fastest-csv 0.0.1 → 0.0.4

data/.gitignore

@@ -3,6 +3,7 @@
 .bundle
 .config
 .yardoc
+.DS_Store
 Gemfile.lock
 InstalledFiles
 _yardoc
@@ -15,3 +16,5 @@ spec/reports
 test/tmp
 test/version_tmp
 tmp
+lib/*.bundle
+lib/*.jar

data/README.md

@@ -1,8 +1,8 @@
 # FastestCSV
 
-Fastest CSV class for MRI Ruby. Faster than faster_csv and fasterer-csv. 
+Fastest CSV class for MRI Ruby and JRuby. Faster than faster_csv and fasterer-csv. 
 
-Uses native C code to parse CSV lines. Not (yet) compatible with JRuby.
+Uses native C code to parse CSV lines in MRI Ruby and Java in JRuby.
 
 Supports standard CSV according to RFC4180. Not the so-called "csv" from Excel.
 

data/Rakefile

@@ -1,2 +1,22 @@
 #!/usr/bin/env rake
 require "bundler/gem_tasks"
+
+spec = Gem::Specification.load('fastest-csv.gemspec')
+
+if RUBY_PLATFORM =~ /java/
+  require 'rake/javaextensiontask'
+  Rake::JavaExtensionTask.new('csv_parser', spec)
+else
+  require 'rake/extensiontask'
+  Rake::ExtensionTask.new('csv_parser', spec)
+end
+
+require 'rake/testtask'
+Rake::TestTask.new do |t|
+  t.libs << "test"
+  t.test_files = FileList['test/tc_*.rb']
+  #test.libs << 'lib' << 'test'
+  #test.pattern = 'test/**/test_*.rb'
+  #test.verbose = true
+end
+

data/ext/csv_parser/CsvParser.java

@@ -0,0 +1,89 @@
+//
+// Copyright (c) Maarten Oelering, BrightCode BV
+//
+
+package org.brightcode;
+
+import java.util.ArrayList;
+import java.util.List;
+
+public class CsvParser {
+
+    private static int DEF_ARRAY_LEN = 32;
+
+    private static int UNQUOTED = 0;
+    private static int IN_QUOTED = 1;
+    private static int QUOTE_IN_QUOTED = 2;
+    
+    public static List parseLine(String line) {
+        int length = line.length();
+        if (length == 0)
+            return null;
+        
+        int state = UNQUOTED;
+        StringBuilder value = new StringBuilder(length);   // field value, no longer than line
+        List<String> array = new ArrayList<String>(DEF_ARRAY_LEN);
+        
+        for (int i = 0; i < length; i++) {
+            char c = line.charAt(i);
+            switch (c) {
+                case ',':
+                    if (state == UNQUOTED) {
+                        if (value.length() == 0) {
+                            array.add(null);
+                        } 
+                        else {
+                            array.add(value.toString());
+                            value.setLength(0);
+                        }
+                    }
+                    else if (state == IN_QUOTED) {
+                        value.append(c);
+                    }
+                    else if (state == 2) {
+                        array.add(value.toString());
+                        value.setLength(0);
+                        state = UNQUOTED;
+                    }
+                    break;
+                case '"':
+                    if (state == UNQUOTED) {
+                        state = IN_QUOTED;
+                    }
+                    else if (state == IN_QUOTED) {
+                        state = QUOTE_IN_QUOTED;
+                    }
+                    else if (state == QUOTE_IN_QUOTED) {
+                        value.append(c);   // escaped quote
+                        state = IN_QUOTED;
+                    }
+                    break;
+                case '\r':
+                case '\n':
+                    if (state == IN_QUOTED) {
+                        value.append(c);
+                    }
+                    else {
+                        i = length;  // only parse first line if multiline
+                    }
+                    break;
+                default:
+                    value.append(c);
+                    break;
+            }
+        }
+        if (state == UNQUOTED) {
+            if (value.length() == 0) {
+                array.add(null);
+            } 
+            else {
+                array.add(value.toString());
+                value.setLength(0);
+            }
+        }
+        else if (state == QUOTE_IN_QUOTED) {
+            array.add(value.toString());
+        }
+        return array;
+    }
+}

data/ext/csv_parser/CsvParserService.java

@@ -0,0 +1,115 @@
+//
+// Copyright (c) Maarten Oelering, BrightCode BV
+//
+
+package org.brightcode;
+
+import java.io.IOException;
+
+import org.jruby.Ruby;
+import org.jruby.RubyArray;
+import org.jruby.RubyModule;
+import org.jruby.RubyString;
+import org.jruby.runtime.Block;
+import org.jruby.runtime.CallbackFactory;
+import org.jruby.runtime.builtin.IRubyObject;
+import org.jruby.runtime.load.BasicLibraryService;
+
+public class CsvParserService implements BasicLibraryService {
+
+    private Ruby runtime;
+
+    private static int DEF_ARRAY_LEN = 32;
+
+    private static int UNQUOTED = 0;
+    private static int IN_QUOTED = 1;
+    private static int QUOTE_IN_QUOTED = 2;
+    
+    // Initial setup function. Takes a reference to the current JRuby runtime and
+    // sets up our modules.
+    public boolean basicLoad(Ruby runtime) throws IOException {
+        this.runtime = runtime;
+
+        RubyModule mCsvParser = runtime.defineModule("CsvParser");
+        // TODO: CallbackFactory#getSingletonMethod is deprecated
+        CallbackFactory callbackFactory = runtime.callbackFactory(CsvParserService.class);
+        mCsvParser.defineModuleFunction("parse_line", 
+            callbackFactory.getSingletonMethod("parseLine", RubyString.class));
+        return true;
+    }
+    
+    public static IRubyObject parseLine(IRubyObject recv, RubyString line, Block unusedBlock) {
+        Ruby runtime = recv.getRuntime();
+ 
+        CharSequence seq = line.getValue();
+        int length = seq.length();
+        if (length == 0)
+            return runtime.getNil();
+        
+        int state = UNQUOTED;
+        StringBuilder value = new StringBuilder(length);   // field value, no longer than line
+        RubyArray array = RubyArray.newArray(runtime, DEF_ARRAY_LEN);
+        
+        for (int i = 0; i < length; i++) {
+            char c = seq.charAt(i);
+            switch (c) {
+                case ',':
+                    if (state == UNQUOTED) {
+                        if (value.length() == 0) {
+                            array.append(runtime.getNil());
+                        } 
+                        else {
+                            array.append(RubyString.newString(runtime, value));
+                            value.setLength(0);
+                        }
+                    }
+                    else if (state == IN_QUOTED) {
+                        value.append(c);
+                    }
+                    else if (state == 2) {
+                        array.append(RubyString.newString(runtime, value));
+                        value.setLength(0);
+                        state = UNQUOTED;
+                    }
+                    break;
+                case '"':
+                    if (state == UNQUOTED) {
+                        state = IN_QUOTED;
+                    }
+                    else if (state == IN_QUOTED) {
+                        state = QUOTE_IN_QUOTED;
+                    }
+                    else if (state == QUOTE_IN_QUOTED) {
+                        value.append(c);   // escaped quote
+                        state = IN_QUOTED;
+                    }
+                    break;
+                case '\r':
+                case '\n':
+                    if (state == IN_QUOTED) {
+                        value.append(c);
+                    }
+                    else {
+                        i = length;  // only parse first line if multiline
+                    }
+                    break;
+                default:
+                    value.append(c);
+                    break;
+            }
+        }
+        if (state == UNQUOTED) {
+            if (value.length() == 0) {
+                array.append(runtime.getNil());
+            } 
+            else {
+                array.append(RubyString.newString(runtime, value));
+                value.setLength(0);
+            }
+        }
+        else if (state == QUOTE_IN_QUOTED) {
+            array.append(RubyString.newString(runtime, value));
+        }
+        return array;
+    }
+}

data/ext/csv_parser/extconf.rb

@@ -1,9 +1,15 @@
 #!/usr/bin/ruby -w
 
 require 'mkmf'
+extension_name = 'csv_parser'
+#dir_config(extension_name)
 
 if RUBY_VERSION =~ /1.8/ then
   $CPPFLAGS += " -DRUBY_18"
 end
 
-create_makefile('csv_parser')
+#if CONFIG["arch"] =~ /mswin32|mingw/
+#  $CFLAGS += " -march=i686"
+#end
+
+create_makefile(extension_name)

data/ext/csv_parser/parser.c

@@ -9,10 +9,13 @@
   #include "ruby/io.h"
 #endif
 
-/* default allocated size is 16 */
 #define DEF_ARRAY_LEN 32
 
-static VALUE cFastestCSV;
+#define UNQUOTED 0
+#define IN_QUOTED 1
+#define QUOTE_IN_QUOTED 2
+
+static VALUE mCsvParser;
 
 static VALUE parse_line(VALUE self, VALUE str)
 {
@@ -25,7 +28,7 @@ static VALUE parse_line(VALUE self, VALUE str)
     if (len == 0)
         return Qnil;
     
-    VALUE array = rb_ary_new2(DEF_ARRAY_LEN); 
+    VALUE array = rb_ary_new2(DEF_ARRAY_LEN); /* default allocated size is 16 */
     char value[len];  /* field value, no longer than line */
     int state = 0;
     int index = 0;
@@ -37,51 +40,49 @@ static VALUE parse_line(VALUE self, VALUE str)
         switch (c)
         {
             case ',':
-                if (state == 0) {
+                if (state == UNQUOTED) {
                     rb_ary_push(array, (index == 0 ? Qnil: rb_str_new(value, index)));
                     index = 0;
                 }
-                else if (state == 1) {
+                else if (state == IN_QUOTED) {
                     value[index++] = c;
                 }
-                else if (state == 2) {
+                else if (state == QUOTE_IN_QUOTED) {
                     rb_ary_push(array, rb_str_new(value, index));
                     index = 0;
-                    state = 0;  /* outside quoted */
+                    state = UNQUOTED;
                 }
                 break;
             case '"':
-                if (state == 0) {
-                    state = 1;  /* in quoted */
+                if (state == UNQUOTED) {
+                    state = IN_QUOTED;
                 }
                 else if (state == 1) {
-                    state = 2;  /* quote in quoted */
+                    state = QUOTE_IN_QUOTED;
                 }
-                else if (state == 2) {
+                else if (state == QUOTE_IN_QUOTED) {
                     value[index++] = c;  /* escaped quote */
-                    state = 1;  /* in quoted */
+                    state = IN_QUOTED;
                 }
                 break;
             case 13:  /* \r */
             case 10:  /* \n */
-                if (state == 1) { /* quoted */
+                if (state == IN_QUOTED) {
                     value[index++] = c;
                 }
                 else {
-                    /* only do first line */
-                    i = len;
+                    i = len;  /* only parse first line if multiline */
                 }
-                /* else eat it ??? or return so far */
                 break;
             default:
                 value[index++] = c;
         }
     }
     
-    if (state == 0) {
+    if (state == UNQUOTED) {
         rb_ary_push(array, (index == 0 ? Qnil: rb_str_new(value, index)));
     }
-    else if (state == 2) {
+    else if (state == QUOTE_IN_QUOTED) {
         rb_ary_push(array, rb_str_new(value, index));
     }
     return array;
@@ -89,6 +90,6 @@ static VALUE parse_line(VALUE self, VALUE str)
 
 void Init_csv_parser()
 {
-    cFastestCSV = rb_define_class("FastestCSV", rb_cObject);
-    rb_define_singleton_method(cFastestCSV, "parse_line", parse_line, 1);
+    mCsvParser = rb_define_module("CsvParser");
+    rb_define_module_function(mCsvParser, "parse_line", parse_line, 1);
 }

data/fastest-csv.gemspec

@@ -4,8 +4,8 @@ require File.expand_path('../lib/fastest-csv/version', __FILE__)
 Gem::Specification.new do |gem|
   gem.authors       = ["Maarten Oelering"]
   gem.email         = ["maarten@brightcode.nl"]
-  gem.description   = %q{Fastest standard CSV parser for MRI Ruby}
-  gem.summary       = %q{Fastest standard CSV parser for MRI Ruby}
+  gem.description   = %q{Fastest standard CSV parser for MRI Ruby and JRuby}
+  gem.summary       = %q{Fastest standard CSV parser for MRI Ruby and JRuby}
   gem.homepage      = "https://github.com/brightcode/fastest-csv"
 
   gem.files         = `git ls-files`.split($\)
@@ -14,6 +14,13 @@ Gem::Specification.new do |gem|
   gem.name          = "fastest-csv"
   gem.require_paths = ["lib"]
   gem.version       = FastestCSV::VERSION
+
+  if RUBY_PLATFORM =~ /java/
+    gem.platform = "java"
+    gem.files << "lib/csv_parser.jar"
+  else
+    gem.extensions  = ['ext/csv_parser/extconf.rb']
+  end
   
-  gem.extensions    = ['ext/csv_parser/extconf.rb']
+  gem.add_development_dependency "rake-compiler"
 end

data/lib/fastest-csv/version.rb

@@ -1,3 +1,3 @@
 class FastestCSV
-  VERSION = "0.0.1"
+  VERSION = "0.0.4"
 end

data/lib/fastest_csv.rb

@@ -1,17 +1,30 @@
+# This loads either csv_parser.so, csv_parser.bundle or
+# csv_parser.jar, depending on your Ruby platform and OS
 require 'csv_parser'
 require 'stringio'
 
+# Fast CSV parser using native code
 class FastestCSV
+  include Enumerable
   
-  # This method opens an accounting file and passes each record to the provided +block+.
+  if RUBY_PLATFORM =~ /java/
+    if JRUBY_VERSION =~ /^1\.[0-6]/
+      require 'jruby'
+      org.brightcode.CsvParserService.new.basicLoad(JRuby.runtime)
+    else
+      include_package "org.brightcode"
+    end
+  end
+
+  # Pass each line of the specified +path+ as array to the provided +block+
   def self.foreach(path, &block)
     open(path) do |reader|
       reader.each(&block)
     end
   end
 
-  # This method opens a csv file. It will pass a Reader object to the provided block,
-  # or return a Reader object when no block is provided.
+  # Opens a csv file. Pass a FastestCSV instance to the provided block,
+  # or return it when no block is provided
   def self.open(path, mode = "rb")
     csv = new(File.open(path, mode))
     if block_given?
@@ -25,14 +38,17 @@ class FastestCSV
     end
   end
 
+  # Read all lines from the specified +path+ into an array of arrays
   def self.read(path)
     open(path, "rb") { |csv| csv.read }
   end
 
+  # Alias for FastestCSV.read
   def self.readlines(path)
     read(path)
   end
 
+  # Read all lines from the specified String into an array of arrays
   def self.parse(data, &block)
     csv = new(StringIO.new(data))
     if block.nil?
@@ -45,27 +61,44 @@ class FastestCSV
       csv.each(&block)
     end
   end
+  
+  def self.parse_line(line)
+    CsvParser.parse_line(line)
+  end
 
+  # Create new FastestCSV wrapping the specified IO object
   def initialize(io)
     @io = io
   end
   
+  # Read from the wrapped IO passing each line as array to the specified block
   def each
-    while row = shift
-      yield row
+    if block_given?
+      while row = shift
+        yield row
+      end
+    else
+      to_enum # return enumerator
     end
   end
   
+  # Read all remaining lines from the wrapped IO into an array of arrays
   def read
     table = Array.new
     each {|row| table << row}
     table
   end
   alias_method :readlines, :read
+  
+  # Rewind the underlying IO object and reset line counter
+  def rewind
+    @io.rewind
+  end
 
+  # Read next line from the wrapped IO and return as array or nil at EOF
   def shift
     if line = @io.gets
-      FastestCSV.parse_line(line)
+      CsvParser.parse_line(line)
     else
       nil
     end
@@ -73,6 +106,7 @@ class FastestCSV
   alias_method :gets,     :shift
   alias_method :readline, :shift
   
+  # Close the wrapped IO
   def close
     @io.close
   end
@@ -83,9 +117,9 @@ class FastestCSV
 end
 
 class String
-  # Equivalent to <tt>FasterCSV::parse_line(self, options)</tt>.
+  # Equivalent to <tt>FasterCSV::parse_line(self)</tt>
   def parse_csv
-    FastestCSV.parse_line(self)
+    CsvParser.parse_line(self)
   end
 end
 

data/test/tc_csv_parsing.rb

@@ -13,11 +13,15 @@ require 'fastest_csv'
 # 
 class TestCSVParsing < Test::Unit::TestCase
 
+  if RUBY_PLATFORM =~ /java/
+    include_package "org.brightcode"
+  end
+
   def test_mastering_regex_example
     ex = %Q{Ten Thousand,10000, 2710 ,,"10,000","It's ""10 Grand"", baby",10K}
     assert_equal( [ "Ten Thousand", "10000", " 2710 ", nil, "10,000",
                     "It's \"10 Grand\", baby", "10K" ],
-                  FastestCSV.parse_line(ex) )
+                  CsvParser.parse_line(ex) )
   end
 
   # Pulled from:  http://www.ruby-lang.org/cgi-bin/cvsweb.cgi/ruby/test/csv/test_csv.rb?rev=1.12.2.2;content-type=text%2Fplain
@@ -49,7 +53,7 @@ class TestCSVParsing < Test::Unit::TestCase
       ["foo,\"\r\n\n\",baz", ["foo", "\r\n\n", "baz"]],
       ["foo,\"foo,bar\",baz", ["foo", "foo,bar", "baz"]],
       [";,;", [";", ";"]] ].each do |csv_test|
-      assert_equal(csv_test.last, FastestCSV.parse_line(csv_test.first))
+      assert_equal(csv_test.last, CsvParser.parse_line(csv_test.first))
     end
     
     [ ["foo,\"\"\"\"\"\",baz", ["foo", "\"\"", "baz"]],
@@ -68,7 +72,7 @@ class TestCSVParsing < Test::Unit::TestCase
       ["foo,bar", ["foo", "bar"]],
       ["foo,\"\r\n\n\",baz", ["foo", "\r\n\n", "baz"]],
       ["foo,\"foo,bar\",baz", ["foo", "foo,bar", "baz"]] ].each do |csv_test|
-      assert_equal(csv_test.last, FastestCSV.parse_line(csv_test.first))
+      assert_equal(csv_test.last, CsvParser.parse_line(csv_test.first))
      end
   end
 
@@ -91,20 +95,20 @@ class TestCSVParsing < Test::Unit::TestCase
       [%Q{,"\r"},             [nil,"\r"]],
       [%Q{"\r\n,"},           ["\r\n,"]],
       [%Q{"\r\n,",},          ["\r\n,", nil]] ].each do |edge_case|
-        assert_equal(edge_case.last, FastestCSV.parse_line(edge_case.first))
+        assert_equal(edge_case.last, CsvParser.parse_line(edge_case.first))
       end
   end
 
   def test_james_edge_cases
     # A read at eof? should return nil.
-    assert_equal(nil, FastestCSV.parse_line(""))
+    assert_equal(nil, CsvParser.parse_line(""))
     # 
     # With CSV it's impossible to tell an empty line from a line containing a
     # single +nil+ field.  The standard CSV library returns <tt>[nil]</tt>
     # in these cases, but <tt>Array.new</tt> makes more sense to me.
     # 
     #assert_equal(Array.new, FastestCSV.parse_line("\n1,2,3\n"))
-    assert_equal([nil], FastestCSV.parse_line("\n1,2,3\n"))
+    assert_equal([nil], CsvParser.parse_line("\n1,2,3\n"))
   end
 
   def test_rob_edge_cases
@@ -119,7 +123,7 @@ class TestCSVParsing < Test::Unit::TestCase
       [%Q{"a\r\n\r\na","two CRLFs"},       ["a\r\n\r\na", 'two CRLFs']],
       [%Q{with blank,"start\n\nfinish"\n}, ['with blank', "start\n\nfinish"]],
     ].each do |edge_case|
-      assert_equal(edge_case.last, FastestCSV.parse_line(edge_case.first))
+      assert_equal(edge_case.last, CsvParser.parse_line(edge_case.first))
     end
   end
 

data/test/tc_interface.rb

@@ -116,4 +116,13 @@ class TestFastestCSVInterface < Test::Unit::TestCase
     @expected = [%w{1 2} + ['3' * long_field_length]]
     test_shift
   end
+  
+  def test_enumerable
+    FastestCSV.open(@path) do |csv|
+      assert(csv.include?(["1", "2", "3"]))
+      csv.rewind
+      assert_equal([["1", "2", "3"], ["4", "5"]], csv.to_a)
+    end
+  end
+  
 end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: fastest-csv
 version: !ruby/object:Gem::Version
-  version: 0.0.1
+  version: 0.0.4
   prerelease: 
 platform: ruby
 authors:
@@ -9,9 +9,25 @@ authors:
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2012-06-28 00:00:00.000000000 Z
-dependencies: []
-description: Fastest standard CSV parser for MRI Ruby
+date: 2013-08-15 00:00:00.000000000 Z
+dependencies:
+- !ruby/object:Gem::Dependency
+  name: rake-compiler
+  requirement: !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ! '>='
+      - !ruby/object:Gem::Version
+        version: '0'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ! '>='
+      - !ruby/object:Gem::Version
+        version: '0'
+description: Fastest standard CSV parser for MRI Ruby and JRuby
 email:
 - maarten@brightcode.nl
 executables: []
@@ -24,6 +40,8 @@ files:
 - LICENSE
 - README.md
 - Rakefile
+- ext/csv_parser/CsvParser.java
+- ext/csv_parser/CsvParserService.java
 - ext/csv_parser/extconf.rb
 - ext/csv_parser/parser.c
 - fastest-csv.gemspec
@@ -52,10 +70,10 @@ required_rubygems_version: !ruby/object:Gem::Requirement
       version: '0'
 requirements: []
 rubyforge_project: 
-rubygems_version: 1.8.24
+rubygems_version: 1.8.25
 signing_key: 
 specification_version: 3
-summary: Fastest standard CSV parser for MRI Ruby
+summary: Fastest standard CSV parser for MRI Ruby and JRuby
 test_files:
 - test/tc_csv_parsing.rb
 - test/tc_interface.rb