lunar 0.1.0 → 0.1.1

data/VERSION

@@ -1 +1 @@
-0.1.0
+0.1.1

data/examples/ohm.rb

@@ -3,21 +3,35 @@ class Item < Ohm::Model
   attribute :description
 
   def index
-    Lunar::Index.create(
-      self.class.name, 
-      self.id,
-      :name => name,
-      :description => description
-    )
-
-    index = Lunar::Index.new('Item')
-    index.attrs = { :name => name, :description => description }
-    index.create
+    Lunar::Index.create 'Item' do |i|
+      i.key  id
+      i.attr :name, name
+      i.attr :description, description
+    end
+    
+    # You can also do this, no problem
+    Lunar::Index.create Item do |i|
+      i.key  id
+      i.attr :name, name
+      i.attr :description, description
+    end
 
-    Luner::Index.create 'Item' do |i|
+    # Or to avoid name ties...
+    Lunar::Index.create self.class do |i|
       i.key  id
       i.attr :name, name
       i.attr :description, description
     end
   end
 end
+
+# Searching...
+# You can just straight out search keywords
+Lunar.search(Item, "iphone")
+
+# Or opt to filter by field
+Lunar.search(Item, :name => "iphone", :description => "mobile")
+
+# Or using the pagination gem with this:
+@items = Lunar.search(Item, "iphone")
+paginate @items, :per_page => 10, :page => 1

data/lib/lunar.rb

@@ -2,11 +2,28 @@ require 'base64'
 require File.join(File.dirname(__FILE__), '..', 'vendor', 'nest', 'nest')
 
 module Lunar
-  autoload :Scoring, 'lunar/scoring'
-  autoload :Doc,     'lunar/doc'
-  autoload :Index,   'lunar/index'
+  autoload :Scoring,   'lunar/scoring'
+  autoload :Doc,       'lunar/doc'
+  autoload :Index,     'lunar/index'
+  autoload :Words,     'lunar/words'
+  autoload :Search,    'lunar/search'
+  autoload :Sets,      'lunar/sets'
+  autoload :ResultSet, 'lunar/result_set'
 
   def self.redis(connection = defined?(Ohm) ? Ohm.redis : nil)
     @connection ||= connection 
   end
+
+  def self.search(prefix, keywords, &block)
+    search = Search.new(prefix, keywords)
+    search.results(&block)
+  end
+
+  def self.encode(str)
+    Base64.encode64(str).strip 
+  end
+
+  def self.nest
+    Nest.new(:Lunar)
+  end
 end

data/lib/lunar/index.rb

@@ -11,7 +11,7 @@ module Lunar
     end
 
     def initialize(prefix, redis = nil)
-      @ns    = Nest.new(:Lunar)[prefix]
+      @ns    = Lunar.nest[prefix]
       @attrs = {}
       @redis = (redis || Lunar.redis)
     end
@@ -36,13 +36,13 @@ module Lunar
         words   = []
         scoring.scores.each do |word, score|
           words << word
-          @redis.zadd @ns[field][encode(word)], score, key
+          @redis.zadd @ns[field][Lunar.encode(word)], score, key
           @redis.sadd @ns[key][field], word
         end
 
         unused_words = @redis.smembers(@ns[key][field]) - words
         unused_words.each do |word|
-          @redis.zrem @ns[field][encode(word)], key
+          @redis.zrem @ns[field][Lunar.encode(word)], key
           @redis.srem @ns[key][field], word
         end
       end
@@ -56,15 +56,10 @@ module Lunar
         field = k.gsub("#{ @ns[key] }:", '')
         words = @redis.smembers(k)
         words.each do |w|
-          @redis.zrem @ns[field][encode(w)], key
+          @redis.zrem @ns[field][Lunar.encode(w)], key
         end
         @redis.del k 
       end
     end
-
-  private
-    def encode(str)
-      Base64.encode64(str).strip
-    end
   end
 end

data/lib/lunar/result_set.rb

@@ -0,0 +1,27 @@
+module Lunar
+  class ResultSet
+    include Enumerable
+
+    attr :sorted_set_key
+
+    def initialize(sorted_set_key, &block)
+      @sorted_set_key = sorted_set_key
+      @block          = block
+    end
+
+    def each(&block)
+      all.each(&block)
+    end
+
+    def all(options = {})
+      start  = Integer(options[:start] || 0)
+      finish = start + Integer(options[:limit] || 0) - 1
+
+      Lunar.redis.zrevrange(sorted_set_key, start, finish).map(&@block)
+    end
+
+    def size
+      Lunar.redis.zcard(sorted_set_key)
+    end
+  end
+end

data/lib/lunar/scoring.rb

@@ -1,7 +1,7 @@
 module Lunar
   class Scoring
     def initialize(words)
-      @words = words.split(/\s+/).reject { |w| w.to_s.strip.empty? }
+      @words = Words.new(words)
     end
 
     def scores

data/lib/lunar/search.rb

@@ -0,0 +1,44 @@
+module Lunar
+  class Search
+    attr :sets, :prefix, :search_identifier
+
+    def initialize(prefix, keywords)
+      if keywords.is_a?(Hash)
+        @sets = keywords.inject([]) { |a, (field, query)| a | Sets.new(prefix, query, field) }
+        @search_identifier = keywords.hash
+      else
+        @sets = Sets.new(prefix, keywords)
+        @search_identifier = keywords.hash
+      end
+
+      @prefix = prefix
+
+      # Default finder, uses Ohm style finding
+      @finder = lambda { |id| prefix[id] }
+    end
+
+    def results(&block)
+      Lunar.redis.zunion dist_key, sets.size, *sets
+      
+      if block_given?
+        ResultSet.new(dist_key, &block)
+      else
+        ResultSet.new(dist_key, &@finder)
+      end
+
+      # ids = Lunar.redis.zrevrange dist_key, start, finish
+
+      # if block_given?
+      #   ids.map { |e| block.call(e) }
+      # else
+      #   ids.map { |e| @@finder.call(prefix, e) }
+      # end
+    end
+  
+  protected
+    def dist_key
+      Lunar.nest[:Results][search_identifier]
+    end
+  end
+end
+

data/lib/lunar/sets.rb

@@ -0,0 +1,26 @@
+module Lunar
+  class Sets < Array
+    attr :prefix, :words, :field
+
+    def initialize(prefix, keywords, field = '*')
+      @prefix = prefix
+      @words  = Words.new(keywords)
+      @field  = field
+
+      super(redis_set_keys)
+    end
+  
+  protected
+    def redis_set_keys
+      keys_for_each_word.map { |key| Lunar.redis.keys(key) }.flatten
+    end
+
+    def keys_for_each_word
+      words.map { |w| ns[Lunar.encode(w)] }
+    end
+
+    def ns
+      @ns ||= Lunar.nest[prefix][field]
+    end
+  end
+end

data/lib/lunar/words.rb

@@ -0,0 +1,21 @@
+require 'iconv'
+
+module Lunar
+  class Words < Array
+    SEPARATOR = /\s+/
+
+    def initialize(str)
+      words = str.split(SEPARATOR).
+        reject { |w| w.to_s.strip.empty? }.
+        map    { |w| sanitize(w) }
+      
+      super(words)
+    end
+
+  private
+    def sanitize(str)
+      Iconv.iconv('UTF-8//IGNORE', 'UTF-8', str)[0].to_s.
+        gsub(/[^a-zA-Z0-9\-_]/, '')
+    end
+  end
+end

data/lunar.gemspec

@@ -0,0 +1,106 @@
+# Generated by jeweler
+# DO NOT EDIT THIS FILE DIRECTLY
+# Instead, edit Jeweler::Tasks in Rakefile, and run the gemspec command
+# -*- encoding: utf-8 -*-
+
+Gem::Specification.new do |s|
+  s.name = %q{lunar}
+  s.version = "0.1.1"
+
+  s.required_rubygems_version = Gem::Requirement.new(">= 0") if s.respond_to? :required_rubygems_version=
+  s.authors = ["Cyril David"]
+  s.date = %q{2010-05-02}
+  s.description = %q{uses sorted sets and sets, sorting by score}
+  s.email = %q{cyx.ucron@gmail.com}
+  s.extra_rdoc_files = [
+    "LICENSE",
+     "README.rdoc"
+  ]
+  s.files = [
+    ".document",
+     ".gitignore",
+     "LICENSE",
+     "README.rdoc",
+     "Rakefile",
+     "VERSION",
+     "examples/ohm.rb",
+     "lib/lunar.rb",
+     "lib/lunar/doc.rb",
+     "lib/lunar/index.rb",
+     "lib/lunar/result_set.rb",
+     "lib/lunar/scoring.rb",
+     "lib/lunar/search.rb",
+     "lib/lunar/sets.rb",
+     "lib/lunar/words.rb",
+     "lunar.gemspec",
+     "test/helper.rb",
+     "test/test_lunar.rb",
+     "test/test_lunar_document.rb",
+     "test/test_lunar_index.rb",
+     "test/test_lunar_scoring.rb",
+     "test/test_lunar_search.rb",
+     "test/test_sets.rb",
+     "vendor/nest/nest.rb",
+     "vendor/redis/.gitignore",
+     "vendor/redis/LICENSE",
+     "vendor/redis/README.markdown",
+     "vendor/redis/Rakefile",
+     "vendor/redis/benchmarking/logging.rb",
+     "vendor/redis/benchmarking/pipeline.rb",
+     "vendor/redis/benchmarking/speed.rb",
+     "vendor/redis/benchmarking/suite.rb",
+     "vendor/redis/benchmarking/worker.rb",
+     "vendor/redis/bin/distredis",
+     "vendor/redis/examples/basic.rb",
+     "vendor/redis/examples/dist_redis.rb",
+     "vendor/redis/examples/incr-decr.rb",
+     "vendor/redis/examples/list.rb",
+     "vendor/redis/examples/pubsub.rb",
+     "vendor/redis/examples/sets.rb",
+     "vendor/redis/lib/edis.rb",
+     "vendor/redis/lib/redis.rb",
+     "vendor/redis/lib/redis/client.rb",
+     "vendor/redis/lib/redis/dist_redis.rb",
+     "vendor/redis/lib/redis/distributed.rb",
+     "vendor/redis/lib/redis/hash_ring.rb",
+     "vendor/redis/lib/redis/pipeline.rb",
+     "vendor/redis/lib/redis/raketasks.rb",
+     "vendor/redis/lib/redis/subscribe.rb",
+     "vendor/redis/profile.rb",
+     "vendor/redis/tasks/redis.tasks.rb",
+     "vendor/redis/test/db/.gitignore",
+     "vendor/redis/test/distributed_test.rb",
+     "vendor/redis/test/redis_test.rb",
+     "vendor/redis/test/test.conf",
+     "vendor/redis/test/test_helper.rb"
+  ]
+  s.homepage = %q{http://github.com/cyx/lunar}
+  s.rdoc_options = ["--charset=UTF-8"]
+  s.require_paths = ["lib"]
+  s.rubygems_version = %q{1.3.6}
+  s.summary = %q{a minimalistic full text search implementation in redis}
+  s.test_files = [
+    "test/helper.rb",
+     "test/test_lunar.rb",
+     "test/test_lunar_document.rb",
+     "test/test_lunar_index.rb",
+     "test/test_lunar_scoring.rb",
+     "test/test_lunar_search.rb",
+     "test/test_sets.rb",
+     "examples/ohm.rb"
+  ]
+
+  if s.respond_to? :specification_version then
+    current_version = Gem::Specification::CURRENT_SPECIFICATION_VERSION
+    s.specification_version = 3
+
+    if Gem::Version.new(Gem::RubyGemsVersion) >= Gem::Version.new('1.2.0') then
+      s.add_development_dependency(%q<contest>, [">= 0"])
+    else
+      s.add_dependency(%q<contest>, [">= 0"])
+    end
+  else
+    s.add_dependency(%q<contest>, [">= 0"])
+  end
+end
+

data/test/test_lunar_index.rb

@@ -163,9 +163,10 @@ class LunarIndexTest < Test::Unit::TestCase
       assert ! Lunar.redis.exists("Lunar:Item:1001:description")      
     end
   end
+
 protected
   def encode(str)
-    Base64.encode64(str).strip    
+    Lunar.encode(str)
   end
 
   def zscore(key, value)

data/test/test_lunar_scoring.rb

@@ -23,4 +23,14 @@ class LunarScoringTest < Test::Unit::TestCase
       assert_equal 3, scoring.scores['brown']
     end     
   end
+
+  describe 'scores of apple macbook pro 17"' do
+    should "return a hash of apple macbook pro 17" do
+      scoring = Lunar::Scoring.new('apple macbook pro 17"')
+      assert_equal 1, scoring.scores['apple']
+      assert_equal 1, scoring.scores['macbook']
+      assert_equal 1, scoring.scores['pro']
+      assert_equal 1, scoring.scores['17']
+    end
+  end
 end

data/test/test_lunar_search.rb

@@ -0,0 +1,77 @@
+require "helper"
+
+class Item < Struct.new(:id)
+  def self.[](id)
+    new(id)
+  end
+end
+
+class LunarSearchTest < Test::Unit::TestCase
+  context "given an Item named iphone 3GS and apple macbook pro 17 exists" do
+    setup do
+      Lunar::Index.create "Item" do |i|
+        i.key 1001
+        i.attr :name, 'iphone 3GS'
+      end
+
+      Lunar::Index.create "Item" do |i|
+        i.key 1002
+        i.attr :name, 'apple macbook pro 17"'
+        i.attr :desc, 'iphone'
+      end
+    end
+
+    should "be able to search by keyword iphone" do
+      items = Lunar.search(Item, "iphone") { |id| Item.new(id) }
+
+      assert items.map(&:id).include?('1001')
+      assert items.map(&:id).include?('1002')
+    end
+
+    should "be able to search by keyword iphone apple and return the 2 items" do
+      items = Lunar.search(Item, "iphone apple") { |id| Item.new(id) }
+
+      assert items.map(&:id).include?('1001')
+      assert items.map(&:id).include?('1002')
+    end
+
+    should "be able to search 17 and return the macbook pro" do
+      items = Lunar.search(Item, "17") { |id| Item.new(id) }
+
+      assert items.map(&:id).include?('1002')
+    end
+
+    should 'be able to search 17 with quote and return the macbook pro' do
+      items = Lunar.search(Item, '17"') { |id| Item.new(id) }
+
+      assert items.map(&:id).include?('1002')
+    end
+
+    should "not return item 1001 when searching apple only" do
+      items = Lunar.search(Item, "apple") { |id| Item.new(id) }
+
+      assert ! items.map(&:id).include?('1001')
+    end
+
+    should "by default try and find using a bracket syntax" do
+      items = Lunar.search(Item, "apple")
+
+      assert items.map(&:id).include?('1002')
+    end
+
+    should "be able to search by name: 'iphone'" do
+      items = Lunar.search(Item, :name => 'iphone')
+
+      assert items.map(&:id).include?('1001')
+      assert ! items.map(&:id).include?('1002')
+    end
+
+    should "be able to search by name: 'iphone' desc: 'iphone'" do
+      items = Lunar.search(Item, :name => 'iphone', :desc => 'iphone')
+
+      assert items.map(&:id).include?('1001')
+      assert items.map(&:id).include?('1002')
+    end
+
+  end
+end

data/test/test_sets.rb

@@ -0,0 +1,48 @@
+require "helper"
+
+class LunarSetsTest < Test::Unit::TestCase
+  setup do
+    Lunar.redis(Redis.new(:host => "127.0.0.1", :port => "6380"))
+    Lunar.redis.flushdb
+    
+    @keys = [
+      "Lunar:Item:name:#{ Lunar.encode('iphone') }",
+      "Lunar:Item:desc:#{ Lunar.encode('iphone') }",
+      "Lunar:Item:tags:#{ Lunar.encode('asian') }",  
+      "Lunar:Item:tags:#{ Lunar.encode('hacker') }",    
+      "Lunar:Item:tags:#{ Lunar.encode('ninja') }",  
+      "Lunar:Item:body:#{ Lunar.encode('ninja') }",   
+      "Lunar:Item:body:#{ Lunar.encode('assassin') }"    
+    ]
+  
+    @keys.each { |key| Lunar.redis.zadd key, 1, 1001 }
+  end
+
+  context "given prefix Item, keywords iphone" do
+    should "return both keys" do
+      sets = Lunar::Sets.new("Item", "iphone")
+
+      assert sets.include?(@keys[0])
+      assert sets.include?(@keys[1])
+    end
+  end
+
+  context "given prefix Item, keywords iphone, field name" do
+    should "return only the name key" do
+      sets = Lunar::Sets.new("Item", "iphone", "name")
+
+      assert sets.include?(@keys[0])
+      assert ! sets.include?(@keys[1])
+    end
+  end
+
+  context "given asian ninja assassin" do
+    should "return all relevant keys" do
+      sets = Lunar::Sets.new("Item", "asian ninja assassin", "tags")
+
+      assert sets.include?(@keys[2])
+      assert sets.include?(@keys[4])
+    end
+  end
+
+end

metadata

@@ -5,8 +5,8 @@ version: !ruby/object:Gem::Version
   segments: 
   - 0
   - 1
-  - 0
-  version: 0.1.0
+  - 1
+  version: 0.1.1
 platform: ruby
 authors: 
 - Cyril David
@@ -14,7 +14,7 @@ autorequire:
 bindir: bin
 cert_chain: []
 
-date: 2010-05-01 00:00:00 +08:00
+date: 2010-05-02 00:00:00 +08:00
 default_executable: 
 dependencies: 
 - !ruby/object:Gem::Dependency 
@@ -49,12 +49,19 @@ files:
 - lib/lunar.rb
 - lib/lunar/doc.rb
 - lib/lunar/index.rb
+- lib/lunar/result_set.rb
 - lib/lunar/scoring.rb
+- lib/lunar/search.rb
+- lib/lunar/sets.rb
+- lib/lunar/words.rb
+- lunar.gemspec
 - test/helper.rb
 - test/test_lunar.rb
 - test/test_lunar_document.rb
 - test/test_lunar_index.rb
 - test/test_lunar_scoring.rb
+- test/test_lunar_search.rb
+- test/test_sets.rb
 - vendor/nest/nest.rb
 - vendor/redis/.gitignore
 - vendor/redis/LICENSE
@@ -124,4 +131,6 @@ test_files:
 - test/test_lunar_document.rb
 - test/test_lunar_index.rb
 - test/test_lunar_scoring.rb
+- test/test_lunar_search.rb
+- test/test_sets.rb
 - examples/ohm.rb