kristopher-sunspot_rails 0.10.6
Sign up to get free protection for your applications and to get access to all the features.
- data/LICENSE +18 -0
- data/MIT-LICENSE +20 -0
- data/README.rdoc +233 -0
- data/Rakefile +19 -0
- data/VERSION.yml +4 -0
- data/dev_tasks/gemspec.rake +29 -0
- data/dev_tasks/rdoc.rake +7 -0
- data/dev_tasks/release.rake +4 -0
- data/dev_tasks/todo.rake +4 -0
- data/generators/sunspot/sunspot_generator.rb +9 -0
- data/generators/sunspot/templates/sunspot.yml +19 -0
- data/install.rb +1 -0
- data/lib/sunspot/rails.rb +13 -0
- data/lib/sunspot/rails/adapters.rb +54 -0
- data/lib/sunspot/rails/configuration.rb +169 -0
- data/lib/sunspot/rails/request_lifecycle.rb +18 -0
- data/lib/sunspot/rails/searchable.rb +305 -0
- data/lib/sunspot/rails/tasks.rb +47 -0
- data/rails/init.rb +10 -0
- data/spec/configuration_spec.rb +53 -0
- data/spec/mock_app/app/controllers/application.rb +10 -0
- data/spec/mock_app/app/controllers/application_controller.rb +10 -0
- data/spec/mock_app/app/controllers/posts_controller.rb +6 -0
- data/spec/mock_app/app/models/blog.rb +2 -0
- data/spec/mock_app/app/models/post.rb +5 -0
- data/spec/mock_app/app/models/post_with_auto.rb +9 -0
- data/spec/mock_app/config/boot.rb +110 -0
- data/spec/mock_app/config/database.yml +4 -0
- data/spec/mock_app/config/environment.rb +41 -0
- data/spec/mock_app/config/environments/development.rb +27 -0
- data/spec/mock_app/config/environments/test.rb +27 -0
- data/spec/mock_app/config/initializers/new_rails_defaults.rb +19 -0
- data/spec/mock_app/config/initializers/session_store.rb +15 -0
- data/spec/mock_app/config/routes.rb +43 -0
- data/spec/mock_app/config/sunspot.yml +14 -0
- data/spec/model_lifecycle_spec.rb +38 -0
- data/spec/model_spec.rb +278 -0
- data/spec/request_lifecycle_spec.rb +30 -0
- data/spec/schema.rb +14 -0
- data/spec/spec_helper.rb +32 -0
- metadata +184 -0
@@ -0,0 +1,169 @@
|
|
1
|
+
module Sunspot #:nodoc:
|
2
|
+
module Rails #:nodoc:
|
3
|
+
#
|
4
|
+
# Sunspot::Rails is configured via the config/sunspot.yml file, which
|
5
|
+
# contains properties keyed by environment name. A sample sunspot.yml file
|
6
|
+
# would look like:
|
7
|
+
#
|
8
|
+
# development:
|
9
|
+
# solr:
|
10
|
+
# hostname: localhost
|
11
|
+
# port: 8982
|
12
|
+
# test:
|
13
|
+
# solr:
|
14
|
+
# hostname: localhost
|
15
|
+
# port: 8983
|
16
|
+
#
|
17
|
+
# production:
|
18
|
+
# solr:
|
19
|
+
# hostname: localhost
|
20
|
+
# port: 8983
|
21
|
+
# path: /solr/myindex
|
22
|
+
#
|
23
|
+
# Sunspot::Rails uses the configuration to set up the Solr connection, as
|
24
|
+
# well as for starting Solr with the appropriate port using the
|
25
|
+
# <code>rake sunspot:solr:start</code> task.
|
26
|
+
#
|
27
|
+
class Configuration
|
28
|
+
attr_writer :user_configuration
|
29
|
+
#
|
30
|
+
# The host name at which to connect to Solr. Default 'localhost'.
|
31
|
+
#
|
32
|
+
# ==== Returns
|
33
|
+
#
|
34
|
+
# String:: host name
|
35
|
+
#
|
36
|
+
def hostname
|
37
|
+
@hostname ||= (user_configuration_from_key('solr', 'hostname') || 'localhost')
|
38
|
+
end
|
39
|
+
|
40
|
+
#
|
41
|
+
# The port at which to connect to Solr. Default 8983.
|
42
|
+
#
|
43
|
+
# ==== Returns
|
44
|
+
#
|
45
|
+
# Integer:: port
|
46
|
+
#
|
47
|
+
def port
|
48
|
+
@port ||= (user_configuration_from_key('solr', 'port') || 8983).to_i
|
49
|
+
end
|
50
|
+
|
51
|
+
#
|
52
|
+
# The url path to the Solr servlet (useful if you are running multicore).
|
53
|
+
# Default '/solr'.
|
54
|
+
#
|
55
|
+
# ==== Returns
|
56
|
+
#
|
57
|
+
# String:: path
|
58
|
+
#
|
59
|
+
def path
|
60
|
+
@path ||= (user_configuration_from_key('solr', 'path') || '/solr')
|
61
|
+
end
|
62
|
+
|
63
|
+
#
|
64
|
+
# Should the solr index receive a commit after each http-request.
|
65
|
+
# Default true
|
66
|
+
#
|
67
|
+
# ==== Returns
|
68
|
+
#
|
69
|
+
# Boolean:: bool
|
70
|
+
#
|
71
|
+
|
72
|
+
def auto_commit_after_request?
|
73
|
+
@auto_commit_after_request ||=
|
74
|
+
user_configuration_from_key('auto_commit_after_request') != false
|
75
|
+
end
|
76
|
+
|
77
|
+
#
|
78
|
+
# The path to the Solr indexes. (Used by the rake tasks).
|
79
|
+
# Default RAILS_ROOT + '/solr/data/' + ENVIRONMENT
|
80
|
+
#
|
81
|
+
# ==== Returns
|
82
|
+
#
|
83
|
+
# String:: path
|
84
|
+
#
|
85
|
+
def data_path
|
86
|
+
@data_path ||=
|
87
|
+
if user_configuration.has_key?('solr')
|
88
|
+
"#{user_configuration['solr']['data_path'] || File.join(::Rails.root, 'solr', 'data', ::Rails.env)}"
|
89
|
+
end
|
90
|
+
end
|
91
|
+
|
92
|
+
#
|
93
|
+
# The path to the Solr pids
|
94
|
+
# Default RAILS_ROOT + '/solr/pids/' + ENVIRONMENT
|
95
|
+
#
|
96
|
+
# ==== Returns
|
97
|
+
#
|
98
|
+
# String:: path
|
99
|
+
#
|
100
|
+
def pid_path
|
101
|
+
@pids_path ||=
|
102
|
+
if user_configuration.has_key?('solr')
|
103
|
+
"#{user_configuration['solr']['pid_path'] || File.join(::Rails.root, 'solr', 'pids', ::Rails.env)}"
|
104
|
+
end
|
105
|
+
end
|
106
|
+
|
107
|
+
#
|
108
|
+
# The path to the Solr home directory
|
109
|
+
# Default nil (runs the solr with sunspot default settings).
|
110
|
+
#
|
111
|
+
# If you have a custom solr conf directory,
|
112
|
+
# change this to the directory above your solr conf files
|
113
|
+
#
|
114
|
+
# e.g. conf files in RAILS_ROOT/solr/conf
|
115
|
+
# solr_home: RAILS_ROOT/solr
|
116
|
+
#
|
117
|
+
# ==== Returns
|
118
|
+
#
|
119
|
+
# String:: path
|
120
|
+
#
|
121
|
+
def solr_home
|
122
|
+
@solr_home ||=
|
123
|
+
if user_configuration.has_key?('solr')
|
124
|
+
if user_configuration['solr']['solr_home'].present?
|
125
|
+
user_configuration['solr']['solr_home']
|
126
|
+
elsif %w(solrconfig schema).all? { |file| File.exist?(File.join(::Rails.root, 'solr', 'conf', "#{file}.xml")) }
|
127
|
+
File.join(::Rails.root, 'solr')
|
128
|
+
end
|
129
|
+
end
|
130
|
+
end
|
131
|
+
|
132
|
+
private
|
133
|
+
|
134
|
+
#
|
135
|
+
# return a specific key from the user configuration in config/sunspot.yml
|
136
|
+
#
|
137
|
+
# ==== Returns
|
138
|
+
#
|
139
|
+
#
|
140
|
+
def user_configuration_from_key( *keys )
|
141
|
+
keys.inject(user_configuration) do |hash, key|
|
142
|
+
hash[key] if hash
|
143
|
+
end
|
144
|
+
end
|
145
|
+
|
146
|
+
#
|
147
|
+
# Memoized hash of configuration options for the current Rails environment
|
148
|
+
# as specified in config/sunspot.yml
|
149
|
+
#
|
150
|
+
# ==== Returns
|
151
|
+
#
|
152
|
+
# Hash:: configuration options for current environment
|
153
|
+
#
|
154
|
+
def user_configuration
|
155
|
+
@user_configuration ||=
|
156
|
+
begin
|
157
|
+
path = File.join(::Rails.root, 'config', 'sunspot.yml')
|
158
|
+
if File.exist?(path)
|
159
|
+
File.open(path) do |file|
|
160
|
+
YAML.load(file)[::Rails.env]
|
161
|
+
end
|
162
|
+
else
|
163
|
+
{}
|
164
|
+
end
|
165
|
+
end
|
166
|
+
end
|
167
|
+
end
|
168
|
+
end
|
169
|
+
end
|
@@ -0,0 +1,18 @@
|
|
1
|
+
module Sunspot #:nodoc:
|
2
|
+
module Rails #:nodoc:
|
3
|
+
#
|
4
|
+
# This module adds an after_filter to ActionController::Base that commits
|
5
|
+
# the Sunspot session if any documents have been added, changed, or removed
|
6
|
+
# in the course of the request.
|
7
|
+
#
|
8
|
+
module RequestLifecycle
|
9
|
+
class <<self
|
10
|
+
def included(base) #:nodoc:
|
11
|
+
base.after_filter do
|
12
|
+
Sunspot.commit_if_dirty if Sunspot::Rails.configuration.auto_commit_after_request?
|
13
|
+
end
|
14
|
+
end
|
15
|
+
end
|
16
|
+
end
|
17
|
+
end
|
18
|
+
end
|
@@ -0,0 +1,305 @@
|
|
1
|
+
module Sunspot #:nodoc:
|
2
|
+
module Rails #:nodoc:
|
3
|
+
#
|
4
|
+
# This module adds Sunspot functionality to ActiveRecord models. As well as
|
5
|
+
# providing class and instance methods, it optionally adds lifecycle hooks
|
6
|
+
# to automatically add and remove models from the Solr index as they are
|
7
|
+
# created and destroyed.
|
8
|
+
#
|
9
|
+
module Searchable
|
10
|
+
class <<self
|
11
|
+
def included(base) #:nodoc:
|
12
|
+
base.module_eval { extend(ActsAsMethods) }
|
13
|
+
end
|
14
|
+
end
|
15
|
+
|
16
|
+
module ActsAsMethods
|
17
|
+
#
|
18
|
+
# Makes a class searchable if it is not already, or adds search
|
19
|
+
# configuration if it is. Note that the options passed in are only used
|
20
|
+
# the first time this method is called for a particular class; so,
|
21
|
+
# search should be defined before activating any mixins that extend
|
22
|
+
# search configuration.
|
23
|
+
#
|
24
|
+
# The block passed into this method is evaluated by the
|
25
|
+
# <code>Sunspot.setup</code> method. See the Sunspot documentation for
|
26
|
+
# complete information on the functionality provided by that method.
|
27
|
+
#
|
28
|
+
# ==== Options (+options+)
|
29
|
+
#
|
30
|
+
# :auto_index<Boolean>::
|
31
|
+
# Automatically index models in Solr when they are saved.
|
32
|
+
# Default: true
|
33
|
+
# :auto_remove<Boolean>::
|
34
|
+
# Automatically remove models from the Solr index when they are
|
35
|
+
# destroyed. <b>Setting this option to +false+ is not recommended
|
36
|
+
# </b>(see the README).
|
37
|
+
#
|
38
|
+
# ==== Example
|
39
|
+
#
|
40
|
+
# class Post < ActiveRecord::Base
|
41
|
+
# searchable do
|
42
|
+
# text :title, :body
|
43
|
+
# string :sort_title do
|
44
|
+
# title.downcase.sub(/^(an?|the)/, '')
|
45
|
+
# end
|
46
|
+
# integer :blog_id
|
47
|
+
# time :updated_at
|
48
|
+
# end
|
49
|
+
# end
|
50
|
+
#
|
51
|
+
def searchable(options = {}, &block)
|
52
|
+
Sunspot.setup(self, &block)
|
53
|
+
|
54
|
+
unless searchable?
|
55
|
+
extend ClassMethods
|
56
|
+
include InstanceMethods
|
57
|
+
|
58
|
+
unless options[:auto_index] == false
|
59
|
+
after_save do |searchable|
|
60
|
+
searchable.index
|
61
|
+
end
|
62
|
+
end
|
63
|
+
|
64
|
+
unless options[:auto_remove] == false
|
65
|
+
after_destroy do |searchable|
|
66
|
+
searchable.remove_from_index
|
67
|
+
end
|
68
|
+
end
|
69
|
+
end
|
70
|
+
end
|
71
|
+
|
72
|
+
#
|
73
|
+
# This method is defined on all ActiveRecord::Base subclasses. It
|
74
|
+
# is false for classes on which #searchable has not been called, and
|
75
|
+
# true for classes on which #searchable has been called.
|
76
|
+
#
|
77
|
+
# ==== Returns
|
78
|
+
#
|
79
|
+
# +false+
|
80
|
+
#
|
81
|
+
def searchable?
|
82
|
+
false
|
83
|
+
end
|
84
|
+
end
|
85
|
+
|
86
|
+
module ClassMethods
|
87
|
+
#
|
88
|
+
# Search for instances of this class in Solr. The block is delegated to
|
89
|
+
# the Sunspot.search method - see the Sunspot documentation for the full
|
90
|
+
# API.
|
91
|
+
#
|
92
|
+
# ==== Example
|
93
|
+
#
|
94
|
+
# Post.search do
|
95
|
+
# keywords 'best pizza'
|
96
|
+
# with :blog_id, 1
|
97
|
+
# order :updated_at, :desc
|
98
|
+
# facet :category_ids
|
99
|
+
# end
|
100
|
+
#
|
101
|
+
#
|
102
|
+
# ==== Returns
|
103
|
+
#
|
104
|
+
# Sunspot::Search:: Object containing results, totals, facets, etc.
|
105
|
+
#
|
106
|
+
def search(&block)
|
107
|
+
Sunspot.search(self, &block)
|
108
|
+
end
|
109
|
+
|
110
|
+
#
|
111
|
+
# Get IDs of matching results without loading the result objects from
|
112
|
+
# the database. This method may be useful if search is used as an
|
113
|
+
# intermediate step in a larger find operation. The block is the same
|
114
|
+
# as the block provided to the #search method.
|
115
|
+
#
|
116
|
+
# ==== Returns
|
117
|
+
#
|
118
|
+
# Array:: Array of IDs, in the order returned by the search
|
119
|
+
#
|
120
|
+
def search_ids(&block)
|
121
|
+
search(&block).raw_results.map { |raw_result| raw_result.primary_key.to_i }
|
122
|
+
end
|
123
|
+
|
124
|
+
#
|
125
|
+
# Remove instances of this class from the Solr index.
|
126
|
+
#
|
127
|
+
def remove_all_from_index
|
128
|
+
Sunspot.remove_all(self)
|
129
|
+
end
|
130
|
+
|
131
|
+
#
|
132
|
+
# Remove all instances of this class from the Solr index and immediately
|
133
|
+
# commit.
|
134
|
+
#
|
135
|
+
#---
|
136
|
+
# XXX Sunspot should implement remove_all!()
|
137
|
+
#
|
138
|
+
def remove_all_from_index!
|
139
|
+
Sunspot.remove_all(self)
|
140
|
+
Sunspot.commit
|
141
|
+
end
|
142
|
+
|
143
|
+
#
|
144
|
+
# Completely rebuild the index for this class. First removes all
|
145
|
+
# instances from the index, then loads records and save them. The
|
146
|
+
# +batch_size+ argument specifies how many records to load out of the
|
147
|
+
# database at a time. The default batch size is 500; if nil is passed,
|
148
|
+
# records will not be indexed in batches. By default, a commit is issued
|
149
|
+
# after each batch; passing +false+ for +batch_commit+ will disable
|
150
|
+
# this, and only issue a commit at the end of the process. If associated
|
151
|
+
# objects need to indexed also, you can specify +include+ in format
|
152
|
+
# accepted by ActiveRecord to improve your sql select performance
|
153
|
+
#
|
154
|
+
# ==== Options (passed as a hash)
|
155
|
+
#
|
156
|
+
# batch_size<Integer>:: Batch size with which to load records. Passing
|
157
|
+
# 'nil' will skip batches. Default is 500.
|
158
|
+
# batch_commit<Boolean>:: Flag signalling if a commit should be done after
|
159
|
+
# after each batch is indexed, default is 'true'
|
160
|
+
# include<Mixed>:: include option to be passed to the ActiveRecord find,
|
161
|
+
# used for including associated objects that need to be
|
162
|
+
# indexed with the parent object, accepts all formats
|
163
|
+
# ActiveRecord::Base.find does
|
164
|
+
#
|
165
|
+
# ==== Examples
|
166
|
+
#
|
167
|
+
# # reindex in batches of 500, commit after each
|
168
|
+
# Post.reindex
|
169
|
+
#
|
170
|
+
# # index all rows at once, then commit
|
171
|
+
# Post.reindex(:batch_size => nil)
|
172
|
+
#
|
173
|
+
# # reindex in batches of 500, commit when all batches complete
|
174
|
+
# Post.reindex(:batch_commit => false)
|
175
|
+
#
|
176
|
+
# # include the associated +author+ object when loading to index
|
177
|
+
# Post.reindex(:include => :author)
|
178
|
+
#
|
179
|
+
def reindex(opts={})
|
180
|
+
options = { :batch_size => 500, :batch_commit => true, :include => []}.merge(opts)
|
181
|
+
remove_all_from_index
|
182
|
+
unless options[:batch_size]
|
183
|
+
Sunspot.index!(all(:include => options[:include]))
|
184
|
+
else
|
185
|
+
record_count = count
|
186
|
+
counter = 1
|
187
|
+
offset = 0
|
188
|
+
while(offset < record_count)
|
189
|
+
benchmark options[:batch_size], counter do
|
190
|
+
Sunspot.index(all(:include => options[:include], :offset => offset, :limit => options[:batch_size], :order => primary_key))
|
191
|
+
end
|
192
|
+
Sunspot.commit if options[:batch_commit]
|
193
|
+
offset += options[:batch_size]
|
194
|
+
counter += 1
|
195
|
+
end
|
196
|
+
Sunspot.commit unless options[:batch_commit]
|
197
|
+
end
|
198
|
+
end
|
199
|
+
|
200
|
+
#
|
201
|
+
# Return the IDs of records of this class that are indexed in Solr but
|
202
|
+
# do not exist in the database. Under normal circumstances, this should
|
203
|
+
# never happen, but this method is provided in case something goes
|
204
|
+
# wrong. Usually you will want to rectify the situation by calling
|
205
|
+
# #clean_index_orphans or #reindex
|
206
|
+
#
|
207
|
+
# ==== Returns
|
208
|
+
#
|
209
|
+
# Array:: Collection of IDs that exist in Solr but not in the database
|
210
|
+
def index_orphans
|
211
|
+
count = self.count
|
212
|
+
indexed_ids = search_ids { paginate(:page => 1, :per_page => count) }.to_set
|
213
|
+
all(:select => 'id').each do |object|
|
214
|
+
indexed_ids.delete(object.id)
|
215
|
+
end
|
216
|
+
indexed_ids.to_a
|
217
|
+
end
|
218
|
+
|
219
|
+
#
|
220
|
+
# Find IDs of records of this class that are indexed in Solr but do not
|
221
|
+
# exist in the database, and remove them from Solr. Under normal
|
222
|
+
# circumstances, this should not be necessary; this method is provided
|
223
|
+
# in case something goes wrong.
|
224
|
+
#
|
225
|
+
def clean_index_orphans
|
226
|
+
index_orphans.each do |id|
|
227
|
+
new do |fake_instance|
|
228
|
+
fake_instance.id = id
|
229
|
+
end.remove_from_index
|
230
|
+
end
|
231
|
+
end
|
232
|
+
|
233
|
+
#
|
234
|
+
# Classes that have been defined as searchable return +true+ for this
|
235
|
+
# method.
|
236
|
+
#
|
237
|
+
# ==== Returns
|
238
|
+
#
|
239
|
+
# +true+
|
240
|
+
#
|
241
|
+
def searchable?
|
242
|
+
true
|
243
|
+
end
|
244
|
+
|
245
|
+
protected
|
246
|
+
|
247
|
+
#
|
248
|
+
# Does some logging for benchmarking indexing performance
|
249
|
+
#
|
250
|
+
def benchmark(batch_size, counter, &block)
|
251
|
+
start = Time.now
|
252
|
+
logger.info("[#{Time.now}] Start Indexing")
|
253
|
+
yield
|
254
|
+
elapsed = Time.now-start
|
255
|
+
logger.info("[#{Time.now}] Completed Indexing. Rows indexed #{counter * batch_size}. Rows/sec: #{batch_size/elapsed.to_f} (Elapsed: #{elapsed} sec.)")
|
256
|
+
end
|
257
|
+
|
258
|
+
end
|
259
|
+
|
260
|
+
module InstanceMethods
|
261
|
+
#
|
262
|
+
# Index the model in Solr. If the model is already indexed, it will be
|
263
|
+
# updated. Using the defaults, you will usually not need to call this
|
264
|
+
# method, as models are indexed automatically when they are created or
|
265
|
+
# updated. If you have disabled automatic indexing (see
|
266
|
+
# ClassMethods#searchable), this method allows you to manage indexing
|
267
|
+
# manually.
|
268
|
+
#
|
269
|
+
def index
|
270
|
+
Sunspot.index(self)
|
271
|
+
end
|
272
|
+
|
273
|
+
#
|
274
|
+
# Index the model in Solr and immediately commit. See #index
|
275
|
+
#
|
276
|
+
def index!
|
277
|
+
Sunspot.index!(self)
|
278
|
+
end
|
279
|
+
|
280
|
+
#
|
281
|
+
# Remove the model from the Solr index. Using the defaults, this should
|
282
|
+
# not be necessary, as models will automatically be removed from the
|
283
|
+
# index when they are destroyed. If you disable automatic removal
|
284
|
+
# (which is not recommended!), you can use this method to manage removal
|
285
|
+
# manually.
|
286
|
+
#
|
287
|
+
def remove_from_index
|
288
|
+
Sunspot.remove(self)
|
289
|
+
end
|
290
|
+
|
291
|
+
#
|
292
|
+
# Remove the model from the Solr index and commit immediately. See
|
293
|
+
# #remove_from_index
|
294
|
+
#
|
295
|
+
#---
|
296
|
+
#FIXME Sunspot should implement remove!()
|
297
|
+
#
|
298
|
+
def remove_from_index!
|
299
|
+
Sunspot.remove(self)
|
300
|
+
Sunspot.commit
|
301
|
+
end
|
302
|
+
end
|
303
|
+
end
|
304
|
+
end
|
305
|
+
end
|