thinkingtank 0.0.2 → 0.0.3
Sign up to get free protection for your applications and to get access to all the features.
- data/lib/indextank_client.rb +255 -0
- data/lib/thinkingtank/init.rb +45 -26
- data/lib/thinkingtank/tasks.rb +13 -0
- metadata +4 -4
- data/lib/indextank.rb +0 -150
@@ -0,0 +1,255 @@
|
|
1
|
+
require 'net/http'
|
2
|
+
require 'uri'
|
3
|
+
require 'rubygems'
|
4
|
+
require 'json'
|
5
|
+
|
6
|
+
module IndexTank
|
7
|
+
|
8
|
+
private
|
9
|
+
|
10
|
+
class RestClient
|
11
|
+
def GET(path, params={})
|
12
|
+
path = "#{path}?#{to_query(params)}" if params
|
13
|
+
request = Net::HTTP::Get.new "#{@uri}#{path}"
|
14
|
+
authorize request
|
15
|
+
return execute(request)
|
16
|
+
end
|
17
|
+
|
18
|
+
def PUT(path, body={})
|
19
|
+
request = Net::HTTP::Put.new "#{@uri}#{path}"
|
20
|
+
authorize request
|
21
|
+
request.body = body.to_json if body
|
22
|
+
return execute(request)
|
23
|
+
end
|
24
|
+
|
25
|
+
def DELETE(path, params={})
|
26
|
+
path = "#{path}?#{to_query(params)}" if params
|
27
|
+
request = Net::HTTP::Delete.new "#{@uri}#{path}"
|
28
|
+
authorize request
|
29
|
+
return execute(request)
|
30
|
+
end
|
31
|
+
|
32
|
+
private
|
33
|
+
|
34
|
+
def to_query(params)
|
35
|
+
require 'cgi' unless defined?(CGI) && defined?(CGI::escape)
|
36
|
+
r = ''
|
37
|
+
params.each do |k,v|
|
38
|
+
r << "#{CGI.escape(k.to_s)}=#{CGI.escape(v.to_s)}&"
|
39
|
+
end
|
40
|
+
return r
|
41
|
+
end
|
42
|
+
|
43
|
+
def authorize(req)
|
44
|
+
req.basic_auth(@uri.user, @uri.password)
|
45
|
+
end
|
46
|
+
|
47
|
+
def execute(req)
|
48
|
+
res = Net::HTTP.new(@uri.host).start { |http| http.request(req) }
|
49
|
+
if res.is_a? Net::HTTPSuccess
|
50
|
+
if res.body.nil? or res.body.empty?
|
51
|
+
return res.code, nil
|
52
|
+
else
|
53
|
+
begin
|
54
|
+
return res.code, JSON.parse(res.body)
|
55
|
+
rescue
|
56
|
+
raise "Invalid JSON response: #{res.body}"
|
57
|
+
end
|
58
|
+
end
|
59
|
+
elsif res.is_a? Net::HTTPUnauthorized
|
60
|
+
raise SecurityError, "Authorization required"
|
61
|
+
elsif res.is_a? Net::HTTPBadRequest
|
62
|
+
raise ArgumentError, res.body
|
63
|
+
else
|
64
|
+
raise HttpCodeException.new(res.code, res.body)
|
65
|
+
end
|
66
|
+
end
|
67
|
+
|
68
|
+
end
|
69
|
+
|
70
|
+
public
|
71
|
+
|
72
|
+
class ApiClient < RestClient
|
73
|
+
def initialize(api_url)
|
74
|
+
@uri = URI.parse(api_url)
|
75
|
+
end
|
76
|
+
|
77
|
+
def get_index(name)
|
78
|
+
return IndexClient.new("#{@uri}/v1/indexes/#{name}")
|
79
|
+
end
|
80
|
+
|
81
|
+
def create_index(name)
|
82
|
+
index = get_index(name)
|
83
|
+
index.create_index()
|
84
|
+
return index
|
85
|
+
end
|
86
|
+
|
87
|
+
def delete_index(name)
|
88
|
+
return get_index(name).delete_index()
|
89
|
+
end
|
90
|
+
|
91
|
+
def list_indexes()
|
92
|
+
code, indexes = GET "/v1/indexes"
|
93
|
+
return indexes.map do |name,metadata| IndexClient.new "#{@uri}/v1/indexes/#{name}", metadata end
|
94
|
+
end
|
95
|
+
end
|
96
|
+
|
97
|
+
class IndexClient < RestClient
|
98
|
+
def initialize(index_url, metadata=nil)
|
99
|
+
@uri = URI.parse(index_url)
|
100
|
+
@metadata = metadata
|
101
|
+
end
|
102
|
+
|
103
|
+
def code
|
104
|
+
return metadata['code']
|
105
|
+
end
|
106
|
+
|
107
|
+
def running?
|
108
|
+
return metadata!['started']
|
109
|
+
end
|
110
|
+
|
111
|
+
def creation_time
|
112
|
+
return metadata['creation_time']
|
113
|
+
end
|
114
|
+
|
115
|
+
def size
|
116
|
+
return metadata['size']
|
117
|
+
end
|
118
|
+
|
119
|
+
def exists?
|
120
|
+
begin
|
121
|
+
metadata!
|
122
|
+
return true
|
123
|
+
rescue HttpCodeException
|
124
|
+
if $!.code == "404"
|
125
|
+
return false
|
126
|
+
end
|
127
|
+
raise
|
128
|
+
end
|
129
|
+
end
|
130
|
+
|
131
|
+
# the options argument may contain a :variables key
|
132
|
+
# with a Hash from variable numbers to their float values
|
133
|
+
# this variables can be used in the scoring functions
|
134
|
+
# when sorting a search
|
135
|
+
def add_document(docid, fields, options={})
|
136
|
+
options.merge!( :docid => docid, :fields => fields )
|
137
|
+
code, r = PUT "/docs", options
|
138
|
+
return r
|
139
|
+
end
|
140
|
+
|
141
|
+
def update_variables(docid, variables, options={})
|
142
|
+
options.merge!( :docid => docid, :variables => variables )
|
143
|
+
code, r = PUT "/docs/variables", options
|
144
|
+
return r
|
145
|
+
end
|
146
|
+
|
147
|
+
def delete_document(docid, options={})
|
148
|
+
options.merge!( :docid => docid )
|
149
|
+
code, r = DELETE "/docs", options
|
150
|
+
return r
|
151
|
+
end
|
152
|
+
|
153
|
+
# the options argument may contain an :index_code definition to override
|
154
|
+
# this instance's default index_code
|
155
|
+
def promote(docid, query, options={})
|
156
|
+
options.merge!( :docid => docid, :query => query )
|
157
|
+
code, r = PUT "/promote", options
|
158
|
+
return r
|
159
|
+
end
|
160
|
+
|
161
|
+
|
162
|
+
# the options argument may contain an :index_code definition to override
|
163
|
+
# this instance's default index_code
|
164
|
+
# it can also contain any of the following:
|
165
|
+
# :start => an int with the number of results to skip
|
166
|
+
# :len => an int with the number of results to return
|
167
|
+
# :snippet => a comma separated list of field names for which a snippet
|
168
|
+
# should be returned. (requires an index that supports snippets)
|
169
|
+
# :fetch => a comma separated list of field names for which its content
|
170
|
+
# should be returned. (requires an index that supports storage)
|
171
|
+
# :function => an int with the index of the scoring function to be used
|
172
|
+
# for this query
|
173
|
+
def search(query, options={})
|
174
|
+
options = { :start => 0, :len => 10 }.merge(options)
|
175
|
+
options.merge!( :q => query )
|
176
|
+
begin
|
177
|
+
code, r = GET "/search", options
|
178
|
+
return r
|
179
|
+
rescue HttpCodeException
|
180
|
+
raise
|
181
|
+
end
|
182
|
+
end
|
183
|
+
|
184
|
+
def add_function(function_index, definition, options={})
|
185
|
+
options.merge!( :definition => definition )
|
186
|
+
code, r = PUT "/functions/#{function_index}", options
|
187
|
+
return r
|
188
|
+
end
|
189
|
+
|
190
|
+
def del_function(function_index, options={})
|
191
|
+
code, r = DELETE "/functions/#{function_index}", options
|
192
|
+
return r
|
193
|
+
end
|
194
|
+
|
195
|
+
def list_functions(options={})
|
196
|
+
code, r = GET "/functions"
|
197
|
+
return r
|
198
|
+
end
|
199
|
+
|
200
|
+
def create_index()
|
201
|
+
begin
|
202
|
+
code, r = PUT ""
|
203
|
+
raise IndexAlreadyExists if code == "204"
|
204
|
+
return r
|
205
|
+
rescue HttpCodeException
|
206
|
+
if $!.code == "409"
|
207
|
+
puts $!.code
|
208
|
+
raise TooManyIndexes
|
209
|
+
end
|
210
|
+
raise
|
211
|
+
end
|
212
|
+
end
|
213
|
+
|
214
|
+
def delete_index()
|
215
|
+
code, r = DELETE ""
|
216
|
+
return r
|
217
|
+
end
|
218
|
+
|
219
|
+
def metadata
|
220
|
+
metadata! if @metadata.nil?
|
221
|
+
return @metadata
|
222
|
+
end
|
223
|
+
|
224
|
+
def metadata!
|
225
|
+
code, @metadata = GET ""
|
226
|
+
return @metadata
|
227
|
+
end
|
228
|
+
|
229
|
+
end
|
230
|
+
|
231
|
+
class IndexAlreadyExists < StandardError
|
232
|
+
end
|
233
|
+
class TooManyIndexes < StandardError
|
234
|
+
end
|
235
|
+
|
236
|
+
class HttpCodeException < StandardError
|
237
|
+
def initialize(code, message)
|
238
|
+
@code = code
|
239
|
+
@message = message
|
240
|
+
super("#{code}: #{message}")
|
241
|
+
end
|
242
|
+
|
243
|
+
attr_accessor :code
|
244
|
+
attr_accessor :message
|
245
|
+
end
|
246
|
+
|
247
|
+
class HerokuClient < ApiClient
|
248
|
+
def initialize()
|
249
|
+
super(ENV['HEROKUTANK_API_URL'])
|
250
|
+
end
|
251
|
+
end
|
252
|
+
|
253
|
+
|
254
|
+
end
|
255
|
+
|
data/lib/thinkingtank/init.rb
CHANGED
@@ -1,4 +1,4 @@
|
|
1
|
-
require '
|
1
|
+
require 'indextank_client'
|
2
2
|
|
3
3
|
module ThinkingTank
|
4
4
|
class Builder
|
@@ -31,7 +31,7 @@ module ThinkingTank
|
|
31
31
|
return unless File.exists?(path)
|
32
32
|
|
33
33
|
conf = YAML::load(ERB.new(IO.read(path)).result)[environment]
|
34
|
-
self.client = IndexTank.new(conf['
|
34
|
+
self.client = IndexTank::ApiClient.new(conf['api_url']).get_index(conf['index_name'])
|
35
35
|
end
|
36
36
|
def environment
|
37
37
|
if defined?(Merb)
|
@@ -51,11 +51,11 @@ module ThinkingTank
|
|
51
51
|
data = {}
|
52
52
|
self.class.thinkingtank_builder.index_fields.each do |field|
|
53
53
|
val = self.instance_eval(field.to_s)
|
54
|
-
data[
|
54
|
+
data[field.to_s] = val.to_s unless val.nil?
|
55
55
|
end
|
56
|
-
data[:
|
57
|
-
data[:
|
58
|
-
it.
|
56
|
+
data[:__any] = data.values.join " . "
|
57
|
+
data[:__type] = self.class.name
|
58
|
+
it.add_document(docid, data)
|
59
59
|
end
|
60
60
|
end
|
61
61
|
|
@@ -64,26 +64,10 @@ end
|
|
64
64
|
class << ActiveRecord::Base
|
65
65
|
@indexable = false
|
66
66
|
def search(*args)
|
67
|
-
|
68
|
-
|
69
|
-
|
70
|
-
|
71
|
-
field = "_#{field}" # underscore prepended to ActiveRecord fields
|
72
|
-
query += " #{field}:(#{value})"
|
73
|
-
end
|
74
|
-
end
|
75
|
-
# TODO : add relevance functions
|
76
|
-
|
77
|
-
it = ThinkingTank::Configuration.instance.client
|
78
|
-
models = []
|
79
|
-
ok, res = it.search("#{query.to_s} type:#{self.name}")
|
80
|
-
if ok
|
81
|
-
res['docs'].each do |doc|
|
82
|
-
type, docid = doc['docid'].split(" ", 2)
|
83
|
-
models << self.find(id=docid)
|
84
|
-
end
|
85
|
-
end
|
86
|
-
return models
|
67
|
+
return indextank_search(true, *args)
|
68
|
+
end
|
69
|
+
def search_raw(*args)
|
70
|
+
return indextank_search(false, *args)
|
87
71
|
end
|
88
72
|
|
89
73
|
def define_index(name = nil, &block)
|
@@ -100,6 +84,41 @@ class << ActiveRecord::Base
|
|
100
84
|
def thinkingtank_builder
|
101
85
|
return @thinkingtank_builder
|
102
86
|
end
|
87
|
+
|
88
|
+
private
|
89
|
+
|
90
|
+
def indextank_search(models, *args)
|
91
|
+
options = args.extract_options!
|
92
|
+
query = args.join(' ')
|
93
|
+
|
94
|
+
# transform fields in query
|
95
|
+
|
96
|
+
if options.has_key? :conditions
|
97
|
+
options[:conditions].each do |field,value|
|
98
|
+
query += " #{field}:(#{value})"
|
99
|
+
end
|
100
|
+
end
|
101
|
+
|
102
|
+
options.slice!(:snippet, :fetch, :function)
|
103
|
+
|
104
|
+
it = ThinkingTank::Configuration.instance.client
|
105
|
+
models = []
|
106
|
+
res = it.search("__any:(#{query.to_s}) __type:#{self.name}", options)
|
107
|
+
if models
|
108
|
+
res['results'].each do |doc|
|
109
|
+
type, docid = doc['docid'].split(" ", 2)
|
110
|
+
models << self.find(id=docid)
|
111
|
+
end
|
112
|
+
return models
|
113
|
+
else
|
114
|
+
res['results'].each do |doc|
|
115
|
+
type, docid = doc['docid'].split(" ", 2)
|
116
|
+
doc['model'] = self.find(id=docid)
|
117
|
+
end
|
118
|
+
return res
|
119
|
+
end
|
120
|
+
end
|
121
|
+
|
103
122
|
end
|
104
123
|
|
105
124
|
|
data/lib/thinkingtank/tasks.rb
CHANGED
@@ -1,5 +1,6 @@
|
|
1
1
|
require 'erb'
|
2
2
|
require 'active_record'
|
3
|
+
require 'indextank_client'
|
3
4
|
|
4
5
|
def load_models
|
5
6
|
app_root = ThinkingTank::Configuration.instance.app_root
|
@@ -28,6 +29,18 @@ def load_models
|
|
28
29
|
end
|
29
30
|
|
30
31
|
def reindex_models
|
32
|
+
it = ThinkingTank::Configuration.instance.client
|
33
|
+
if it.exists?
|
34
|
+
puts "Deleting existing index"
|
35
|
+
it.delete_index()
|
36
|
+
end
|
37
|
+
puts "Creating a new empty index"
|
38
|
+
it.create_index()
|
39
|
+
puts "Waiting for the index to be ready"
|
40
|
+
while not it.running?
|
41
|
+
sleep 0.5
|
42
|
+
end
|
43
|
+
|
31
44
|
Object.subclasses_of(ActiveRecord::Base).each do |klass|
|
32
45
|
reindex klass if klass.is_indexable?
|
33
46
|
end
|
metadata
CHANGED
@@ -5,8 +5,8 @@ version: !ruby/object:Gem::Version
|
|
5
5
|
segments:
|
6
6
|
- 0
|
7
7
|
- 0
|
8
|
-
-
|
9
|
-
version: 0.0.
|
8
|
+
- 3
|
9
|
+
version: 0.0.3
|
10
10
|
platform: ruby
|
11
11
|
authors:
|
12
12
|
- Flaptor
|
@@ -14,7 +14,7 @@ autorequire:
|
|
14
14
|
bindir: bin
|
15
15
|
cert_chain: []
|
16
16
|
|
17
|
-
date: 2010-
|
17
|
+
date: 2010-07-26 00:00:00 -07:00
|
18
18
|
default_executable:
|
19
19
|
dependencies: []
|
20
20
|
|
@@ -30,7 +30,7 @@ extra_rdoc_files: []
|
|
30
30
|
|
31
31
|
files:
|
32
32
|
- rails/init.rb
|
33
|
-
- lib/
|
33
|
+
- lib/indextank_client.rb
|
34
34
|
- lib/thinkingtank/init.rb
|
35
35
|
- README
|
36
36
|
- tasks/rails.rake
|
data/lib/indextank.rb
DELETED
@@ -1,150 +0,0 @@
|
|
1
|
-
require 'net/http'
|
2
|
-
require 'uri'
|
3
|
-
require 'rubygems'
|
4
|
-
require 'json'
|
5
|
-
|
6
|
-
BASE_URL = 'http://api.indextank.com/api/v0'
|
7
|
-
class IndexTank
|
8
|
-
def initialize(api_key, options={})
|
9
|
-
@api_key = api_key
|
10
|
-
@def_index_code = options[:index_code]
|
11
|
-
@def_index_name = options[:index_name]
|
12
|
-
end
|
13
|
-
|
14
|
-
def create_index(index_name)
|
15
|
-
return api_call("admin/create", "index_code", :index_name => index_name)
|
16
|
-
end
|
17
|
-
|
18
|
-
# the options argument may contain an :index_code definition to override
|
19
|
-
# this instance's default index_code
|
20
|
-
def delete_index(options={})
|
21
|
-
return api_call("admin/delete", nil, options)
|
22
|
-
end
|
23
|
-
|
24
|
-
|
25
|
-
def list_indexes()
|
26
|
-
return api_call("admin/list", "results")
|
27
|
-
end
|
28
|
-
|
29
|
-
|
30
|
-
# the options argument may contain an :index_code definition to override
|
31
|
-
# this instance's default index_code
|
32
|
-
def add(doc_id, content, options={})
|
33
|
-
options.merge!(:document_id => doc_id, :document => content.to_json)
|
34
|
-
options[:boosts] = options[:boosts].to_json if options.key?(:boosts)
|
35
|
-
return api_call("index/add", nil, options)
|
36
|
-
end
|
37
|
-
|
38
|
-
# the options argument may contain an :index_code definition to override
|
39
|
-
# this instance's default index_code
|
40
|
-
def boost(doc_id, boosts, options={})
|
41
|
-
options.merge!(:document_id => doc_id, :boosts => boosts.to_json)
|
42
|
-
return api_call("index/boost", nil, options)
|
43
|
-
end
|
44
|
-
|
45
|
-
|
46
|
-
# the options argument may contain an :index_code definition to override
|
47
|
-
# this instance's default index_code
|
48
|
-
def promote(doc_id, query, options={})
|
49
|
-
options.merge!(:document_id => doc_id, :query => query)
|
50
|
-
return api_call("index/promote", nil, options)
|
51
|
-
end
|
52
|
-
|
53
|
-
|
54
|
-
# the options argument may contain an :index_code definition to override
|
55
|
-
# this instance's default index_code
|
56
|
-
def update(doc_id, content, options={})
|
57
|
-
options.merge!(:document_id => doc_id, :document => content.to_json)
|
58
|
-
return api_call("index/update", nil, options)
|
59
|
-
end
|
60
|
-
|
61
|
-
# the options argument may contain an :index_code definition to override
|
62
|
-
# this instance's default index_code
|
63
|
-
def delete(doc_id, options={})
|
64
|
-
options.merge!(:document_id => doc_id)
|
65
|
-
return api_call("index/delete", nil, options)
|
66
|
-
end
|
67
|
-
|
68
|
-
# the options argument may contain an :index_code definition to override
|
69
|
-
# this instance's default index_code
|
70
|
-
# it can also contain any of the following:
|
71
|
-
# :start => an int with the number of results to skip
|
72
|
-
# :len => an int with the number of results to return
|
73
|
-
# :snippet_fields => a comma separated list of field names for which a snippet
|
74
|
-
# should be returned. (requires an index that supports snippets)
|
75
|
-
# :fetch_fields => a comma separated list of field names for which its content
|
76
|
-
# should be returned. (requires an index that supports storage)
|
77
|
-
# :relevance_function => an int with the index of the relevance function to be used
|
78
|
-
# for this query
|
79
|
-
def search(query, options={})
|
80
|
-
options = { :start => 0, :len => 10 }.merge(options)
|
81
|
-
options.merge!(:query => query)
|
82
|
-
return api_call("search/query", "results", options)
|
83
|
-
end
|
84
|
-
|
85
|
-
# the options argument may contain an :index_code definition to override
|
86
|
-
# this instance's default index_code
|
87
|
-
def add_function(function_index, definition, options={})
|
88
|
-
options.merge!( :function_id => function_index, :definition => definition )
|
89
|
-
return api_call("index/add_function", nil, options)
|
90
|
-
end
|
91
|
-
|
92
|
-
# the options argument may contain an :index_code definition to override
|
93
|
-
# this instance's default index_code
|
94
|
-
def del_function(function_index, options={})
|
95
|
-
options.merge!( :function_id => function_index )
|
96
|
-
return api_call("index/remove_function", nil, options)
|
97
|
-
end
|
98
|
-
|
99
|
-
# the options argument may contain an :index_code definition to override
|
100
|
-
# this instance's default index_code
|
101
|
-
def list_functions(options={})
|
102
|
-
return api_call("index/list_functions", "results", options)
|
103
|
-
end
|
104
|
-
|
105
|
-
|
106
|
-
# the options argument may contain an :index_code definition to override
|
107
|
-
# this instance's default index_code
|
108
|
-
def index_stats(options={})
|
109
|
-
return api_call("index/stats", "results", options)
|
110
|
-
end
|
111
|
-
|
112
|
-
|
113
|
-
# the options argument may contain an :index_code definition to override
|
114
|
-
# this instance's default index_code
|
115
|
-
def search_stats(options={})
|
116
|
-
return api_call("search/stats", "results", options)
|
117
|
-
end
|
118
|
-
|
119
|
-
private
|
120
|
-
|
121
|
-
def base_url
|
122
|
-
return ENV['INDEXTANK_BASE_URL'] || BASE_URL
|
123
|
-
end
|
124
|
-
|
125
|
-
def api_call(method, return_key, params={})
|
126
|
-
params = { "api_key" => @api_key }.merge(params)
|
127
|
-
params = { :index_code => @def_index_code }.merge(params) unless @def_index_code.nil?
|
128
|
-
params = { :index_name => @def_index_name }.merge(params) unless @def_index_name.nil?
|
129
|
-
url = base_url + '/' + method
|
130
|
-
req = Net::HTTP::Post.new(url)
|
131
|
-
req.set_form_data(params, ';')
|
132
|
-
res = Net::HTTP.new(URI.parse(base_url).host).start {|http| http.request(req) }
|
133
|
-
if res.is_a? Net::HTTPOK
|
134
|
-
result = JSON.parse(res.body)
|
135
|
-
ok = result['status'] == 'OK'
|
136
|
-
return ok ? [ok, (return_key.nil? ? nil : result[return_key])] : [ok, result['message']]
|
137
|
-
elsif res.is_a? Net::HTTPForbidden
|
138
|
-
return false, "Access forbidden"
|
139
|
-
elsif res.is_a? Net::HTTPClientError
|
140
|
-
return false, "Unknown client error"
|
141
|
-
elsif res.is_a? Net::HTTPServerError
|
142
|
-
puts res.body
|
143
|
-
return false, "Unknown server error"
|
144
|
-
else
|
145
|
-
puts res.body
|
146
|
-
return false, "Unexpected response"
|
147
|
-
end
|
148
|
-
end
|
149
|
-
|
150
|
-
end
|