redis-search 0.1
Sign up to get free protection for your applications and to get access to all the features.
- data/README.textile +55 -0
- data/lib/redis_search.rb +11 -0
- data/lib/redis_search/base.rb +53 -0
- data/lib/redis_search/config.rb +14 -0
- data/lib/redis_search/search.rb +173 -0
- metadata +74 -0
data/README.textile
ADDED
@@ -0,0 +1,55 @@
|
|
1
|
+
h1. RedisSearch
|
2
|
+
|
3
|
+
High performance real-time search (Support Chinese), index in Redis for Rails application
|
4
|
+
|
5
|
+
h2. Features
|
6
|
+
|
7
|
+
* Real-time search
|
8
|
+
* High performance
|
9
|
+
|
10
|
+
h2. Requirements
|
11
|
+
|
12
|
+
* Redis 2.2
|
13
|
+
* Libmmseg
|
14
|
+
|
15
|
+
h2. Install
|
16
|
+
|
17
|
+
* gem install redis-search
|
18
|
+
|
19
|
+
h2. Configure
|
20
|
+
|
21
|
+
# config/initializers/redis_search.rb
|
22
|
+
require "redis_search"
|
23
|
+
redis = Redis.new(:host => redis_config[:host],:port => redis_config[:port])
|
24
|
+
redis.select("app_name.search")
|
25
|
+
RedisSearch.configure do |config|
|
26
|
+
config.redis = redis
|
27
|
+
end
|
28
|
+
|
29
|
+
h2. Usage
|
30
|
+
|
31
|
+
class Post
|
32
|
+
include Mongoid::Document
|
33
|
+
include RedisSearch
|
34
|
+
|
35
|
+
field :title
|
36
|
+
field :body
|
37
|
+
|
38
|
+
belongs_to :user
|
39
|
+
belongs_to :category
|
40
|
+
|
41
|
+
# bind RedisSearch callback event, it will to rebuild search indexes when data create or update.
|
42
|
+
redis_search_index(:title_field => :title,
|
43
|
+
:ext_fields => [:category_name])
|
44
|
+
|
45
|
+
def category_name
|
46
|
+
self.category.name
|
47
|
+
end
|
48
|
+
end
|
49
|
+
|
50
|
+
# GET /searchs?q=title
|
51
|
+
class SearchController < ApplicationController
|
52
|
+
def index
|
53
|
+
RedisSearch::Search.query(params[:q], :type => "Post")
|
54
|
+
end
|
55
|
+
end
|
data/lib/redis_search.rb
ADDED
@@ -0,0 +1,53 @@
|
|
1
|
+
module RedisSearch
|
2
|
+
extend ActiveSupport::Concern
|
3
|
+
|
4
|
+
module ClassMethods
|
5
|
+
def redis_search_index(options = {})
|
6
|
+
title_field = options[:title_field] || :title
|
7
|
+
ext_fields = options[:ext_fields] || []
|
8
|
+
class_eval %(
|
9
|
+
def redis_search_ext_fields(ext_fields)
|
10
|
+
exts = {}
|
11
|
+
ext_fields.each do |f|
|
12
|
+
exts[f] = instance_eval(f.to_s)
|
13
|
+
end
|
14
|
+
exts
|
15
|
+
end
|
16
|
+
|
17
|
+
after_create :redis_search_index_create
|
18
|
+
def redis_search_index_create
|
19
|
+
s = Search.new(:title => self.#{title_field}, :id => self.id,
|
20
|
+
:exts => self.redis_search_ext_fields(#{ext_fields}),
|
21
|
+
:type => self.class.to_s)
|
22
|
+
s.save
|
23
|
+
end
|
24
|
+
|
25
|
+
before_destroy :redis_search_index_remove
|
26
|
+
def redis_search_index_remove
|
27
|
+
Search.remove(:id => self.id, :title => self.#{title_field}, :type => self.class.to_s)
|
28
|
+
end
|
29
|
+
|
30
|
+
before_update :redis_search_index_update
|
31
|
+
def redis_search_index_update
|
32
|
+
index_fields_changed = false
|
33
|
+
#{ext_fields}.each do |f|
|
34
|
+
next if f.to_s == "id"
|
35
|
+
if instance_eval(f.to_s + "_changed?")
|
36
|
+
index_fields_changed = true
|
37
|
+
end
|
38
|
+
end
|
39
|
+
begin
|
40
|
+
if(self.#{title_field}_changed?)
|
41
|
+
index_fields_changed = true
|
42
|
+
end
|
43
|
+
rescue
|
44
|
+
end
|
45
|
+
if index_fields_changed
|
46
|
+
Search.remove(:id => self.id, :title => self.#{title_field}_was, :type => self.class.to_s)
|
47
|
+
self.redis_search_index_create
|
48
|
+
end
|
49
|
+
end
|
50
|
+
)
|
51
|
+
end
|
52
|
+
end
|
53
|
+
end
|
@@ -0,0 +1,173 @@
|
|
1
|
+
# coding: utf-8
|
2
|
+
require "mmseg"
|
3
|
+
module RedisSearch
|
4
|
+
class Search
|
5
|
+
attr_accessor :type, :title, :id, :exts
|
6
|
+
def initialize(options = {})
|
7
|
+
self.exts = []
|
8
|
+
options.keys.each do |k|
|
9
|
+
eval("self.#{k} = options[k]")
|
10
|
+
end
|
11
|
+
end
|
12
|
+
|
13
|
+
def self.split(text)
|
14
|
+
algor = RMMSeg::Algorithm.new(text)
|
15
|
+
words = []
|
16
|
+
loop do
|
17
|
+
tok = algor.next_token
|
18
|
+
break if tok.nil?
|
19
|
+
words << tok.text
|
20
|
+
end
|
21
|
+
words
|
22
|
+
end
|
23
|
+
|
24
|
+
def self.warn(msg)
|
25
|
+
puts "[RedisSearch][warn]: #{msg}"
|
26
|
+
end
|
27
|
+
|
28
|
+
# 生成 uuid,用于作为 hashes 的 field, sets 关键词的值
|
29
|
+
def self.mk_sets_key(type, key)
|
30
|
+
"#{type}:#{key.downcase}"
|
31
|
+
end
|
32
|
+
|
33
|
+
def self.mk_complete_key(type)
|
34
|
+
"Compl#{type}"
|
35
|
+
end
|
36
|
+
|
37
|
+
def self.word?(word)
|
38
|
+
return !/^[\w\u4e00-\u9fa5]+$/i.match(word.force_encoding("UTF-8")).blank?
|
39
|
+
end
|
40
|
+
|
41
|
+
def save
|
42
|
+
return if self.title.blank?
|
43
|
+
data = {:title => self.title, :id => self.id, :type => self.type}
|
44
|
+
self.exts.each do |f|
|
45
|
+
data[f[0]] = f[1]
|
46
|
+
end
|
47
|
+
|
48
|
+
# 将原始数据存入 hashes
|
49
|
+
res = RedisSearch.config.redis.hset(self.type, self.id, data.to_json)
|
50
|
+
# 保存 sets 索引,以分词的单词为key,用于后面搜索,里面存储 ids
|
51
|
+
words = Search.split(self.title)
|
52
|
+
return if words.blank?
|
53
|
+
words.each do |word|
|
54
|
+
next if not Search.word?(word)
|
55
|
+
save_zindex(word)
|
56
|
+
key = Search.mk_sets_key(self.type,word)
|
57
|
+
RedisSearch.config.redis.sadd(key, self.id)
|
58
|
+
end
|
59
|
+
end
|
60
|
+
|
61
|
+
def save_zindex(word)
|
62
|
+
return if not Search.word?(word)
|
63
|
+
word = word.downcase
|
64
|
+
key = Search.mk_complete_key(self.type)
|
65
|
+
(1..(word.length)).each do |l|
|
66
|
+
prefix = word[0...l]
|
67
|
+
RedisSearch.config.redis.zadd(key, 0, prefix)
|
68
|
+
end
|
69
|
+
RedisSearch.config.redis.zadd(key, 0, word + "*")
|
70
|
+
end
|
71
|
+
|
72
|
+
def self.remove(options = {})
|
73
|
+
puts options.inspect
|
74
|
+
type = options[:type]
|
75
|
+
RedisSearch.config.redis.hdel(type,options[:id])
|
76
|
+
words = Search.split(options[:title])
|
77
|
+
words.each do |word|
|
78
|
+
next if not Search.word?(word)
|
79
|
+
key = Search.mk_sets_key(type,word)
|
80
|
+
RedisSearch.config.redis.srem(key, options[:id])
|
81
|
+
end
|
82
|
+
end
|
83
|
+
|
84
|
+
# Use for short title search, this method is search by chars, for example Tag, User, Category ...
|
85
|
+
#
|
86
|
+
# h3. params:
|
87
|
+
# type model name
|
88
|
+
# w search char
|
89
|
+
# :limit result limit
|
90
|
+
# h3. usage:
|
91
|
+
# * RedisSearch::Search.complete("Tag","r") => ["Ruby","Rails", "REST", "Redis", "Redmine"]
|
92
|
+
# * RedisSearch::Search.complete("Tag","re") => ["Redis", "Redmine"]
|
93
|
+
# * RedisSearch::Search.complete("Tag","red") => ["Redis", "Redmine"]
|
94
|
+
# * RedisSearch::Search.complete("Tag","redi") => ["Redis"]
|
95
|
+
def self.complete(type, w, options = {})
|
96
|
+
limit = options[:limit] || 10
|
97
|
+
|
98
|
+
prefix_matchs = []
|
99
|
+
rangelen = 100 # This is not random, try to get replies < MTU size
|
100
|
+
prefix = w.downcase
|
101
|
+
key = Search.mk_complete_key(type)
|
102
|
+
start = RedisSearch.config.redis.zrank(key,prefix)
|
103
|
+
|
104
|
+
return [] if !start
|
105
|
+
count = limit
|
106
|
+
while prefix_matchs.length <= count
|
107
|
+
range = RedisSearch.config.redis.zrange(key,start,start+rangelen-1)
|
108
|
+
start += rangelen
|
109
|
+
break if !range or range.length == 0
|
110
|
+
range.each {|entry|
|
111
|
+
minlen = [entry.length,prefix.length].min
|
112
|
+
if entry[0...minlen] != prefix[0...minlen]
|
113
|
+
count = prefix_matchs.count
|
114
|
+
break
|
115
|
+
end
|
116
|
+
if entry[-1..-1] == "*" and prefix_matchs.length != count
|
117
|
+
prefix_matchs << entry[0...-1]
|
118
|
+
end
|
119
|
+
}
|
120
|
+
end
|
121
|
+
words = []
|
122
|
+
words = prefix_matchs.uniq.collect { |w| Search.mk_sets_key(type,w) }
|
123
|
+
ids = RedisSearch.config.redis.sunion(*words)
|
124
|
+
return [] if ids.blank?
|
125
|
+
hmget(type,ids, :limit => limit)
|
126
|
+
end
|
127
|
+
|
128
|
+
# Search items, this will split words by Libmmseg
|
129
|
+
#
|
130
|
+
# h3. params:
|
131
|
+
# type model name
|
132
|
+
# text search text
|
133
|
+
# :limit result limit
|
134
|
+
# h3. usage:
|
135
|
+
# * RedisSearch::Search.query("Tag","Ruby vs Python")
|
136
|
+
def self.query(type, text,options = {})
|
137
|
+
result = []
|
138
|
+
return result if text.strip.blank?
|
139
|
+
|
140
|
+
words = Search.split(text)
|
141
|
+
limit = options[:limit] || 10
|
142
|
+
sort_field = options[:sort_field] || "id"
|
143
|
+
words = words.collect { |w| Search.mk_sets_key(type,w) }
|
144
|
+
return result if words.blank?
|
145
|
+
ids = RedisSearch.config.redis.sinter(*words)
|
146
|
+
hmget(type,ids, :limit => limit, :sort_field => sort_field)
|
147
|
+
end
|
148
|
+
|
149
|
+
private
|
150
|
+
def self.hmget(type, ids, options = {})
|
151
|
+
result = []
|
152
|
+
limit = options[:limit] || 10
|
153
|
+
sort_field = options[:sort_field] || "id"
|
154
|
+
return result if ids.blank?
|
155
|
+
RedisSearch.config.redis.hmget(type,*ids).each do |r|
|
156
|
+
begin
|
157
|
+
result << JSON.parse(r)
|
158
|
+
rescue => e
|
159
|
+
Search.warn("Search.query failed: #{e}")
|
160
|
+
end
|
161
|
+
end
|
162
|
+
items = sort_result(result, type, sort_field)
|
163
|
+
items = items[0..limit-1] if items.length > limit
|
164
|
+
items
|
165
|
+
end
|
166
|
+
|
167
|
+
def self.sort_result(items, type, sort_field)
|
168
|
+
return items if items.blank?
|
169
|
+
items = items.sort { |x,y| y[sort_field] <=> x[sort_field] }
|
170
|
+
items
|
171
|
+
end
|
172
|
+
end
|
173
|
+
end
|
metadata
ADDED
@@ -0,0 +1,74 @@
|
|
1
|
+
--- !ruby/object:Gem::Specification
|
2
|
+
name: redis-search
|
3
|
+
version: !ruby/object:Gem::Version
|
4
|
+
version: '0.1'
|
5
|
+
prerelease:
|
6
|
+
platform: ruby
|
7
|
+
authors:
|
8
|
+
- Jason Lee
|
9
|
+
autorequire:
|
10
|
+
bindir: bin
|
11
|
+
cert_chain: []
|
12
|
+
date: 2011-08-16 00:00:00.000000000Z
|
13
|
+
dependencies:
|
14
|
+
- !ruby/object:Gem::Dependency
|
15
|
+
name: rmmseg-cpp-huacnlee
|
16
|
+
requirement: &2159834180 !ruby/object:Gem::Requirement
|
17
|
+
none: false
|
18
|
+
requirements:
|
19
|
+
- - ~>
|
20
|
+
- !ruby/object:Gem::Version
|
21
|
+
version: 0.2.8
|
22
|
+
type: :runtime
|
23
|
+
prerelease: false
|
24
|
+
version_requirements: *2159834180
|
25
|
+
- !ruby/object:Gem::Dependency
|
26
|
+
name: redis
|
27
|
+
requirement: &2159833700 !ruby/object:Gem::Requirement
|
28
|
+
none: false
|
29
|
+
requirements:
|
30
|
+
- - ~>
|
31
|
+
- !ruby/object:Gem::Version
|
32
|
+
version: 2.1.1
|
33
|
+
type: :runtime
|
34
|
+
prerelease: false
|
35
|
+
version_requirements: *2159833700
|
36
|
+
description: High performance real-time search (Support Chinese), index in Redis for
|
37
|
+
Rails application.
|
38
|
+
email:
|
39
|
+
- huacnlee@gmail.com
|
40
|
+
executables: []
|
41
|
+
extensions: []
|
42
|
+
extra_rdoc_files: []
|
43
|
+
files:
|
44
|
+
- lib/redis_search/base.rb
|
45
|
+
- lib/redis_search/config.rb
|
46
|
+
- lib/redis_search/search.rb
|
47
|
+
- lib/redis_search.rb
|
48
|
+
- README.textile
|
49
|
+
homepage: http://github.com/huacnlee/redis-search
|
50
|
+
licenses: []
|
51
|
+
post_install_message:
|
52
|
+
rdoc_options: []
|
53
|
+
require_paths:
|
54
|
+
- lib
|
55
|
+
required_ruby_version: !ruby/object:Gem::Requirement
|
56
|
+
none: false
|
57
|
+
requirements:
|
58
|
+
- - ! '>='
|
59
|
+
- !ruby/object:Gem::Version
|
60
|
+
version: '0'
|
61
|
+
required_rubygems_version: !ruby/object:Gem::Requirement
|
62
|
+
none: false
|
63
|
+
requirements:
|
64
|
+
- - ! '>='
|
65
|
+
- !ruby/object:Gem::Version
|
66
|
+
version: 1.3.6
|
67
|
+
requirements: []
|
68
|
+
rubyforge_project:
|
69
|
+
rubygems_version: 1.8.6
|
70
|
+
signing_key:
|
71
|
+
specification_version: 3
|
72
|
+
summary: High performance real-time search (Support Chinese), index in Redis for Rails
|
73
|
+
application.
|
74
|
+
test_files: []
|