Merge pull request #20 from gitlabhq/gitlab-grit-gem

WIP: gitlab-grit gem to replace grit fork

.gitignore

 pkg
 .DS_Store
 repos/grit/
+*.gem

.travis.yml

 env:
   - TRAVIS=true
+before_install:
+  - sudo apt-get install libicu-dev -y
 branches:
   only:
     - 'master'
 rvm:
   - 1.9.3
   - 2.0.0
+before_script:
+  - "bundle install"
 script: "bundle exec rake test"

Gemfile

 source "http://rubygems.org"
+gemspec
+
 group :development, :test do
+  gem 'pry'
   gem 'rake'
   gem 'posix-spawn', "~> 0.3.6"
   gem 'mime-types', "~> 1.15"
   gem 'diff-lcs', "~> 1.1"
   gem 'mocha', "~> 0.13.2"
   gem 'coveralls', require: false
+  gem 'rspec', '~> 2.11'
 end

Gemfile.lock

+PATH
+  remote: .
+  specs:
+    gitlab-grit (2.5.0)
+      charlock_holmes (~> 0.6.9)
+      diff-lcs (~> 1.1)
+      mime-types (~> 1.15)
+      posix-spawn (~> 0.3.6)
+
 GEM
   remote: http://rubygems.org/
   specs:
+    charlock_holmes (0.6.9.4)
+    coderay (1.0.9)
     colorize (0.5.8)
     coveralls (0.6.2)
       colorize
@@ -10,18 +21,32 @@ GEM
       thor
     diff-lcs (1.1.3)
     metaclass (0.0.1)
+    method_source (0.8.1)
     mime-types (1.18)
     mocha (0.13.2)
       metaclass (~> 0.0.1)
     multi_json (1.7.2)
     posix-spawn (0.3.6)
+    pry (0.9.12.2)
+      coderay (~> 1.0.5)
+      method_source (~> 0.8)
+      slop (~> 3.4)
     rake (0.9.2.2)
     rest-client (1.6.7)
       mime-types (>= 1.16)
+    rspec (2.13.0)
+      rspec-core (~> 2.13.0)
+      rspec-expectations (~> 2.13.0)
+      rspec-mocks (~> 2.13.0)
+    rspec-core (2.13.1)
+    rspec-expectations (2.13.0)
+      diff-lcs (>= 1.1.3, < 2.0)
+    rspec-mocks (2.13.1)
     simplecov (0.7.1)
       multi_json (~> 1.0)
       simplecov-html (~> 0.7.1)
     simplecov-html (0.7.1)
+    slop (3.4.4)
     thor (0.18.0)
 PLATFORMS
@@ -30,7 +55,10 @@ PLATFORMS
 DEPENDENCIES
   coveralls
   diff-lcs (~> 1.1)
+  gitlab-grit!
   mime-types (~> 1.15)
   mocha (~> 0.13.2)
   posix-spawn (~> 0.3.6)
+  pry
   rake
+  rspec (~> 2.11)

Rakefile

@@ -2,48 +2,7 @@ require 'rubygems'
 require 'rake'
 require 'date'
-#############################################################################
-#
-# Helper functions
-#
-#############################################################################
-
-def name
-  @name ||= Dir['*.gemspec'].first.split('.').first
-end
-
-def version
-  line = File.read("lib/#{name}.rb")[/^\s*VERSION\s*=\s*.*/]
-  line.match(/.*VERSION\s*=\s*['"](.*)['"]/)[1]
-end
-
-def date
-  Date.today.to_s
-end
-
-def rubyforge_project
-  name
-end
-
-def gemspec_file
-  "#{name}.gemspec"
-end
-
-def gem_file
-  "#{name}-#{version}.gem"
-end
-
-def replace_header(head, header_name)
-  head.sub!(/(\.#{header_name}\s*= ').*'/) { "#{$1}#{send(header_name)}'"}
-end
-
-#############################################################################
-#
-# Standard tasks
-#
-#############################################################################
-
-task :default => :test
+task default: :test
 require 'rake/testtask'
 Rake::TestTask.new(:test) do |test|
@@ -52,98 +11,7 @@ Rake::TestTask.new(:test) do |test|
   test.verbose = true
 end
-desc "Generate RCov test coverage and open in your browser"
-task :coverage do
-  require 'rcov'
-  sh "rm -fr coverage"
-  sh "rcov test/test_*.rb"
-  sh "open coverage/index.html"
-end
-
-require 'rdoc/task'
-Rake::RDocTask.new do |rdoc|
-  rdoc.rdoc_dir = 'rdoc'
-  rdoc.title = "#{name} #{version}"
-  rdoc.rdoc_files.include('README*')
-  rdoc.rdoc_files.include('lib/**/*.rb')
-end
-
 desc "Open an irb session preloaded with this library"
 task :console do
   sh "irb -rubygems -r ./lib/#{name}.rb"
 end
-
-#############################################################################
-#
-# Custom tasks (add your own tasks here)
-#
-#############################################################################
-
-desc "Upload site to Rubyforge"
-task :site do
-  sh "scp -r doc/* mojombo@grit.rubyforge.org:/var/www/gforge-projects/grit"
-end
-
-#############################################################################
-#
-# Packaging tasks
-#
-#############################################################################
-
-task :release => :build do
-  unless `git branch` =~ /^\* master$/
-    puts "You must be on the master branch to release!"
-    exit!
-  end
-  sh "git commit --allow-empty -a -m 'Release #{version}'"
-  sh "git tag v#{version}"
-  sh "git push origin master"
-  sh "git push origin v#{version}"
-  sh "gem push pkg/#{name}-#{version}.gem"
-end
-
-task :build => :gemspec do
-  sh "mkdir -p pkg"
-  sh "gem build #{gemspec_file}"
-  sh "mv #{gem_file} pkg"
-end
-
-task :gemspec => :validate do
-  # read spec file and split out manifest section
-  spec = File.read(gemspec_file)
-  head, manifest, tail = spec.split("  # = MANIFEST =\n")
-
-  # replace name version and date
-  replace_header(head, :name)
-  replace_header(head, :version)
-  replace_header(head, :date)
-  #comment this out if your rubyforge_project has a different name
-  replace_header(head, :rubyforge_project)
-
-  # determine file list from git ls-files
-  files = `git ls-files`.
-    split("\n").
-    sort.
-    reject { |file| file =~ /^\./ }.
-    reject { |file| file =~ /^(rdoc|pkg|test)/ }.
-    map { |file| "    #{file}" }.
-    join("\n")
-
-  # piece file back together and write
-  manifest = "  s.files = %w[\n#{files}\n  ]\n"
-  spec = [head, manifest, tail].join("  # = MANIFEST =\n")
-  File.open(gemspec_file, 'w') { |io| io.write(spec) }
-  puts "Updated #{gemspec_file}"
-end
-
-task :validate do
-  libfiles = Dir['lib/*'] - ["lib/#{name}.rb", "lib/#{name}"]
-  unless libfiles.empty?
-    puts "Directory `lib` should only contain a `#{name}.rb` file and `#{name}` dir."
-    exit!
-  end
-  unless Dir['VERSION*'].empty?
-    puts "A `VERSION` file at root level violates Gem best practices."
-    exit!
-  end
-end

gitlab-grit.gemspec

+Gem::Specification.new do |s|
+  s.name        = 'gitlab-grit'
+  s.version     = '2.5.1'
+  s.date        = '2013-05-06'
+  s.license     = 'MIT'
+  s.summary     = "Ruby Git bindings."
+  s.description = "Grit is a Ruby library for extracting information from a git repository in an object oriented manner. GitLab fork"
+  s.authors     = ["Tom Preston-Werner", "Scott Chacon", "Dmitriy Zaporozhets"]
+  s.email       = 'm@gitlabhq.com'
+  s.homepage    = 'http://github.com/gitlabhq/grit'
+  s.require_paths = %w[lib]
+  s.rdoc_options = ["--charset=UTF-8"]
+  s.extra_rdoc_files = %w[README.md LICENSE]
+  s.files = `git ls-files lib/`.split("\n")
+  s.test_files = s.files.select { |path| path =~ /^test\/test_.*\.rb/ }
+
+  s.add_dependency("charlock_holmes", "~> 0.6.9")
+  s.add_dependency('posix-spawn', "~> 0.3.6")
+  s.add_dependency('mime-types', "~> 1.15")
+  s.add_dependency('diff-lcs', "~> 1.1")
+  s.add_development_dependency('mocha')
+end

grit.gemspec

-Gem::Specification.new do |s|
-  s.specification_version = 2 if s.respond_to? :specification_version=
-  s.required_rubygems_version = Gem::Requirement.new(">= 0") if s.respond_to? :required_rubygems_version=
-  s.rubygems_version = '1.3.5'
-
-  s.name              = 'grit'
-  s.version           = '2.5.0'
-  s.date              = '2013-03-02'
-  s.rubyforge_project = 'grit'
-
-  s.summary     = "Ruby Git bindings."
-  s.description = "Grit is a Ruby library for extracting information from a git repository in an object oriented manner."
-
-  s.authors  = ["Tom Preston-Werner", "Scott Chacon"]
-  s.email    = 'm@gitlabhq.com'
-  s.homepage = 'http://github.com/gitlabhq/grit'
-
-  s.require_paths = %w[lib]
-
-  s.rdoc_options = ["--charset=UTF-8"]
-  s.extra_rdoc_files = %w[README.md LICENSE]
-
-  s.add_dependency('posix-spawn', "~> 0.3.6")
-  s.add_dependency('mime-types', "~> 1.15")
-  s.add_dependency('diff-lcs', "~> 1.1")
-
-  s.add_development_dependency('mocha')
-
-  # = MANIFEST =
-  s.files = %w[
-    API.txt
-    History.txt
-    LICENSE
-    PURE_TODO
-    README.md
-    Rakefile
-    benchmarks.rb
-    benchmarks.txt
-    examples/ex_add_commit.rb
-    examples/ex_index.rb
-    grit.gemspec
-    lib/grit.rb
-    lib/grit/actor.rb
-    lib/grit/blame.rb
-    lib/grit/blob.rb
-    lib/grit/commit.rb
-    lib/grit/commit_stats.rb
-    lib/grit/config.rb
-    lib/grit/diff.rb
-    lib/grit/errors.rb
-    lib/grit/git-ruby.rb
-    lib/grit/git-ruby/commit_db.rb
-    lib/grit/git-ruby/git_object.rb
-    lib/grit/git-ruby/internal/file_window.rb
-    lib/grit/git-ruby/internal/loose.rb
-    lib/grit/git-ruby/internal/pack.rb
-    lib/grit/git-ruby/internal/raw_object.rb
-    lib/grit/git-ruby/repository.rb
-    lib/grit/git.rb
-    lib/grit/index.rb
-    lib/grit/lazy.rb
-    lib/grit/merge.rb
-    lib/grit/ref.rb
-    lib/grit/repo.rb
-    lib/grit/ruby1.9.rb
-    lib/grit/status.rb
-    lib/grit/submodule.rb
-    lib/grit/tag.rb
-    lib/grit/tree.rb
-  ]
-  # = MANIFEST =
-
-  s.test_files = s.files.select { |path| path =~ /^test\/test_.*\.rb/ }
-end

lib/grit.rb

@@ -74,3 +74,7 @@ module Grit
     VERSION
   end
 end
+
+# Include grit_ext
+require 'charlock_holmes'
+require 'grit_ext'

lib/grit_ext.rb

+require "charlock_holmes"
+require "grit_ext/actor"
+require "grit_ext/blob"
+require "grit_ext/commit"
+require "grit_ext/tree"
+require "grit_ext/diff"
+require "grit_ext/version"
+
+module GritExt
+  extend self
+
+  def encode!(message)
+    return nil unless message.respond_to? :force_encoding
+
+    # if message is utf-8 encoding, just return it
+    message.force_encoding("UTF-8")
+    return message if message.valid_encoding?
+
+    # return message if message type is binary
+    detect = CharlockHolmes::EncodingDetector.detect(message)
+    return message.force_encoding("BINARY") if detect && detect[:type] == :binary
+
+    # encoding message to detect encoding
+    if detect && detect[:encoding]
+      message.force_encoding(detect[:encoding])
+    end
+
+    # encode and clean the bad chars
+    message.replace clean(message)
+  rescue
+    encoding = detect ? detect[:encoding] : "unknown"
+    "--broken encoding: #{encoding}"
+  end
+
+  private
+  def clean(message)
+    message.encode("UTF-16BE", :undef => :replace, :invalid => :replace, :replace => "")
+           .encode("UTF-8")
+           .gsub("\0".encode("UTF-8"), "")
+  end
+end

lib/grit_ext/actor.rb

+module Grit
+  class Actor
+
+    alias_method :old_name, :name
+    alias_method :old_email, :email
+
+    def name
+      GritExt.encode! old_name
+    end
+
+    def email
+      GritExt.encode! old_email
+    end
+  end
+end

lib/grit_ext/blob.rb

+module Grit
+  class Blob
+
+    alias_method :old_name, :name
+    alias_method :old_data, :data
+
+    def name
+      GritExt.encode! old_name
+    end
+
+    def data
+      GritExt.encode! old_data
+    end
+
+    class << self
+      alias_method :old_blame, :blame
+
+      def blame(repo, commit, file)
+        old_blame(repo, commit, file).map do |b,lines|
+          [b, GritExt.encode!(lines.join('\n')).split('\n')]
+        end
+      end
+    end
+
+  end
+end

lib/grit_ext/commit.rb

+module Grit
+  class Commit
+
+    alias_method :old_message, :message
+    alias_method :old_short_message, :short_message
+
+    def message
+      GritExt.encode! old_message
+    end
+
+    def short_message
+      GritExt.encode! old_short_message
+    end
+  end
+end

lib/grit_ext/diff.rb

+module Grit
+  class Diff
+
+    def old_path
+      GritExt.encode! @a_path
+    end
+
+    def new_path
+      GritExt.encode! @b_path
+    end
+
+    def diff
+      if @diff.nil?
+        @diff = ""
+      else
+        lines = @diff.lines.to_a
+        path = GritExt.encode! lines.shift(2).join
+        body = GritExt.encode! lines.join
+        @diff = path + body
+      end
+    end
+  end
+end

lib/grit_ext/tag.rb

+module Grit
+  class Tag
+
+    alias_method :old_message, :message
+
+    def message
+      GritExt.encode! old_message
+    end
+  end
+end

lib/grit_ext/tree.rb

+module Grit
+  class Tree
+
+    alias_method :old_name, :name
+
+    def name
+      GritExt.encode! old_name
+    end
+  end
+end

lib/grit_ext/version.rb

+module GritExt
+  extend self
+
+  def version
+    "0.8.1"
+  end
+end

test/helper.rb

@@ -8,7 +8,7 @@ require File.join(File.dirname(__FILE__), *%w[.. lib grit])
 require 'rubygems'
 require 'test/unit'
 require 'mocha/setup'
-
+require 'pry'
 REPOS_PATH = File.expand_path(File.join(File.dirname(__FILE__), '..', 'repos'))
 GRIT_REPO = ENV["GRIT_REPO"] || File.join(REPOS_PATH, 'grit')

test/test_blob.rb

@@ -53,7 +53,7 @@ class TestBlob < Test::Unit::TestCase
     Git.any_instance.expects(:blame).returns(fixture('blame'))
     b = Blob.blame(@r, 'master', 'lib/grit.rb')
     assert_equal 13, b.size
-    assert_equal 25, b.inject(0) { |acc, x| acc + x.last.size }
+    assert_equal 24, b.inject(0) { |acc, x| acc + x.last.size }
     assert_equal b[0].first.object_id, b[9].first.object_id
     c = b.first.first
     c.expects(:__bake__).times(0)
@@ -80,4 +80,4 @@ class TestBlob < Test::Unit::TestCase
     @b = Blob.create(@r, :name => 'foo/bar.rb')
     assert_equal "bar.rb", @b.basename
   end
-end
\ No newline at end of file
+end

test/test_diff.rb

@@ -13,7 +13,7 @@ class TestDiff < Test::Unit::TestCase
     diffs = Grit::Diff.list_from_string(@r, output)
     assert_equal 2,   diffs.size
     assert_equal 10,  diffs.first.diff.split("\n").size
-    assert_nil        diffs.last.diff
+    assert_equal "",  diffs.last.diff
   end
   def test_list_from_string_with_renames

test/test_encoding.rb

+# encoding: UTF-8
+
+require File.dirname(__FILE__) + '/helper'
+
+class TestEncoding < Test::Unit::TestCase
+  def test_nil_message
+    message = GritExt.encode! nil
+    assert_equal message, nil
+  end
+
+  def test_binary_message
+    message = "\xFF\xD8\xFF\xE0"
+    encoded_message = GritExt.encode!(message)
+    assert_equal encoded_message.bytes.to_a, message.bytes.to_a
+    assert_equal message.encoding.name, "UTF-8"
+  end
+
+  def test_invalid_encoding
+    message = GritExt.encode!("yummy\xE2 \xF0\x9F\x8D\x94 \x9F\x8D\x94")
+    assert_equal message, "yummy 🍔 "
+    assert_equal message.encoding.name, "UTF-8"
+  end
+
+  def test_encode_string
+    message = GritExt.encode!("{foo \xC3 'bar'}")
+    assert_equal message, "{foo à 'bar'}"
+    assert_equal message.encoding.name, "UTF-8"
+
+    message = "我爱你".encode("GBK")
+    assert_equal message.encoding.name, "GBK"
+
+    GritExt.encode!(message)
+    assert_equal message, "我爱你"
+    assert_equal message.encoding.name, "UTF-8"
+  end
+end
+