::SheetsV4::SheetsService+ instance configured to connect to the sheets API
     #
     # @return [Google::Apis::SheetsV4::SheetsService]
-    def self.sheets_client
-      ::Google::Apis::SheetsV4::SheetsService.new.tap do |s|
-        s.authorization = ::Google::Auth.get_application_default(['https://www.googleapis.com/auth/spreadsheets'].freeze)
-      end
+    def sheets_client
+      ::T.must(
+        @sheets_client ||= ::T.let(
+          ::Google::Apis::SheetsV4::SheetsService.new.tap do |s|
+            s.authorization = ::Google::Auth.get_application_default(['https://www.googleapis.com/auth/spreadsheets'].freeze)
+          end,
+          ::T.nilable(::Google::Apis::SheetsV4::SheetsService)
+        )
+      )
     end
 
     sig { returns(::Google::Apis::DriveV3::DriveService) }
     # Get a +Google::Apis::DriveV3::DriveService+ instance connected to the drive API
     #
     # @return [Google::Apis::DriveV3::DriveService]
-    def self.drive_client
-      ::Google::Apis::DriveV3::DriveService.new.tap do |d|
-        d.authorization = ::Google::Auth.get_application_default(['https://www.googleapis.com/auth/drive.file'].freeze)
-      end
+    def drive_client
+      ::T.must(
+        @drive_client ||= ::T.let(
+          ::Google::Apis::DriveV3::DriveService.new.tap do |d|
+            d.authorization = ::Google::Auth.get_application_default(['https://www.googleapis.com/auth/drive.file'].freeze)
+          end,
+          ::T.nilable(::Google::Apis::DriveV3::DriveService)
+        )
+      )
     end
   end
 end

data/lib/csv_plus_plus/lexer/racc_lexer.rb

@@ -0,0 +1,144 @@
+# typed: strict
+# frozen_string_literal: true
+
+module CSVPlusPlus
+  module Lexer
+    # TODO: ugh clean this up
+    RaccToken =
+      ::T.type_alias do
+        ::T.any(
+          [::String, ::Symbol],
+          [::Symbol, ::String],
+          [::String, ::String],
+          [::Symbol, ::Symbol],
+          [::FalseClass, ::FalseClass]
+        )
+      end
+    public_constant :RaccToken
+
+    # Common methods to be mixed into the Racc parsers
+    #
+    # @attr_reader tokens [Array]
+    module RaccLexer
+      extend ::T::Sig
+      extend ::T::Helpers
+      extend ::T::Generic
+      include ::Kernel
+
+      abstract!
+
+      ReturnType = type_member
+      public_constant :ReturnType
+
+      sig { returns(::T::Array[::CSVPlusPlus::Lexer::RaccToken]) }
+      attr_reader :tokens
+
+      sig { params(tokens: ::T::Array[::CSVPlusPlus::Lexer::RaccToken]).void }
+      # Initialize a lexer instance with an empty +@tokens+
+      def initialize(tokens: [])
+        @tokens = ::T.let(tokens, ::T::Array[::CSVPlusPlus::Lexer::RaccToken])
+      end
+
+      sig { returns(::T.nilable(::CSVPlusPlus::Lexer::RaccToken)) }
+      # Used by racc to iterate each token
+      #
+      # @return [Array<(Regexp, Symbol) | (false, false)>]
+      def next_token
+        @tokens.shift
+      end
+
+      sig { params(input: ::String).returns(::CSVPlusPlus::Lexer::RaccLexer::ReturnType) }
+      # Orchestate the tokenizing, parsing and error handling of parsing input.  Each instance will implement their own
+      # +#tokenizer+ method
+      #
+      # @return [RaccLexer#] Each instance will define it's own +return_value+ with the result of parsing
+      # rubocop:disable Metrics/MethodLength
+      def parse(input)
+        return return_value unless anything_to_parse?(input)
+
+        tokenize(input)
+        do_parse
+        return_value
+      rescue ::Racc::ParseError => e
+        raise(
+          ::CSVPlusPlus::Error::FormulaSyntaxError.new(
+            "Error parsing #{parse_subject}",
+            bad_input: e.message,
+            wrapped_error: e
+          )
+        )
+      end
+      # rubocop:enable Metrics/MethodLength
+
+      protected
+
+      sig { abstract.params(input: ::String).returns(::T::Boolean) }
+      # Is the input even worth parsing? for example we don't want to parse cells unless they're a formula (start
+      # with '=')
+      #
+      # @param input [String]
+      #
+      # @return [Boolean]
+      def anything_to_parse?(input); end
+
+      sig { abstract.returns(::String) }
+      # Used for error messages, what is the thing being parsed? ("cell value", "modifier", "code section")
+      def parse_subject; end
+
+      sig { abstract.returns(::CSVPlusPlus::Lexer::RaccLexer::ReturnType) }
+      # The output of the parser
+      def return_value; end
+
+      sig { abstract.returns(::CSVPlusPlus::Lexer::Tokenizer) }
+      # Returns a +Lexer::Tokenizer+ configured for the given
+      def tokenizer; end
+
+      private
+
+      sig { params(input: ::String).void }
+      def tokenize(input)
+        return if input.nil?
+
+        t = tokenizer.scan(input)
+
+        until t.scanner.empty?
+          next if t.matches_ignore?
+
+          return if t.stop?
+
+          t.scan_tokens!
+          consume_token(t)
+        end
+
+        @tokens << %i[EOL EOL]
+      end
+
+      sig { params(tokenizer: ::CSVPlusPlus::Lexer::Tokenizer).void }
+      # rubocop:disable Metrics/AbcSize, Metrics/MethodLength
+      def consume_token(tokenizer)
+        if tokenizer.last_token&.token && tokenizer.last_match
+          @tokens << [::T.must(tokenizer.last_token).token, ::T.must(tokenizer.last_match)]
+        elsif tokenizer.scan_catchall
+          @tokens << [::T.must(tokenizer.last_match), ::T.must(tokenizer.last_match)]
+        # TODO: checking the +parse_subject+ like this is a little hacky... but we need to know if we're parsing
+        # modifiers or code_section (or formulas in a cell)
+        elsif parse_subject == 'modifier'
+          raise(
+            ::CSVPlusPlus::Error::ModifierSyntaxError.new(
+              "Unable to parse #{parse_subject} starting at",
+              bad_input: tokenizer.peek
+            )
+          )
+        else
+          raise(
+            ::CSVPlusPlus::Error::FormulaSyntaxError.new(
+              "Unable to parse #{parse_subject} starting at",
+              bad_input: tokenizer.peek
+            )
+          )
+        end
+      end
+      # rubocop:enable Metrics/AbcSize, Metrics/MethodLength
+    end
+  end
+end

data/lib/csv_plus_plus/lexer/tokenizer.rb

@@ -1,24 +1,33 @@
-# typed: true
+# typed: strict
 # frozen_string_literal: true
 
-require 'strscan'
-
 module CSVPlusPlus
   module Lexer
     # A class that contains the use-case-specific regexes for parsing
     #
-    # @attr_reader last_token [String] The last token that's been matched.
-    # @attr_reader scanner [StringScanner] The StringScanner instance that's parsing the input.
+    # @attr_reader last_token [String, nil] The last token that's been matched.
     class Tokenizer
-      attr_reader :last_token, :scanner
+      extend ::T::Sig
+
+      sig { returns(::T.nilable(::CSVPlusPlus::Lexer::Token)) }
+      attr_reader :last_token
 
+      sig do
+        params(
+          tokens: ::T::Enumerable[::CSVPlusPlus::Lexer::Token],
+          catchall: ::T.nilable(::Regexp),
+          ignore: ::T.nilable(::Regexp),
+          alter_matches: ::T::Hash[::Symbol, ::T.proc.params(s: ::String).returns(::String)],
+          stop_fn: ::T.nilable(::T.proc.params(s: ::StringScanner).returns(::T::Boolean))
+        ).void
+      end
       # @param tokens [Array<Regexp, String>] The list of tokens to scan
       # @param catchall [Regexp] A final regexp to try if nothing else matches
       # @param ignore [Regexp] Ignore anything matching this regexp
       # @param alter_matches [Object] A map of matches to alter
       # @param stop_fn [Proc] Stop parsing when this is true
       def initialize(tokens:, catchall: nil, ignore: nil, alter_matches: {}, stop_fn: nil)
-        @last_token = nil
+        @last_token = ::T.let(nil, ::T.nilable(::CSVPlusPlus::Lexer::Token))
 
         @catchall = catchall
         @ignore = ignore
@@ -27,67 +36,94 @@ module CSVPlusPlus
         @alter_matches = alter_matches
       end
 
+      sig { params(input: ::String).returns(::T.self_type) }
       # Initializers a scanner for the given input to be parsed
       #
       # @param input The input to be tokenized
+      #
       # @return [Tokenizer]
       def scan(input)
-        @scanner = ::StringScanner.new(input.strip)
+        @scanner = ::T.let(::StringScanner.new(input.strip), ::T.nilable(::StringScanner))
         self
       end
 
+      sig { returns(::StringScanner) }
+      # Returns the currently initialized +StringScanner+.  You must call +#scan+ first or else this will throw an
+      # exception.
+      #
+      # @return [StringScanner]
+      def scanner
+        # The caller needs to initialize this class with a call to #scan before we can do anything.  it sets up the
+        # +@scanner+ with it's necessary input.
+        unless @scanner
+          raise(::CSVPlusPlus::Error::CompilerError, 'Called Tokenizer#scanner without calling #scan first')
+        end
+
+        @scanner
+      end
+
+      sig { void }
       # Scan tokens and set +@last_token+ if any match
       #
       # @return [String, nil]
       def scan_tokens!
-        m = @tokens.find { |t| @scanner.scan(t.first) }
-        @last_token = m ? m[1] : nil
+        @last_token = @tokens.find { |t| scanner.scan(t.regexp) }
       end
 
+      sig { returns(::T.nilable(::String)) }
       # Scan input against the catchall pattern
       #
       # @return [String, nil]
       def scan_catchall
-        @scanner.scan(@catchall) if @catchall
+        scanner.scan(@catchall) if @catchall
       end
 
+      sig { returns(::T.nilable(::String)) }
       # Scan input against the ignore pattern
       #
       # @return [boolean]
       def matches_ignore?
-        @scanner.scan(@ignore) if @ignore
+        scanner.scan(@ignore) if @ignore
       end
 
+      sig { returns(::T.nilable(::String)) }
       # The value of the last token matched
       #
       # @return [String, nil]
       def last_match
-        return @alter_matches[@last_token].call(@scanner.matched) if @alter_matches.key?(@last_token)
+        # rubocop:disable Style/MissingElse
+        if @last_token && @alter_matches.key?(@last_token.token.to_sym)
+          # rubocop:enable Style/MissingElse
+          return ::T.must(@alter_matches[@last_token.token.to_sym]).call(scanner.matched)
+        end
 
-        @scanner.matched
+        scanner.matched
       end
 
+      sig { params(peek_characters: ::Integer).returns(::String) }
       # Read the input but don't consume it
       #
       # @param peek_characters [Integer]
       #
       # @return [String]
       def peek(peek_characters: 100)
-        @scanner.peek(peek_characters)
+        scanner.peek(peek_characters)
       end
 
+      sig { returns(::T::Boolean) }
       # Scan for our stop token (if there is one - some parsers stop early and some don't)
       #
       # @return [boolean]
       def stop?
-        @stop_fn ? @stop_fn.call(@scanner) : false
+        @stop_fn ? @stop_fn.call(scanner) : false
       end
 
+      sig { returns(::String) }
       # The rest of the un-parsed input.  The tokenizer might not need to parse the entire input
       #
       # @return [String]
       def rest
-        @scanner.rest
+        scanner.rest
       end
     end
   end

data/lib/csv_plus_plus/lexer.rb

@@ -1,7 +1,7 @@
 # typed: strict
 # frozen_string_literal: true
 
-require_relative './lexer/lexer'
+require_relative './lexer/racc_lexer'
 require_relative './lexer/tokenizer'
 
 module CSVPlusPlus
@@ -9,12 +9,51 @@ module CSVPlusPlus
   module Lexer
     extend ::T::Sig
 
+    # A token that's matched by +regexp+ and presented with +token+
+    class Token < ::T::Struct
+      const :regexp, ::Regexp
+      const :token, ::T.any(::String, ::Symbol)
+    end
+
     END_OF_CODE_SECTION = '---'
     public_constant :END_OF_CODE_SECTION
 
     VARIABLE_REF = '$$'
     public_constant :VARIABLE_REF
 
+    # @see https://github.com/ruby/racc/blob/master/lib/racc/parser.rb#L121
+    TOKEN_LIBRARY = ::T.let(
+      {
+        # A1_NOTATION: ::CSVPlusPlus::Lexer::Token.new(
+        # regexp: ::CSVPlusPlus::A1Reference::A1_NOTATION_REGEXP, token: :A1_NOTATION
+        # ),
+        FALSE: ::CSVPlusPlus::Lexer::Token.new(regexp: /false/i, token: :FALSE),
+        HEX_COLOR: ::CSVPlusPlus::Lexer::Token.new(regexp: ::CSVPlusPlus::Color::HEX_STRING_REGEXP, token: :HEX_COLOR),
+        INFIX_OP: ::CSVPlusPlus::Lexer::Token.new(regexp: %r{\^|\+|-|\*|/|&|<|>|<=|>=|<>}, token: :INFIX_OP),
+        NUMBER: ::CSVPlusPlus::Lexer::Token.new(regexp: /-?[\d.]+/, token: :NUMBER),
+        REF: ::CSVPlusPlus::Lexer::Token.new(regexp: /[$!\w:]+/, token: :REF),
+        STRING: ::CSVPlusPlus::Lexer::Token.new(
+          regexp: %r{"(?:[^"\\]|\\(?:["\\/bfnrt]|u[0-9a-fA-F]{4}))*"},
+          token: :STRING
+        ),
+        TRUE: ::CSVPlusPlus::Lexer::Token.new(regexp: /true/i, token: :TRUE),
+        VAR_REF: ::CSVPlusPlus::Lexer::Token.new(regexp: /\$\$/, token: :VAR_REF)
+      }.freeze,
+      ::T::Hash[::Symbol, ::CSVPlusPlus::Lexer::Token]
+    )
+    public_constant :TOKEN_LIBRARY
+
+    sig { params(str: ::String).returns(::String) }
+    # Run any transformations to the input before going into the CSV parser
+    #
+    # The CSV parser in particular does not like if there is whitespace after a double quote and before the next comma
+    #
+    # @param str [String]
+    # @return [String]
+    def self.preprocess(str)
+      str.gsub(/"\s*,/, '",')
+    end
+
     sig { params(str: ::String).returns(::String) }
     # When parsing a modifier with a quoted string field, we need a way to unescape.  Some examples of quoted and
     # unquoted results:

data/lib/csv_plus_plus/modifier/data_validation.rb

@@ -44,7 +44,7 @@ module CSVPlusPlus
       end
 
       def a1_notation(arg)
-        return arg if ::CSVPlusPlus::Entities::CellReference.valid_cell_reference?(arg)
+        return arg if ::CSVPlusPlus::A1Reference.valid_cell_reference?(arg)
       end
 
       def a_date(arg, allow_relative_date: false)

data/lib/csv_plus_plus/modifier/expand.rb

@@ -56,6 +56,23 @@ module CSVPlusPlus
         @starts_at = row_index
         @ends_at = row_index + @repetitions unless @repetitions.nil?
       end
+
+      sig { params(position: ::CSVPlusPlus::Runtime::Position).returns(::T::Boolean) }
+      # Does the given +position+ fall within this expand?
+      #
+      # @param position [Runtime::Position]
+      #
+      # @return [boolean]
+      def position_within?(position)
+        unless starts_at
+          raise(
+            ::CSVPlusPlus::Error::CompilerError,
+            'Must call Template.expand_rows! before checking the scope of expands.'
+          )
+        end
+
+        position.row_index >= ::T.must(starts_at) && (ends_at.nil? || position.row_index <= ::T.must(ends_at))
+      end
     end
   end
 end

data/lib/csv_plus_plus/modifier.rb

@@ -76,7 +76,12 @@ module CSVPlusPlus
       end
     end
 
-    sig { params(options: ::CSVPlusPlus::Options, row_level: ::T::Boolean).returns(::CSVPlusPlus::Modifier::Modifier) }
+    sig do
+      params(
+        options: ::CSVPlusPlus::Options::Options,
+        row_level: ::T::Boolean
+      ).returns(::CSVPlusPlus::Modifier::Modifier)
+    end
     # Return a +Modifier+ with the proper validation and helper functions attached for the given output
     #
     # @param options [boolean] is this a row level modifier? (otherwise cell-level)

data/lib/csv_plus_plus/options/file_options.rb

@@ -0,0 +1,49 @@
+# typed: strict
+# frozen_string_literal: true
+
+module CSVPlusPlus
+  module Options
+    # The options that are specific for compiling to a file
+    #
+    # @attr output_filename [Pathname] The file to write our compiled results to
+    class FileOptions < ::CSVPlusPlus::Options::Options
+      extend ::T::Sig
+      extend ::T::Helpers
+
+      sig { returns(::Pathname) }
+      attr_accessor :output_filename
+
+      sig { params(sheet_name: ::String, output_filename: ::String).void }
+      # Initialize an +Options+ object for writing to a file
+      def initialize(sheet_name, output_filename)
+        super(sheet_name)
+
+        @output_filename = ::T.let(::Pathname.new(output_filename), ::Pathname)
+      end
+
+      sig { override.returns(::CSVPlusPlus::Options::OutputFormat) }
+      # Given the options, figure out which type of +OutputFormat+ we'll be writing to
+      #
+      # @return [Options::OutputFormat]
+      def output_format
+        case output_filename.extname
+        when '.csv' then ::CSVPlusPlus::Options::OutputFormat::CSV
+        when '.ods' then ::CSVPlusPlus::Options::OutputFormat::OpenDocument
+        when /\.xl(sx|sm|tx|tm)$/ then ::CSVPlusPlus::Options::OutputFormat::Excel
+        else raise(::CSVPlusPlus::Error::CLIError, "Unsupported file extension: #{output_filename}")
+        end
+      end
+
+      sig { override.returns(::String) }
+      # Verbose summary for options specific to compiling to a file
+      #
+      # @return [String]
+      def verbose_summary
+        shared_summary(
+          <<~OUTPUT)
+            > Output filename                     | #{output_filename}
+          OUTPUT
+      end
+    end
+  end
+end

data/lib/csv_plus_plus/options/google_sheets_options.rb

@@ -0,0 +1,42 @@
+# typed: strict
+# frozen_string_literal: true
+
+module CSVPlusPlus
+  module Options
+    # The Google-specific options a user can supply.
+    #
+    # @attr sheet_id [String] The ID of the Google Sheet to write to.
+    class GoogleSheetsOptions < Options
+      extend ::T::Sig
+
+      sig { returns(::String) }
+      attr_reader :sheet_id
+
+      sig { params(sheet_name: ::String, sheet_id: ::String).void }
+      # @param sheet_name [String] The name of the sheet
+      # @param sheet_id [String] The unique ID Google uses to reference the sheet
+      def initialize(sheet_name, sheet_id)
+        super(sheet_name)
+
+        @sheet_id = sheet_id
+      end
+
+      sig { override.returns(::CSVPlusPlus::Options::OutputFormat) }
+      # @return [OutputFormat]
+      def output_format
+        ::CSVPlusPlus::Options::OutputFormat::GoogleSheets
+      end
+
+      sig { override.returns(::String) }
+      # Format a string with a verbose description of Google-specific options
+      #
+      # @return [String]
+      def verbose_summary
+        shared_summary(
+          <<~SUMMARY)
+            > Sheet ID | #{@sheet_id}
+          SUMMARY
+      end
+    end
+  end
+end