csv 3.2.2 → 3.2.3

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 8682072d16c079d3d25b3f22ca9f06cae36210998194ae3f3de627c74c062453
-  data.tar.gz: 755bddbed0b08dd681939a76c5f6a80f2c536a0e72edcf6e8c770be860e5fcae
+  metadata.gz: 915b3ed5a51bf4836f08f7bb06efc3b07bdc90e09209a5253092130e2cad2ab6
+  data.tar.gz: 6bce2e39329afcf200691b4b2f422b6a48d45da66368f5d5e136e0c761cd6217
 SHA512:
-  metadata.gz: 28e191df9cb41c6ca04a9969eace28deac9217de9f2677b7edec8803531c28b4c47e4866d00fd6edbc5ea0b21f10a4f59b376473475a8c96317949b84b53f49a
-  data.tar.gz: 4c747ddbdb78e4e6e8dc725199b2f35aea7949b7eb5b49929c2e818d15cbebfaa07825f57a5ee092a5da7f5c75fc2ea3c035ff204f1a95a12283cea85720e31c
+  metadata.gz: 5c1434c8e91c16de40d19d4d1200f193248e786720b67f2bbecf26a481859fe814b8cbaa02d22027668ff02588541266c8ff5d00b9fc1cfc2163b358b8e9ece9
+  data.tar.gz: 1978e933549049129f0ec99e80a10f2838b3c75a282103aa177d8421fe7589d428308e2786b29a961a4a7a5565ede77e3b1ef44ba8f4bc91b593a5a884ded7aa

data/NEWS.md

@@ -1,5 +1,54 @@
 # News
 
+## 3.2.3 - 2022-04-09
+
+### Improvements
+
+  * Added contents summary to `CSV::Table#inspect`.
+    [GitHub#229][Patch by Eriko Sugiyama]
+    [GitHub#235][Patch by Sampat Badhe]
+
+  * Suppressed `$INPUT_RECORD_SEPARATOR` deprecation warning by
+    `Warning.warn`.
+    [GitHub#233][Reported by Jean byroot Boussier]
+
+  * Improved error message for liberal parsing with quoted values.
+    [GitHub#231][Patch by Nikolay Rys]
+
+  * Fixed typos in documentation.
+    [GitHub#236][Patch by Sampat Badhe]
+
+  * Added `:max_field_size` option and deprecated `:field_size_limit` option.
+    [GitHub#238][Reported by Dan Buettner]
+
+  * Added `:symbol_raw` to built-in header converters.
+    [GitHub#237][Reported by taki]
+    [GitHub#239][Patch by Eriko Sugiyama]
+
+### Fixes
+
+  * Fixed a bug that some texts may be dropped unexpectedly.
+    [Bug #18245][ruby-core:105587][Reported by Hassan Abdul Rehman]
+
+  * Fixed a bug that `:field_size_limit` doesn't work with not complex row.
+    [GitHub#238][Reported by Dan Buettner]
+
+### Thanks
+
+  * Hassan Abdul Rehman
+
+  * Eriko Sugiyama
+
+  * Jean byroot Boussier
+
+  * Nikolay Rys
+
+  * Sampat Badhe
+
+  * Dan Buettner
+
+  * taki
+
 ## 3.2.2 - 2021-12-24
 
 ### Improvements
@@ -15,9 +64,6 @@
   * Fixed a bug that all of `ARGF` contents may not be consumed.
     [GitHub#228][Reported by Rafael Navaza]
 
-  * Fixed a bug that some texts may be dropped unexpectedly.
-    [Bug #18245][ruby-core:105587][Reported by Hassan Abdul Rehman]
-
 ### Thanks
 
   * adamroyjones
@@ -26,8 +72,6 @@
 
   * Rafael Navaza
 
-  * Hassan Abdul Rehman
-
 ## 3.2.1 - 2021-10-23
 
 ### Improvements

data/doc/csv/options/generating/write_headers.rdoc

@@ -19,7 +19,7 @@ Without +write_headers+:
 
 With +write_headers+":
   CSV.open(file_path,'w',
-      :write_headers=> true,
+      :write_headers => true,
       :headers => ['Name','Value']
     ) do |csv|
       csv << ['foo', '0']

data/doc/csv/recipes/generating.rdoc

@@ -148,7 +148,7 @@ This example defines and uses a custom write converter to strip whitespace from
 
 ==== Recipe: Specify Multiple Write Converters
 
-Use option <tt>:write_converters</tt> and multiple custom coverters
+Use option <tt>:write_converters</tt> and multiple custom converters
 to convert field values when generating \CSV.
 
 This example defines and uses two custom write converters to strip and upcase generated fields:

data/doc/csv/recipes/parsing.rdoc

@@ -83,7 +83,7 @@ Use instance method CSV#each with option +headers+ to read a source \String one
   CSV.new(string, headers: true).each do |row|
     p row
   end
-Ouput:
+Output:
   #<CSV::Row "Name":"foo" "Value":"0">
   #<CSV::Row "Name":"bar" "Value":"1">
   #<CSV::Row "Name":"baz" "Value":"2">

data/lib/csv/input_record_separator.rb

@@ -4,20 +4,7 @@ require "stringio"
 class CSV
   module InputRecordSeparator
     class << self
-      is_input_record_separator_deprecated = false
-      verbose, $VERBOSE = $VERBOSE, true
-      stderr, $stderr = $stderr, StringIO.new
-      input_record_separator = $INPUT_RECORD_SEPARATOR
-      begin
-        $INPUT_RECORD_SEPARATOR = "\r\n"
-        is_input_record_separator_deprecated = (not $stderr.string.empty?)
-      ensure
-        $INPUT_RECORD_SEPARATOR = input_record_separator
-        $stderr = stderr
-        $VERBOSE = verbose
-      end
-
-      if is_input_record_separator_deprecated
+      if RUBY_VERSION >= "3.0.0"
         def value
           "\n"
         end

data/lib/csv/parser.rb

@@ -27,6 +27,10 @@ class CSV
     class InvalidEncoding < StandardError
     end
 
+    # Raised when unexpected case is happen.
+    class UnexpectedError < StandardError
+    end
+
     #
     # CSV::Scanner receives a CSV output, scans it and return the content.
     # It also controls the life cycle of the object with its methods +keep_start+,
@@ -78,10 +82,10 @@ class CSV
     # +keep_end+, +keep_back+, +keep_drop+.
     #
     # CSV::InputsScanner.scan() tries to match with pattern at the current position.
-    # If there's a match, the scanner advances the “scan pointer” and returns the matched string.
+    # If there's a match, the scanner advances the "scan pointer" and returns the matched string.
     # Otherwise, the scanner returns nil.
     #
-    # CSV::InputsScanner.rest() returns the “rest” of the string (i.e. everything after the scan pointer).
+    # CSV::InputsScanner.rest() returns the "rest" of the string (i.e. everything after the scan pointer).
     # If there is no more data (eos? = true), it returns "".
     #
     class InputsScanner
@@ -96,11 +100,13 @@ class CSV
       end
 
       def each_line(row_separator)
+        return enum_for(__method__, row_separator) unless block_given?
         buffer = nil
         input = @scanner.rest
         position = @scanner.pos
         offset = 0
         n_row_separator_chars = row_separator.size
+        # trace(__method__, :start, line, input)
         while true
           input.each_line(row_separator) do |line|
             @scanner.pos += line.bytesize
@@ -140,25 +146,28 @@ class CSV
       end
 
       def scan(pattern)
+        # trace(__method__, pattern, :start)
         value = @scanner.scan(pattern)
+        # trace(__method__, pattern, :done, :last, value) if @last_scanner
         return value if @last_scanner
 
-        if value
-          read_chunk if @scanner.eos?
-          return value
-        else
-          nil
-        end
+        read_chunk if value and @scanner.eos?
+        # trace(__method__, pattern, :done, value)
+        value
       end
 
       def scan_all(pattern)
+        # trace(__method__, pattern, :start)
         value = @scanner.scan(pattern)
+        # trace(__method__, pattern, :done, :last, value) if @last_scanner
         return value if @last_scanner
 
         return nil if value.nil?
         while @scanner.eos? and read_chunk and (sub_value = @scanner.scan(pattern))
+          # trace(__method__, pattern, :sub, sub_value)
           value << sub_value
         end
+        # trace(__method__, pattern, :done, value)
         value
       end
 
@@ -167,68 +176,126 @@ class CSV
       end
 
       def keep_start
-        @keeps.push([@scanner.pos, nil])
+        # trace(__method__, :start)
+        adjust_last_keep
+        @keeps.push([@scanner, @scanner.pos, nil])
+        # trace(__method__, :done)
       end
 
       def keep_end
-        start, buffer = @keeps.pop
-        keep = @scanner.string.byteslice(start, @scanner.pos - start)
+        # trace(__method__, :start)
+        scanner, start, buffer = @keeps.pop
+        if scanner == @scanner
+          keep = @scanner.string.byteslice(start, @scanner.pos - start)
+        else
+          keep = @scanner.string.byteslice(0, @scanner.pos)
+        end
         if buffer
           buffer << keep
           keep = buffer
         end
+        # trace(__method__, :done, keep)
         keep
       end
 
       def keep_back
-        start, buffer = @keeps.pop
+        # trace(__method__, :start)
+        scanner, start, buffer = @keeps.pop
         if buffer
+          # trace(__method__, :rescan, start, buffer)
           string = @scanner.string
-          keep = string.byteslice(start, string.bytesize - start)
+          if scanner == @scanner
+            keep = string.byteslice(start, string.bytesize - start)
+          else
+            keep = string
+          end
           if keep and not keep.empty?
             @inputs.unshift(StringIO.new(keep))
             @last_scanner = false
           end
           @scanner = StringScanner.new(buffer)
         else
+          if @scanner != scanner
+            message = "scanners are different but no buffer: "
+            message += "#{@scanner.inspect}(#{@scanner.object_id}): "
+            message += "#{scanner.inspect}(#{scanner.object_id})"
+            raise UnexpectedError, message
+          end
+          # trace(__method__, :repos, start, buffer)
           @scanner.pos = start
         end
         read_chunk if @scanner.eos?
       end
 
       def keep_drop
-        @keeps.pop
+        _, _, buffer = @keeps.pop
+        # trace(__method__, :done, :empty) unless buffer
+        return unless buffer
+
+        last_keep = @keeps.last
+        # trace(__method__, :done, :no_last_keep) unless last_keep
+        return unless last_keep
+
+        if last_keep[2]
+          last_keep[2] << buffer
+        else
+          last_keep[2] = buffer
+        end
+        # trace(__method__, :done)
       end
 
       def rest
         @scanner.rest
       end
 
+      def check(pattern)
+        @scanner.check(pattern)
+      end
+
       private
-      def read_chunk
-        return false if @last_scanner
+      def trace(*args)
+        pp([*args, @scanner, @scanner&.string, @scanner&.pos, @keeps])
+      end
 
-        unless @keeps.empty?
-          keep = @keeps.last
-          keep_start = keep[0]
-          string = @scanner.string
-          keep_data = string.byteslice(keep_start, @scanner.pos - keep_start)
-          if keep_data
-            keep_buffer = keep[1]
-            if keep_buffer
-              keep_buffer << keep_data
-            else
-              keep[1] = keep_data.dup
-            end
+      def adjust_last_keep
+        # trace(__method__, :start)
+
+        keep = @keeps.last
+        # trace(__method__, :done, :empty) if keep.nil?
+        return if keep.nil?
+
+        scanner, start, buffer = keep
+        string = @scanner.string
+        if @scanner != scanner
+          start = 0
+        end
+        if start == 0 and @scanner.eos?
+          keep_data = string
+        else
+          keep_data = string.byteslice(start, @scanner.pos - start)
+        end
+        if keep_data
+          if buffer
+            buffer << keep_data
+          else
+            keep[2] = keep_data.dup
           end
-          keep[0] = 0
         end
 
+        # trace(__method__, :done)
+      end
+
+      def read_chunk
+        return false if @last_scanner
+
+        adjust_last_keep
+
         input = @inputs.first
         case input
         when StringIO
           string = input.read
           raise InvalidEncoding unless string.valid_encoding?
+          # trace(__method__, :stringio, string)
           @scanner = StringScanner.new(string)
           @inputs.shift
           @last_scanner = @inputs.empty?
@@ -237,6 +304,7 @@ class CSV
           chunk = input.gets(@row_separator, @chunk_size)
           if chunk
             raise InvalidEncoding unless chunk.valid_encoding?
+            # trace(__method__, :chunk, chunk)
             @scanner = StringScanner.new(chunk)
             if input.respond_to?(:eof?) and input.eof?
               @inputs.shift
@@ -244,6 +312,7 @@ class CSV
             end
             true
           else
+            # trace(__method__, :no_chunk)
             @scanner = StringScanner.new("".encode(@encoding))
             @inputs.shift
             @last_scanner = @inputs.empty?
@@ -278,7 +347,11 @@ class CSV
     end
 
     def field_size_limit
-      @field_size_limit
+      @max_field_size&.succ
+    end
+
+    def max_field_size
+      @max_field_size
     end
 
     def skip_lines
@@ -346,6 +419,16 @@ class CSV
         end
         message = "Invalid byte sequence in #{@encoding}"
         raise MalformedCSVError.new(message, lineno)
+      rescue UnexpectedError => error
+        if @scanner
+          ignore_broken_line
+          lineno = @lineno
+        else
+          lineno = @lineno + 1
+        end
+        message = "This should not be happen: #{error.message}: "
+        message += "Please report this to https://github.com/ruby/csv/issues"
+        raise MalformedCSVError.new(message, lineno)
       end
     end
 
@@ -390,7 +473,7 @@ class CSV
         @backslash_quote = false
       end
       @unconverted_fields = @options[:unconverted_fields]
-      @field_size_limit = @options[:field_size_limit]
+      @max_field_size = @options[:max_field_size]
       @skip_blanks = @options[:skip_blanks]
       @fields_converter = @options[:fields_converter]
       @header_fields_converter = @options[:header_fields_converter]
@@ -729,28 +812,28 @@ class CSV
       sample[0, 128].index(@quote_character)
     end
 
-    SCANNER_TEST = (ENV["CSV_PARSER_SCANNER_TEST"] == "yes")
-    if SCANNER_TEST
-      class UnoptimizedStringIO
-        def initialize(string)
-          @io = StringIO.new(string, "rb:#{string.encoding}")
-        end
+    class UnoptimizedStringIO # :nodoc:
+      def initialize(string)
+        @io = StringIO.new(string, "rb:#{string.encoding}")
+      end
 
-        def gets(*args)
-          @io.gets(*args)
-        end
+      def gets(*args)
+        @io.gets(*args)
+      end
 
-        def each_line(*args, &block)
-          @io.each_line(*args, &block)
-        end
+      def each_line(*args, &block)
+        @io.each_line(*args, &block)
+      end
 
-        def eof?
-          @io.eof?
-        end
+      def eof?
+        @io.eof?
       end
+    end
 
-      SCANNER_TEST_CHUNK_SIZE =
-        Integer((ENV["CSV_PARSER_SCANNER_TEST_CHUNK_SIZE"] || "1"), 10)
+    SCANNER_TEST = (ENV["CSV_PARSER_SCANNER_TEST"] == "yes")
+    if SCANNER_TEST
+      SCANNER_TEST_CHUNK_SIZE_NAME = "CSV_PARSER_SCANNER_TEST_CHUNK_SIZE"
+      SCANNER_TEST_CHUNK_SIZE_VALUE = ENV[SCANNER_TEST_CHUNK_SIZE_NAME]
       def build_scanner
         inputs = @samples.collect do |sample|
           UnoptimizedStringIO.new(sample)
@@ -760,10 +843,17 @@ class CSV
         else
           inputs << @input
         end
+        begin
+          chunk_size_value = ENV[SCANNER_TEST_CHUNK_SIZE_NAME]
+        rescue # Ractor::IsolationError
+          # Ractor on Ruby 3.0 can't read ENV value.
+          chunk_size_value = SCANNER_TEST_CHUNK_SIZE_VALUE
+        end
+        chunk_size = Integer((chunk_size_value || "1"), 10)
         InputsScanner.new(inputs,
                           @encoding,
                           @row_separator,
-                          chunk_size: SCANNER_TEST_CHUNK_SIZE)
+                          chunk_size: chunk_size)
       end
     else
       def build_scanner
@@ -826,6 +916,14 @@ class CSV
       end
     end
 
+    def validate_field_size(field)
+      return unless @max_field_size
+      return if field.size <= @max_field_size
+      ignore_broken_line
+      message = "Field size exceeded: #{field.size} > #{@max_field_size}"
+      raise MalformedCSVError.new(message, @lineno)
+    end
+
     def parse_no_quote(&block)
       @scanner.each_line(@row_separator) do |line|
         next if @skip_lines and skip_line?(line)
@@ -838,6 +936,11 @@ class CSV
         else
           line = strip_value(line)
           row = line.split(@split_column_separator, -1)
+          if @max_field_size
+            row.each do |column|
+              validate_field_size(column)
+            end
+          end
           n_columns = row.size
           i = 0
           while i < n_columns
@@ -893,6 +996,7 @@ class CSV
                 @need_robust_parsing = true
                 return parse_quotable_robust(&block)
               end
+              validate_field_size(row[i])
             end
             i += 1
           end
@@ -916,10 +1020,7 @@ class CSV
         value = parse_column_value
         if value
           @scanner.scan_all(@strip_value) if @strip_value
-          if @field_size_limit and value.size >= @field_size_limit
-            ignore_broken_line
-            raise MalformedCSVError.new("Field size exceeded", @lineno)
-          end
+          validate_field_size(value)
         end
         if parse_column_end
           row << value
@@ -940,8 +1041,14 @@ class CSV
           break
         else
           if @quoted_column_value
+            if liberal_parsing? and (new_line = @scanner.check(@line_end))
+              message =
+                "Illegal end-of-line sequence outside of a quoted field " +
+                "<#{new_line.inspect}>"
+            else
+              message = "Any value after quoted field isn't allowed"
+            end
             ignore_broken_line
-            message = "Any value after quoted field isn't allowed"
             raise MalformedCSVError.new(message, @lineno)
           elsif @unquoted_column_value and
                 (new_line = @scanner.scan(@line_end))

data/lib/csv/table.rb

@@ -999,9 +999,15 @@ class CSV
     # Omits the headers if option +write_headers+ is given as +false+
     # (see {Option +write_headers+}[../CSV.html#class-CSV-label-Option+write_headers]):
     #   table.to_csv(write_headers: false) # => "foo,0\nbar,1\nbaz,2\n"
-    def to_csv(write_headers: true, **options)
+    #
+    # Limit rows if option +limit+ is given like +2+:
+    #   table.to_csv(limit: 2) # => "Name,Value\nfoo,0\nbar,1\n"
+    def to_csv(write_headers: true, limit: nil, **options)
       array = write_headers ? [headers.to_csv(**options)] : []
-      @table.each do |row|
+      limit ||= @table.size
+      limit = @table.size + 1 + limit if limit < 0
+      limit = 0 if limit < 0
+      @table.first(limit).each do |row|
         array.push(row.fields.to_csv(**options)) unless row.header_row?
       end
 
@@ -1038,9 +1044,13 @@ class CSV
     # Example:
     #   source = "Name,Value\nfoo,0\nbar,1\nbaz,2\n"
     #   table = CSV.parse(source, headers: true)
-    #   table.inspect # => "#<CSV::Table mode:col_or_row row_count:4>"
+    #   table.inspect # => "#<CSV::Table mode:col_or_row row_count:4>\nName,Value\nfoo,0\nbar,1\nbaz,2\n"
+    #
     def inspect
-      "#<#{self.class} mode:#{@mode} row_count:#{to_a.size}>".encode("US-ASCII")
+      inspected = +"#<#{self.class} mode:#{@mode} row_count:#{to_a.size}>"
+      summary = to_csv(limit: 5)
+      inspected << "\n" << summary if summary.encoding.ascii_compatible?
+      inspected
     end
   end
 end

data/lib/csv/version.rb

@@ -2,5 +2,5 @@
 
 class CSV
   # The version of the installed library.
-  VERSION = "3.2.2"
+  VERSION = "3.2.3"
 end

data/lib/csv.rb

@@ -357,7 +357,9 @@ using CSV::MatchP if CSV.const_defined?(:MatchP)
 # - +row_sep+: Specifies the row separator; used to delimit rows.
 # - +col_sep+: Specifies the column separator; used to delimit fields.
 # - +quote_char+: Specifies the quote character; used to quote fields.
-# - +field_size_limit+: Specifies the maximum field size allowed.
+# - +field_size_limit+: Specifies the maximum field size + 1 allowed.
+#   Deprecated since 3.2.3. Use +max_field_size+ instead.
+# - +max_field_size+: Specifies the maximum field size allowed.
 # - +converters+: Specifies the field converters to be used.
 # - +unconverted_fields+: Specifies whether unconverted fields are to be available.
 # - +headers+: Specifies whether data contains headers,
@@ -926,7 +928,8 @@ class CSV
     symbol:   lambda { |h|
       h.encode(ConverterEncoding).downcase.gsub(/[^\s\w]+/, "").strip.
                                            gsub(/\s+/, "_").to_sym
-    }
+    },
+    symbol_raw: lambda { |h| h.encode(ConverterEncoding).to_sym }
   }
 
   # Default values for method options.
@@ -937,6 +940,7 @@ class CSV
     quote_char:         '"',
     # For parsing.
     field_size_limit:   nil,
+    max_field_size:     nil,
     converters:         nil,
     unconverted_fields: nil,
     headers:            false,
@@ -1865,6 +1869,7 @@ class CSV
                  row_sep: :auto,
                  quote_char: '"',
                  field_size_limit: nil,
+                 max_field_size: nil,
                  converters: nil,
                  unconverted_fields: nil,
                  headers: false,
@@ -1907,11 +1912,14 @@ class CSV
     @initial_header_converters = header_converters
     @initial_write_converters = write_converters
 
+    if max_field_size.nil? and field_size_limit
+      max_field_size = field_size_limit - 1
+    end
     @parser_options = {
       column_separator: col_sep,
       row_separator: row_sep,
       quote_character: quote_char,
-      field_size_limit: field_size_limit,
+      max_field_size: max_field_size,
       unconverted_fields: unconverted_fields,
       headers: headers,
       return_headers: return_headers,
@@ -1979,10 +1987,24 @@ class CSV
   # Returns the limit for field size; used for parsing;
   # see {Option +field_size_limit+}[#class-CSV-label-Option+field_size_limit]:
   #   CSV.new('').field_size_limit # => nil
+  #
+  # Deprecated since 3.2.3. Use +max_field_size+ instead.
   def field_size_limit
     parser.field_size_limit
   end
 
+  # :call-seq:
+  #   csv.max_field_size -> integer or nil
+  #
+  # Returns the limit for field size; used for parsing;
+  # see {Option +max_field_size+}[#class-CSV-label-Option+max_field_size]:
+  #   CSV.new('').max_field_size # => nil
+  #
+  # Since 3.2.3.
+  def max_field_size
+    parser.max_field_size
+  end
+
   # :call-seq:
   #   csv.skip_lines -> regexp or nil
   #

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: csv
 version: !ruby/object:Gem::Version
-  version: 3.2.2
+  version: 3.2.3
 platform: ruby
 authors:
 - James Edward Gray II
@@ -9,7 +9,7 @@ authors:
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2021-12-24 00:00:00.000000000 Z
+date: 2022-04-08 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: bundler
@@ -147,7 +147,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
-rubygems_version: 3.3.0
+rubygems_version: 3.4.0.dev
 signing_key: 
 specification_version: 4
 summary: CSV Reading and Writing