theman 0.0.2 → 0.0.3

data/.gitignore

@@ -14,9 +14,9 @@ tmtags
 *.swp
 
 ## PROJECT::GENERAL
-coverage
-rdoc
-pkg
+pkg/*
+*.gem
+.bundle
 
 ## PROJECT::SPECIFIC
 spec/database.yml

data/Gemfile

@@ -0,0 +1,4 @@
+source :gemcutter
+
+# Specify your gem's dependencies in theman.gemspec
+gemspec

data/Rakefile

@@ -1,18 +1,2 @@
-require 'rubygems'
-require 'rake'
-
-begin
-  require 'jeweler'
-  Jeweler::Tasks.new do |gem|
-    gem.name = "theman"
-    gem.summary = %Q{PostgreSQL AR temporary table generator using PostgreSQL COPY}
-    gem.description = %Q{longer description of your gem}
-    gem.email = "rufuspost@gmail.com"
-    gem.homepage = "http://github.com/mynameisrufus/theman"
-    gem.authors = ["Rufus Post"]
-    gem.add_development_dependency "rspec", ">= 2.0.0.beta.20"
-  end
-  Jeweler::GemcutterTasks.new
-rescue LoadError
-  puts "Jeweler (or a dependency) not available. Install it with: gem install jeweler"
-end
+require 'bundler'
+Bundler::GemHelper.install_tasks

data/lib/theman/themans_agency.rb

@@ -1,8 +1,15 @@
 module Theman
   class Agency
-    attr_reader :instance, :table_proc, :null_replacements, :stream_location, :column_names
+    attr_reader :instance, :column_names, :null_replacements, :sed_commands
 
-    def initialize(*args)
+    attr_writer :stream
+    
+    def initialize(stream = nil, parent = ::ActiveRecord::Base)
+      # source of the data
+      @stream = stream
+
+      # create a new class that extends an active record model
+      # use instance_parent(klass) if not ActiveRecord::Base
       cabinet_id = "c#{10.times.map{rand(9)}.join}" 
       @column_names = {}
       @instance = Class.new(parent) do
@@ -16,34 +23,24 @@ module Theman
           end
         EOV
       end
-      yield self if block_given?
-      unless args.first.nil?
-        @stream_location = args.first
+
+      # if stream given table will be created
+      # other wise create_table and pipe_it will need to called
+      # proceduraly
+      if stream
+        if block_given?
+          yield self
+        end
         create_table
         pipe_it
       end
     end
-    
-    def instance_parent(klass)
-      @parent = klass
-    end
 
-    def table(&block)
+    def table
       yield self if block_given?
     end
 
-    def nulls(*args)
-      @null_replacements = args
-    end
-
-    def seds(*args)
-      @extra_seds = args
-    end
-    
-    def symbolize(name)
-      name.gsub(/ /,"_").gsub(/\W/, "").downcase.to_sym
-    end
-
+    # overide ActiveRecord column types to be used in a block
     %w( string text integer float decimal datetime timestamp time date binary boolean ).each do |column_type|
       class_eval <<-EOV, __FILE__, __LINE__ + 1
         def #{column_type}(*args)
@@ -51,27 +48,14 @@ module Theman
         end
       EOV
     end
-    
-    protected
-    def sed_to_s
-      seds = []
-      seds << "| sed #{nulls_to_sed.join(" ")}" unless @null_replacements.nil?
-      seds << "| sed #{@extra_seds.join("| sed ")}" unless @extra_seds.nil?
-      return seds.join(" ") unless seds.empty?
-    end
 
-    def nulls_to_sed
-      @null_replacements.map do |null|
-        "-e 's/#{null.source}//g'"
-      end
-    end
-
-    def parent
-      @parent ||= ::ActiveRecord::Base
+    # overides the default string type column
+    def column(name, type, options)
+      @column_names.merge! name.to_sym => [name, type, options]
     end
 
     def create_table
-      f = File.open(stream_location, 'r')
+      f = File.open(@stream, 'r')
       instance.connection.create_table(instance.table_name, :temporary => true, :id => false) do |t|
         f.each_line do |line|
           line.split(/,/).each do |col|
@@ -87,14 +71,12 @@ module Theman
       end
     end
     
-    def column(name, type, options)
-      @column_names.merge! name.to_sym => [name, type, options]
-    end
-
+    # use postgress COPY command using STDIN with CSV HEADER
+    # reads chunks of 8192 bytes to save memory
     def pipe_it(l = "")
       raw = instance.connection.raw_connection
       raw.query "COPY #{instance.table_name} FROM STDIN WITH CSV HEADER"
-      command = "cat #{stream_location} #{sed_to_s}"
+      command = "cat #{@stream} #{seds_join}"
       f = IO.popen(command)
       begin
         while f.read(8192, l)
@@ -105,5 +87,34 @@ module Theman
       end
       raw.put_copy_end
     end
+
+    def nulls(*args)
+      @null_replacements = args
+    end
+
+    def seds(*args)
+      @sed_commands = args
+    end
+    
+    def symbolize(name)
+      name.gsub(/ /,"_").gsub(/\W/, "").downcase.to_sym
+    end
+    
+    # join together the sed commands to apply to stream
+    def seds_join(commands = [])
+      unless null_replacements.nil?
+        commands << "| sed #{nulls_to_sed.join(" ")}"
+      end
+      unless sed_commands.nil?
+        commands << "| sed #{sed_commands.join("| sed ")}"
+      end
+      commands.join(" ")
+    end
+
+    def nulls_to_sed
+      @null_replacements.map do |null|
+        "-e 's/#{null.source}//g'"
+      end
+    end
   end
 end

data/lib/theman/version.rb

@@ -0,0 +1,3 @@
+module Theman
+  VERSION = "0.0.3"
+end

data/spec/spec_helper.rb

@@ -1,15 +1,16 @@
 $LOAD_PATH.unshift(File.dirname(__FILE__))
 $LOAD_PATH.unshift(File.join(File.dirname(__FILE__), '..', 'lib'))
-require 'rails'
 
 require 'active_record'
-Rails.env = 'test'
+require 'fileutils'
+require 'logger'
+require 'rspec'
+require 'rspec/autorun'
+require 'theman'
+
 ActiveRecord::Base.configurations = YAML.load_file(File.join("spec", "database.yml"))
 FileUtils.mkdir_p "#{Dir.pwd}/log"
-logfile= File.new("#{Dir.pwd}/log/database.log", "w")
+logfile = "#{Dir.pwd}/log/database.log"
 ActiveRecord::Base.logger = Logger.new(File.open(logfile, 'w'))
-ActiveRecord::Base.establish_connection(ActiveRecord::Base.configurations.fetch(Rails.env))
+ActiveRecord::Base.establish_connection(ActiveRecord::Base.configurations.fetch('test'))
 
-require 'theman'
-require 'rspec'
-require 'rspec/autorun'

data/theman.gemspec

@@ -1,55 +1,24 @@
-# Generated by jeweler
-# DO NOT EDIT THIS FILE DIRECTLY
-# Instead, edit Jeweler::Tasks in Rakefile, and run the gemspec command
 # -*- encoding: utf-8 -*-
+require File.expand_path("../lib/theman/version", __FILE__)
 
 Gem::Specification.new do |s|
-  s.name = %q{theman}
-  s.version = "0.0.2"
+  s.name        = "theman"
+  s.version     = Theman::VERSION
+  s.platform    = Gem::Platform::RUBY
+  s.authors     = ["Rufus Post"]
+  s.email       = ["rufuspost@gmail.com"]
+  s.homepage    = "http://github.com/mynameisrufus/theman"
+  s.summary     = "PostgreSQL AR temporary table generator using PostgreSQL COPY"
+  s.description = "FasterCSV is great and all but when you get to 100mb files it takes a while and you may only be looking for certain records that match some criteria, enter theman"
 
-  s.required_rubygems_version = Gem::Requirement.new(">= 0") if s.respond_to? :required_rubygems_version=
-  s.authors = ["Rufus Post"]
-  s.date = %q{2010-10-26}
-  s.description = %q{longer description of your gem}
-  s.email = %q{rufuspost@gmail.com}
-  s.extra_rdoc_files = [
-    "README.rdoc"
-  ]
-  s.files = [
-    ".document",
-     ".gitignore",
-     "README.rdoc",
-     "Rakefile",
-     "VERSION",
-     "lib/theman.rb",
-     "lib/theman/themans_agency.rb",
-     "spec/fixtures/temp_one.csv",
-     "spec/fixtures/temp_two.csv",
-     "spec/spec_helper.rb",
-     "spec/theman_spec.rb",
-     "theman.gemspec"
-  ]
-  s.homepage = %q{http://github.com/mynameisrufus/theman}
-  s.rdoc_options = ["--charset=UTF-8"]
-  s.require_paths = ["lib"]
-  s.rubygems_version = %q{1.3.7}
-  s.summary = %q{PostgreSQL AR temporary table generator using PostgreSQL COPY}
-  s.test_files = [
-    "spec/theman_spec.rb",
-     "spec/spec_helper.rb"
-  ]
+  s.required_rubygems_version = ">= 1.3.6"
+  s.rubyforge_project         = "theman"
 
-  if s.respond_to? :specification_version then
-    current_version = Gem::Specification::CURRENT_SPECIFICATION_VERSION
-    s.specification_version = 3
+  s.add_development_dependency "bundler", ">= 1.0.0"
+  s.add_development_dependency "rspec", ">= 2.0.0"
+  s.add_development_dependency "activerecord", ">= 3.0.0"
 
-    if Gem::Version.new(Gem::VERSION) >= Gem::Version.new('1.2.0') then
-      s.add_development_dependency(%q<rspec>, [">= 2.0.0.beta.20"])
-    else
-      s.add_dependency(%q<rspec>, [">= 2.0.0.beta.20"])
-    end
-  else
-    s.add_dependency(%q<rspec>, [">= 2.0.0.beta.20"])
-  end
+  s.files        = `git ls-files`.split("\n")
+  s.executables  = `git ls-files`.split("\n").map{|f| f =~ /^bin\/(.*)/ ? $1 : nil}.compact
+  s.require_path = 'lib'
 end
-

metadata

@@ -5,8 +5,8 @@ version: !ruby/object:Gem::Version
   segments: 
   - 0
   - 0
-  - 2
-  version: 0.0.2
+  - 3
+  version: 0.0.3
 platform: ruby
 authors: 
 - Rufus Post
@@ -14,11 +14,11 @@ autorequire:
 bindir: bin
 cert_chain: []
 
-date: 2010-10-26 00:00:00 +11:00
+date: 2010-10-28 00:00:00 +11:00
 default_executable: 
 dependencies: 
 - !ruby/object:Gem::Dependency 
-  name: rspec
+  name: bundler
   prerelease: false
   requirement: &id001 !ruby/object:Gem::Requirement 
     none: false
@@ -26,30 +26,59 @@ dependencies:
     - - ">="
       - !ruby/object:Gem::Version 
         segments: 
-        - 2
+        - 1
         - 0
         - 0
-        - beta
-        - 20
-        version: 2.0.0.beta.20
+        version: 1.0.0
   type: :development
   version_requirements: *id001
-description: longer description of your gem
-email: rufuspost@gmail.com
+- !ruby/object:Gem::Dependency 
+  name: rspec
+  prerelease: false
+  requirement: &id002 !ruby/object:Gem::Requirement 
+    none: false
+    requirements: 
+    - - ">="
+      - !ruby/object:Gem::Version 
+        segments: 
+        - 2
+        - 0
+        - 0
+        version: 2.0.0
+  type: :development
+  version_requirements: *id002
+- !ruby/object:Gem::Dependency 
+  name: activerecord
+  prerelease: false
+  requirement: &id003 !ruby/object:Gem::Requirement 
+    none: false
+    requirements: 
+    - - ">="
+      - !ruby/object:Gem::Version 
+        segments: 
+        - 3
+        - 0
+        - 0
+        version: 3.0.0
+  type: :development
+  version_requirements: *id003
+description: FasterCSV is great and all but when you get to 100mb files it takes a while and you may only be looking for certain records that match some criteria, enter theman
+email: 
+- rufuspost@gmail.com
 executables: []
 
 extensions: []
 
-extra_rdoc_files: 
-- README.rdoc
+extra_rdoc_files: []
+
 files: 
-- .document
 - .gitignore
+- Gemfile
 - README.rdoc
 - Rakefile
-- VERSION
 - lib/theman.rb
 - lib/theman/themans_agency.rb
+- lib/theman/version.rb
 - spec/fixtures/temp_one.csv
 - spec/fixtures/temp_two.csv
 - spec/spec_helper.rb
@@ -60,8 +89,8 @@ homepage: http://github.com/mynameisrufus/theman
 licenses: []
 
 post_install_message: 
-rdoc_options: 
-- --charset=UTF-8
+rdoc_options: []
+
 require_paths: 
 - lib
 required_ruby_version: !ruby/object:Gem::Requirement 
@@ -78,15 +107,16 @@ required_rubygems_version: !ruby/object:Gem::Requirement
   - - ">="
     - !ruby/object:Gem::Version 
       segments: 
-      - 0
-      version: "0"
+      - 1
+      - 3
+      - 6
+      version: 1.3.6
 requirements: []
 
-rubyforge_project: 
+rubyforge_project: theman
 rubygems_version: 1.3.7
 signing_key: 
 specification_version: 3
 summary: PostgreSQL AR temporary table generator using PostgreSQL COPY
-test_files: 
-- spec/theman_spec.rb
-- spec/spec_helper.rb
+test_files: []
+

data/.document

@@ -1,5 +0,0 @@
-README.rdoc
-lib/**/*.rb
-bin/*
-features/**/*.feature
-LICENSE

data/VERSION

@@ -1 +0,0 @@
-0.0.2