sphinx 0.9.10.2043 → 0.9.10.2091
Sign up to get free protection for your applications and to get access to all the features.
- data/.gitignore +2 -0
- data/README.rdoc +172 -18
- data/Rakefile +24 -15
- data/VERSION.yml +1 -1
- data/lib/sphinx.rb +15 -3
- data/lib/sphinx/buffered_io.rb +4 -0
- data/lib/sphinx/client.rb +1472 -666
- data/lib/sphinx/request.rb +2 -0
- data/lib/sphinx/response.rb +2 -0
- data/lib/sphinx/server.rb +4 -1
- data/lib/sphinx/timeout.rb +4 -1
- data/spec/sphinx/sphinx-id64.conf +1 -1
- data/spec/sphinx/sphinx.conf +1 -1
- data/spec/sphinx/sphinx_test.sql +4 -2
- data/sphinx.gemspec +2 -2
- metadata +2 -2
data/.gitignore
CHANGED
data/README.rdoc
CHANGED
@@ -1,10 +1,10 @@
|
|
1
|
-
=Sphinx Client API 0.9.10
|
1
|
+
= Sphinx Client API 0.9.10
|
2
2
|
|
3
|
-
This document gives an overview of what is Sphinx itself and how to use
|
4
|
-
|
3
|
+
This document gives an overview of what is Sphinx itself and how to use it
|
4
|
+
from your Ruby on Rails application. For more information or documentation,
|
5
5
|
please go to http://www.sphinxsearch.com
|
6
6
|
|
7
|
-
==Sphinx
|
7
|
+
== Sphinx
|
8
8
|
|
9
9
|
Sphinx is a standalone full-text search engine, meant to provide fast,
|
10
10
|
size-efficient and relevant fulltext search functions to other applications.
|
@@ -12,37 +12,191 @@ Sphinx was specially designed to integrate well with SQL databases and
|
|
12
12
|
scripting languages. Currently built-in data sources support fetching data
|
13
13
|
either via direct connection to MySQL, or from an XML pipe.
|
14
14
|
|
15
|
-
Simplest way to communicate with Sphinx is to use <tt>searchd</tt>
|
16
|
-
a daemon to search through
|
15
|
+
Simplest way to communicate with Sphinx is to use <tt>searchd</tt> —
|
16
|
+
a daemon to search through full text indexes from external software.
|
17
17
|
|
18
|
-
==
|
18
|
+
== Installation
|
19
19
|
|
20
|
-
|
20
|
+
There are two options when approaching sphinx plugin installation:
|
21
21
|
|
22
|
-
*
|
23
|
-
*
|
22
|
+
* using the gem (recommended)
|
23
|
+
* install as a Rails plugin
|
24
24
|
|
25
|
-
|
25
|
+
To install as a gem, add this to your environment.rb:
|
26
26
|
|
27
|
-
|
27
|
+
config.gem 'sphinx', :source => 'http://gemcutter.org'
|
28
28
|
|
29
|
-
|
29
|
+
And then run the command:
|
30
30
|
|
31
|
-
|
31
|
+
sudo rake gems:install
|
32
32
|
|
33
|
-
|
33
|
+
To install Sphinx as a Rails plugin use this:
|
34
|
+
|
35
|
+
script/plugin install git://github.com/kpumuk/sphinx.git
|
36
|
+
|
37
|
+
== Documentation
|
38
|
+
|
39
|
+
Complete Sphinx plugin documentation could be found here:
|
40
|
+
http://kpumuk.github.com/sphinx
|
41
|
+
|
42
|
+
Also you can find documentation on rdoc.info:
|
43
|
+
http://rdoc.info/projects/kpumuk/sphinx
|
44
|
+
|
45
|
+
You can build the documentation locally by running:
|
46
|
+
|
47
|
+
rake yard
|
48
|
+
|
49
|
+
Please note: you should have yard gem installed on your system:
|
50
|
+
|
51
|
+
sudo gem install yard --source http://gemcutter.org
|
52
|
+
|
53
|
+
Complete Sphinx API documentation could be found on Sphinx Search Engine
|
54
|
+
site: http://www.sphinxsearch.com/docs/current.html
|
55
|
+
This plugin is fully compatible with original PHP API implementation.
|
56
|
+
|
57
|
+
== Ruby naming conventions
|
58
|
+
|
59
|
+
Sphinx Client API supports Ruby naming conventions, so every API
|
60
|
+
method name is in underscored, lowercase form:
|
61
|
+
|
62
|
+
SetServer -> set_server
|
63
|
+
RunQueries -> run_queries
|
64
|
+
SetMatchMode -> set_match_mode
|
65
|
+
|
66
|
+
Every method is aliased to a corresponding one from standard Sphinx
|
67
|
+
API, so you can use both <tt>SetServer</tt> and <tt>set_server</tt>
|
68
|
+
with no differrence.
|
69
|
+
|
70
|
+
There are three exceptions to this naming rule:
|
71
|
+
|
72
|
+
GetLastError -> last_error
|
73
|
+
GetLastWarning -> last_warning
|
74
|
+
IsConnectError -> connect_error?
|
75
|
+
|
76
|
+
Of course, all of them are aliased to the original method names.
|
77
|
+
|
78
|
+
== Using multiple Sphinx servers
|
79
|
+
|
80
|
+
Since we actively use this plugin in our Scribd development workflow,
|
81
|
+
there are several methods have been added to accommodate our needs.
|
82
|
+
You can find documentation on Ruby-specific methods in documentation:
|
83
|
+
http://rdoc.info/projects/kpumuk/sphinx
|
84
|
+
|
85
|
+
First of all, we added support of multiple Sphinx servers to balance
|
86
|
+
load between them. Also it means that in case of any problems with one
|
87
|
+
of servers, library will try to fetch the results from another one.
|
88
|
+
Every consequence request will be executed on the next server in list
|
89
|
+
(round-robin technique).
|
90
|
+
|
91
|
+
sphinx.set_servers([
|
92
|
+
{ :host => 'browse01.local', :port => 3312 },
|
93
|
+
{ :host => 'browse02.local', :port => 3312 },
|
94
|
+
{ :host => 'browse03.local', :port => 3312 }
|
95
|
+
])
|
96
|
+
|
97
|
+
By default library will try to fetch results from a single server, and
|
98
|
+
fail if it does not respond. To setup number of retries being performed,
|
99
|
+
you can use second (additional) parameter of the <tt>set_connect_timeout</tt>
|
100
|
+
and <tt>set_request_timeout</tt> methods:
|
101
|
+
|
102
|
+
sphinx.set_connect_timeout(1, 3)
|
103
|
+
sphinx.set_request_timeout(1, 3)
|
104
|
+
|
105
|
+
There is a big difference between these two methods. First will affect
|
106
|
+
only on requests experiencing problems with connection (socket error,
|
107
|
+
pipe error, etc), second will be used when request is broken somehow
|
108
|
+
(temporary searchd error, incomplete reply, etc). The workflow looks like
|
109
|
+
this:
|
110
|
+
|
111
|
+
1. Increase retries number. If is less or equal to configured value,
|
112
|
+
try to connect to the next server. Otherwise, raise an error.
|
113
|
+
2. In case of connection problem go to 1.
|
114
|
+
3. Increase request retries number. If it less or equal to configured
|
115
|
+
value, try to perform request. Otherwise, raise an error.
|
116
|
+
4. In case of connection problem go to 1.
|
117
|
+
5. In case of request problem, go to 3.
|
118
|
+
6. Parse and return response.
|
119
|
+
|
120
|
+
Withdrawals:
|
121
|
+
|
122
|
+
1. Request could be performed <tt>connect_retries</tt> * <tt>request_retries</tt>
|
123
|
+
times. E.g., it could be tried <tt>request_retries</tt> times on each
|
124
|
+
of <tt>connect_retries</tt> servers (when you have 1 server configured,
|
125
|
+
but <tt>connect_retries</tt> is 5, library will try to connect to this
|
126
|
+
server 5 times).
|
127
|
+
2. Request could be tried to execute on each server <tt>1..request_retries</tt>
|
128
|
+
times. In case of connection problem, request will be moved to another
|
129
|
+
server immediately.
|
130
|
+
|
131
|
+
Usually you will set <tt>connect_retries</tt> equal to servers number,
|
132
|
+
so you will be sure each failing request will be performed on all servers.
|
133
|
+
This means that if one of servers is live, but others are dead, you request
|
134
|
+
will be finally executed successfully.
|
135
|
+
|
136
|
+
== Sphinx constants
|
137
|
+
|
138
|
+
Most Sphinx API methods expecting for special constants will be passed.
|
139
|
+
For example:
|
140
|
+
|
141
|
+
sphinx.set_match_mode(Sphinx::Client::SPH_MATCH_ANY)
|
142
|
+
|
143
|
+
Please note that these constants defined in a <tt>Sphinx::Client</tt>
|
144
|
+
namespace. You can use symbols or strings instead of these awful
|
145
|
+
constants:
|
146
|
+
|
147
|
+
sphinx.set_match_mode(:any)
|
148
|
+
sphinx.set_match_mode('any')
|
149
|
+
|
150
|
+
== Setting query filters
|
151
|
+
|
152
|
+
Every <tt>set_</tt> method returns <tt>Sphinx::Client</tt> object itself.
|
153
|
+
It means that you can chain filtering methods:
|
154
|
+
|
155
|
+
results = Sphinx::Client.new.
|
156
|
+
set_match_mode(:any).
|
157
|
+
set_ranking_mode(:bm25).
|
158
|
+
set_id_range(10, 1000).
|
159
|
+
query('test')
|
160
|
+
|
161
|
+
== Example
|
162
|
+
|
163
|
+
This simple example illustrates base connection establishing,
|
164
|
+
search results retrieving, and excerpts building. Please note
|
165
|
+
how does it perform database select using ActiveRecord to
|
166
|
+
save the order of records established by Sphinx.
|
167
|
+
|
168
|
+
sphinx = Sphinx::Client.new
|
169
|
+
result = sphinx.query('test')
|
170
|
+
ids = result['matches'].map { |match| match['id'] }
|
171
|
+
posts = Post.all :conditions => { :id => ids },
|
172
|
+
:order => "FIELD(id,#{ids.join(',')})"
|
173
|
+
|
174
|
+
docs = posts.map(&:body)
|
175
|
+
excerpts = sphinx.build_excerpts(docs, 'index', 'test')
|
176
|
+
|
177
|
+
== Support
|
178
|
+
|
179
|
+
Source code:
|
180
|
+
http://github.com/kpumuk/sphinx
|
181
|
+
|
182
|
+
To suggest a feature or report a bug:
|
183
|
+
http://github.com/kpumuk/sphinx/issues
|
184
|
+
|
185
|
+
Project home page:
|
34
186
|
http://kpumuk.info/projects/ror-plugins/sphinx
|
35
187
|
|
36
|
-
==Credits
|
188
|
+
== Credits
|
37
189
|
|
38
190
|
Dmytro Shteflyuk <kpumuk@kpumuk.info> http://kpumuk.info
|
39
191
|
|
40
|
-
Andrew Aksyonoff http://sphinxsearch.com
|
192
|
+
Andrew Aksyonoff http://sphinxsearch.com
|
41
193
|
|
42
194
|
Special thanks to Alexey Kovyrin <alexey@kovyrin.net> http://blog.kovyrin.net
|
43
195
|
|
196
|
+
Special thanks to Mike Perham http://www.mikeperham.com for his awesome
|
197
|
+
memcache-client gem, where latest Sphinx gem got new sockets handling from.
|
198
|
+
|
44
199
|
==License
|
45
200
|
|
46
201
|
This library is distributed under the terms of the Ruby license.
|
47
202
|
You can freely distribute/modify this library.
|
48
|
-
|
data/Rakefile
CHANGED
@@ -1,6 +1,4 @@
|
|
1
1
|
require 'rake'
|
2
|
-
require 'spec/rake/spectask'
|
3
|
-
require 'rake/rdoctask'
|
4
2
|
|
5
3
|
begin
|
6
4
|
require 'jeweler'
|
@@ -17,20 +15,31 @@ rescue LoadError
|
|
17
15
|
puts 'Jeweler not available. Install it with: sudo gem install jeweler'
|
18
16
|
end
|
19
17
|
|
20
|
-
|
21
|
-
|
18
|
+
begin
|
19
|
+
require 'spec/rake/spectask'
|
20
|
+
|
21
|
+
desc 'Default: run specs'
|
22
|
+
task :default => :spec
|
22
23
|
|
23
|
-
desc 'Test the sphinx plugin'
|
24
|
-
Spec::Rake::SpecTask.new
|
25
|
-
|
26
|
-
|
24
|
+
desc 'Test the sphinx plugin'
|
25
|
+
Spec::Rake::SpecTask.new do |t|
|
26
|
+
t.libs << 'lib'
|
27
|
+
t.pattern = 'spec/*_spec.rb'
|
28
|
+
end
|
29
|
+
rescue LoadError
|
30
|
+
puts 'RSpec not available. Install it with: sudo gem install rspec'
|
27
31
|
end
|
28
32
|
|
29
|
-
|
30
|
-
|
31
|
-
|
32
|
-
|
33
|
-
|
34
|
-
|
35
|
-
|
33
|
+
begin
|
34
|
+
require 'yard'
|
35
|
+
YARD::Rake::YardocTask.new(:yard) do |t|
|
36
|
+
t.options = ['--title', 'Sphinx Client API Documentation']
|
37
|
+
if ENV['PRIVATE']
|
38
|
+
t.options.concat ['--protected', '--private']
|
39
|
+
else
|
40
|
+
t.options << '--no-private'
|
41
|
+
end
|
42
|
+
end
|
43
|
+
rescue LoadError
|
44
|
+
puts 'Yard not available. Install it with: sudo gem install yard'
|
36
45
|
end
|
data/VERSION.yml
CHANGED
data/lib/sphinx.rb
CHANGED
@@ -1,9 +1,21 @@
|
|
1
|
-
|
2
|
-
|
3
|
-
|
1
|
+
# Sphinx Client API
|
2
|
+
#
|
3
|
+
# Author:: Dmytro Shteflyuk <mailto:kpumuk@kpumuk.info>.
|
4
|
+
# Copyright:: Copyright (c) 2006 — 2009 Dmytro Shteflyuk
|
5
|
+
# License:: Distributes under the same terms as Ruby
|
6
|
+
# Version:: 0.9.10-r2091
|
7
|
+
# Website:: http://kpumuk.info/projects/ror-plugins/sphinx
|
8
|
+
# Sources:: http://github.com/kpumuk/sphinx
|
9
|
+
#
|
10
|
+
# This library is distributed under the terms of the Ruby license.
|
11
|
+
# You can freely distribute/modify this library.
|
12
|
+
#
|
4
13
|
module Sphinx
|
5
14
|
end
|
6
15
|
|
16
|
+
require 'socket'
|
17
|
+
require 'net/protocol'
|
18
|
+
|
7
19
|
require File.dirname(__FILE__) + '/sphinx/request'
|
8
20
|
require File.dirname(__FILE__) + '/sphinx/response'
|
9
21
|
require File.dirname(__FILE__) + '/sphinx/timeout'
|
data/lib/sphinx/buffered_io.rb
CHANGED
data/lib/sphinx/client.rb
CHANGED
@@ -1,120 +1,149 @@
|
|
1
|
-
# = client.rb - Sphinx Client API
|
2
|
-
#
|
3
|
-
# Author:: Dmytro Shteflyuk <mailto:kpumuk@kpumuk.info>.
|
4
|
-
# Copyright:: Copyright (c) 2006 — 2009 Dmytro Shteflyuk
|
5
|
-
# License:: Distributes under the same terms as Ruby
|
6
|
-
# Version:: 0.9.10-r2043
|
7
|
-
# Website:: http://kpumuk.info/projects/ror-plugins/sphinx
|
8
|
-
#
|
9
|
-
# This library is distributed under the terms of the Ruby license.
|
10
|
-
# You can freely distribute/modify this library.
|
11
|
-
|
12
|
-
# ==Sphinx Client API
|
13
|
-
#
|
14
|
-
# The Sphinx Client API is used to communicate with <tt>searchd</tt>
|
15
|
-
# daemon and get search results from Sphinx.
|
16
|
-
#
|
17
|
-
# ===Usage
|
18
|
-
#
|
19
|
-
# sphinx = Sphinx::Client.new
|
20
|
-
# result = sphinx.Query('test')
|
21
|
-
# ids = result['matches'].map { |match| match['id'] }.join(',')
|
22
|
-
# posts = Post.find :all, :conditions => "id IN (#{ids})"
|
23
|
-
#
|
24
|
-
# docs = posts.map(&:body)
|
25
|
-
# excerpts = sphinx.BuildExcerpts(docs, 'index', 'test')
|
26
|
-
|
27
1
|
module Sphinx
|
28
|
-
#
|
29
|
-
|
2
|
+
# Base class for all Sphinx errors
|
30
3
|
class SphinxError < StandardError; end
|
4
|
+
# Connect error occurred on the API side.
|
31
5
|
class SphinxConnectError < SphinxError; end
|
6
|
+
# Request error occurred on the API side.
|
32
7
|
class SphinxResponseError < SphinxError; end
|
8
|
+
# Internal error occurred inside searchd.
|
33
9
|
class SphinxInternalError < SphinxError; end
|
10
|
+
# Temporary error occurred inside searchd.
|
34
11
|
class SphinxTemporaryError < SphinxError; end
|
12
|
+
# Unknown error occurred inside searchd.
|
35
13
|
class SphinxUnknownError < SphinxError; end
|
36
14
|
|
37
|
-
#
|
38
|
-
|
15
|
+
# The Sphinx Client API is used to communicate with <tt>searchd</tt>
|
16
|
+
# daemon and perform requests.
|
17
|
+
#
|
18
|
+
# @example
|
19
|
+
# sphinx = Sphinx::Client.new
|
20
|
+
# result = sphinx.query('test')
|
21
|
+
# ids = result['matches'].map { |match| match['id'] }
|
22
|
+
# posts = Post.all :conditions => { :id => ids },
|
23
|
+
# :order => "FIELD(id,#{ids.join(',')})"
|
24
|
+
#
|
25
|
+
# docs = posts.map(&:body)
|
26
|
+
# excerpts = sphinx.build_excerpts(docs, 'index', 'test')
|
27
|
+
#
|
39
28
|
class Client
|
40
|
-
|
41
|
-
|
29
|
+
#=================================================================
|
42
30
|
# Known searchd commands
|
43
|
-
|
31
|
+
#=================================================================
|
32
|
+
|
44
33
|
# search command
|
34
|
+
# @private
|
45
35
|
SEARCHD_COMMAND_SEARCH = 0
|
46
36
|
# excerpt command
|
37
|
+
# @private
|
47
38
|
SEARCHD_COMMAND_EXCERPT = 1
|
48
39
|
# update command
|
40
|
+
# @private
|
49
41
|
SEARCHD_COMMAND_UPDATE = 2
|
50
42
|
# keywords command
|
43
|
+
# @private
|
51
44
|
SEARCHD_COMMAND_KEYWORDS = 3
|
52
45
|
# persist command
|
46
|
+
# @private
|
53
47
|
SEARCHD_COMMAND_PERSIST = 4
|
54
48
|
# status command
|
49
|
+
# @private
|
55
50
|
SEARCHD_COMMAND_STATUS = 5
|
56
51
|
# query command
|
52
|
+
# @private
|
57
53
|
SEARCHD_COMMAND_QUERY = 6
|
58
54
|
# flushattrs command
|
55
|
+
# @private
|
59
56
|
SEARCHD_COMMAND_FLUSHATTRS = 7
|
60
|
-
|
57
|
+
|
58
|
+
#=================================================================
|
61
59
|
# Current client-side command implementation versions
|
62
|
-
|
60
|
+
#=================================================================
|
61
|
+
|
63
62
|
# search command version
|
63
|
+
# @private
|
64
64
|
VER_COMMAND_SEARCH = 0x117
|
65
65
|
# excerpt command version
|
66
|
+
# @private
|
66
67
|
VER_COMMAND_EXCERPT = 0x100
|
67
68
|
# update command version
|
69
|
+
# @private
|
68
70
|
VER_COMMAND_UPDATE = 0x102
|
69
71
|
# keywords command version
|
72
|
+
# @private
|
70
73
|
VER_COMMAND_KEYWORDS = 0x100
|
71
74
|
# persist command version
|
75
|
+
# @private
|
72
76
|
VER_COMMAND_PERSIST = 0x000
|
73
77
|
# status command version
|
78
|
+
# @private
|
74
79
|
VER_COMMAND_STATUS = 0x100
|
75
80
|
# query command version
|
81
|
+
# @private
|
76
82
|
VER_COMMAND_QUERY = 0x100
|
77
83
|
# flushattrs command version
|
84
|
+
# @private
|
78
85
|
VER_COMMAND_FLUSHATTRS = 0x100
|
79
|
-
|
86
|
+
|
87
|
+
#=================================================================
|
80
88
|
# Known searchd status codes
|
81
|
-
|
89
|
+
#=================================================================
|
90
|
+
|
82
91
|
# general success, command-specific reply follows
|
92
|
+
# @private
|
83
93
|
SEARCHD_OK = 0
|
84
94
|
# general failure, command-specific reply may follow
|
95
|
+
# @private
|
85
96
|
SEARCHD_ERROR = 1
|
86
97
|
# temporaty failure, client should retry later
|
98
|
+
# @private
|
87
99
|
SEARCHD_RETRY = 2
|
88
|
-
# general success, warning message and command-specific reply follow
|
89
|
-
|
100
|
+
# general success, warning message and command-specific reply follow
|
101
|
+
# @private
|
102
|
+
SEARCHD_WARNING = 3
|
90
103
|
|
104
|
+
#=================================================================
|
105
|
+
# Some internal attributes to use inside client API
|
106
|
+
#=================================================================
|
107
|
+
|
108
|
+
# List of searchd servers to connect to.
|
109
|
+
# @private
|
91
110
|
attr_reader :servers
|
111
|
+
# Connection timeout in seconds.
|
112
|
+
# @private
|
92
113
|
attr_reader :timeout
|
114
|
+
# Number of connection retries.
|
115
|
+
# @private
|
93
116
|
attr_reader :retries
|
117
|
+
# Request timeout in seconds.
|
118
|
+
# @private
|
94
119
|
attr_reader :reqtimeout
|
120
|
+
# Number of request retries.
|
121
|
+
# @private
|
95
122
|
attr_reader :reqretries
|
96
|
-
|
97
|
-
|
98
|
-
|
123
|
+
|
124
|
+
#=================================================================
|
99
125
|
# Known match modes
|
100
|
-
|
126
|
+
#=================================================================
|
127
|
+
|
101
128
|
# match all query words
|
102
|
-
SPH_MATCH_ALL = 0
|
129
|
+
SPH_MATCH_ALL = 0
|
103
130
|
# match any query word
|
104
|
-
SPH_MATCH_ANY = 1
|
131
|
+
SPH_MATCH_ANY = 1
|
105
132
|
# match this exact phrase
|
106
|
-
SPH_MATCH_PHRASE = 2
|
133
|
+
SPH_MATCH_PHRASE = 2
|
107
134
|
# match this boolean query
|
108
|
-
SPH_MATCH_BOOLEAN = 3
|
135
|
+
SPH_MATCH_BOOLEAN = 3
|
109
136
|
# match this extended query
|
110
|
-
SPH_MATCH_EXTENDED = 4
|
137
|
+
SPH_MATCH_EXTENDED = 4
|
111
138
|
# match all document IDs w/o fulltext query, apply filters
|
112
139
|
SPH_MATCH_FULLSCAN = 5
|
113
140
|
# extended engine V2 (TEMPORARY, WILL BE REMOVED IN 0.9.8-RELEASE)
|
114
141
|
SPH_MATCH_EXTENDED2 = 6
|
115
|
-
|
142
|
+
|
143
|
+
#=================================================================
|
116
144
|
# Known ranking modes (ext2 only)
|
117
|
-
|
145
|
+
#=================================================================
|
146
|
+
|
118
147
|
# default mode, phrase proximity major factor and BM25 minor one
|
119
148
|
SPH_RANK_PROXIMITY_BM25 = 0
|
120
149
|
# statistical mode, BM25 ranking only (faster but worse quality)
|
@@ -131,9 +160,11 @@ module Sphinx
|
|
131
160
|
SPH_RANK_FIELDMASK = 6
|
132
161
|
# codename SPH04, phrase proximity + bm25 + head/exact boost
|
133
162
|
SPH_RANK_SPH04 = 7
|
134
|
-
|
163
|
+
|
164
|
+
#=================================================================
|
135
165
|
# Known sort modes
|
136
|
-
|
166
|
+
#=================================================================
|
167
|
+
|
137
168
|
# sort by document relevance desc, then by date
|
138
169
|
SPH_SORT_RELEVANCE = 0
|
139
170
|
# sort by document date desc, then by relevance desc
|
@@ -146,23 +177,27 @@ module Sphinx
|
|
146
177
|
SPH_SORT_EXTENDED = 4
|
147
178
|
# sort by arithmetic expression in descending order (eg. "@id + max(@weight,1000)*boost + log(price)")
|
148
179
|
SPH_SORT_EXPR = 5
|
149
|
-
|
180
|
+
|
181
|
+
#=================================================================
|
150
182
|
# Known filter types
|
151
|
-
|
183
|
+
#=================================================================
|
184
|
+
|
152
185
|
# filter by integer values set
|
153
186
|
SPH_FILTER_VALUES = 0
|
154
187
|
# filter by integer range
|
155
188
|
SPH_FILTER_RANGE = 1
|
156
189
|
# filter by float range
|
157
190
|
SPH_FILTER_FLOATRANGE = 2
|
158
|
-
|
191
|
+
|
192
|
+
#=================================================================
|
159
193
|
# Known attribute types
|
160
|
-
|
194
|
+
#=================================================================
|
195
|
+
|
161
196
|
# this attr is just an integer
|
162
197
|
SPH_ATTR_INTEGER = 1
|
163
198
|
# this attr is a timestamp
|
164
199
|
SPH_ATTR_TIMESTAMP = 2
|
165
|
-
# this attr is an ordinal string number (integer at search time,
|
200
|
+
# this attr is an ordinal string number (integer at search time,
|
166
201
|
# specially handled at indexing time)
|
167
202
|
SPH_ATTR_ORDINAL = 3
|
168
203
|
# this attr is a boolean bit field
|
@@ -175,23 +210,25 @@ module Sphinx
|
|
175
210
|
SPH_ATTR_STRING = 7
|
176
211
|
# this attr has multiple values (0 or more)
|
177
212
|
SPH_ATTR_MULTI = 0x40000000
|
178
|
-
|
213
|
+
|
214
|
+
#=================================================================
|
179
215
|
# Known grouping functions
|
180
|
-
|
216
|
+
#=================================================================
|
217
|
+
|
181
218
|
# group by day
|
182
219
|
SPH_GROUPBY_DAY = 0
|
183
220
|
# group by week
|
184
|
-
SPH_GROUPBY_WEEK = 1
|
221
|
+
SPH_GROUPBY_WEEK = 1
|
185
222
|
# group by month
|
186
|
-
SPH_GROUPBY_MONTH = 2
|
223
|
+
SPH_GROUPBY_MONTH = 2
|
187
224
|
# group by year
|
188
225
|
SPH_GROUPBY_YEAR = 3
|
189
226
|
# group by attribute value
|
190
227
|
SPH_GROUPBY_ATTR = 4
|
191
228
|
# group by sequential attrs pair
|
192
229
|
SPH_GROUPBY_ATTRPAIR = 5
|
193
|
-
|
194
|
-
# Constructs the <tt>Sphinx::Client</tt> object and sets options to their default values.
|
230
|
+
|
231
|
+
# Constructs the <tt>Sphinx::Client</tt> object and sets options to their default values.
|
195
232
|
def initialize
|
196
233
|
# per-query settings
|
197
234
|
@offset = 0 # how many records to seek from result-set start (default is 0)
|
@@ -214,16 +251,16 @@ module Sphinx
|
|
214
251
|
@anchor = [] # geographical anchor point
|
215
252
|
@indexweights = [] # per-index weights
|
216
253
|
@ranker = SPH_RANK_PROXIMITY_BM25 # ranking mode (default is SPH_RANK_PROXIMITY_BM25)
|
217
|
-
@maxquerytime = 0 # max query time, milliseconds (default is 0, do not limit)
|
254
|
+
@maxquerytime = 0 # max query time, milliseconds (default is 0, do not limit)
|
218
255
|
@fieldweights = {} # per-field-name weights
|
219
256
|
@overrides = [] # per-query attribute values overrides
|
220
257
|
@select = '*' # select-list (attributes or expressions, with optional aliases)
|
221
|
-
|
258
|
+
|
222
259
|
# per-reply fields (for single-query case)
|
223
260
|
@error = '' # last error message
|
224
261
|
@warning = '' # last warning message
|
225
262
|
@connerror = false # connection error vs remote error flag
|
226
|
-
|
263
|
+
|
227
264
|
@reqs = [] # requests storage (for multi-query case)
|
228
265
|
@mbenc = '' # stored mbstring encoding
|
229
266
|
@timeout = 0 # connect timeout
|
@@ -233,58 +270,104 @@ module Sphinx
|
|
233
270
|
|
234
271
|
# per-client-object settings
|
235
272
|
# searchd servers list
|
236
|
-
@servers = [Sphinx::Server.new(self, 'localhost',
|
273
|
+
@servers = [Sphinx::Server.new(self, 'localhost', 9312, false)].freeze
|
237
274
|
@lastserver = -1
|
238
275
|
end
|
239
|
-
|
276
|
+
|
277
|
+
#=================================================================
|
278
|
+
# General API functions
|
279
|
+
#=================================================================
|
280
|
+
|
240
281
|
# Returns last error message, as a string, in human readable format. If there
|
241
282
|
# were no errors during the previous API call, empty string is returned.
|
242
283
|
#
|
243
|
-
# You should call it when any other function (such as
|
284
|
+
# You should call it when any other function (such as {#query}) fails (typically,
|
244
285
|
# the failing function returns false). The returned string will contain the
|
245
286
|
# error description.
|
246
287
|
#
|
247
288
|
# The error message is not reset by this call; so you can safely call it
|
248
289
|
# several times if needed.
|
249
290
|
#
|
250
|
-
|
291
|
+
# @return [String] last error message.
|
292
|
+
#
|
293
|
+
# @example
|
294
|
+
# puts sphinx.last_error
|
295
|
+
#
|
296
|
+
# @see #last_warning
|
297
|
+
# @see #connect_error?
|
298
|
+
#
|
299
|
+
def last_error
|
251
300
|
@error
|
252
301
|
end
|
253
|
-
|
302
|
+
alias :GetLastError :last_error
|
303
|
+
|
254
304
|
# Returns last warning message, as a string, in human readable format. If there
|
255
305
|
# were no warnings during the previous API call, empty string is returned.
|
256
306
|
#
|
257
|
-
# You should call it to verify whether your request (such as
|
307
|
+
# You should call it to verify whether your request (such as {#query}) was
|
258
308
|
# completed but with warnings. For instance, search query against a distributed
|
259
309
|
# index might complete succesfully even if several remote agents timed out.
|
260
310
|
# In that case, a warning message would be produced.
|
261
|
-
#
|
311
|
+
#
|
262
312
|
# The warning message is not reset by this call; so you can safely call it
|
263
313
|
# several times if needed.
|
264
314
|
#
|
265
|
-
|
315
|
+
# @return [String] last warning message.
|
316
|
+
#
|
317
|
+
# @example
|
318
|
+
# puts sphinx.last_warning
|
319
|
+
#
|
320
|
+
# @see #last_error
|
321
|
+
# @see #connect_error?
|
322
|
+
#
|
323
|
+
def last_warning
|
266
324
|
@warning
|
267
325
|
end
|
268
|
-
|
326
|
+
alias :GetLastWarning :last_warning
|
327
|
+
|
269
328
|
# Checks whether the last error was a network error on API side, or a
|
270
329
|
# remote error reported by searchd. Returns true if the last connection
|
271
330
|
# attempt to searchd failed on API side, false otherwise (if the error
|
272
331
|
# was remote, or there were no connection attempts at all).
|
273
332
|
#
|
274
|
-
|
333
|
+
# @return [Boolean] the value indicating whether last error was a
|
334
|
+
# nework error on API side.
|
335
|
+
#
|
336
|
+
# @example
|
337
|
+
# puts "Connection failed!" if sphinx.connect_error?
|
338
|
+
#
|
339
|
+
# @see #last_error
|
340
|
+
# @see #last_warning
|
341
|
+
#
|
342
|
+
def connect_error?
|
275
343
|
@connerror || false
|
276
344
|
end
|
277
|
-
|
345
|
+
alias :IsConnectError :connect_error?
|
346
|
+
|
278
347
|
# Sets searchd host name and TCP port. All subsequent requests will
|
279
348
|
# use the new host and port settings. Default +host+ and +port+ are
|
280
|
-
# 'localhost' and
|
349
|
+
# 'localhost' and 9312, respectively.
|
281
350
|
#
|
282
351
|
# Also, you can specify an absolute path to Sphinx's UNIX socket as +host+,
|
283
352
|
# in this case pass port as +0+ or +nil+.
|
284
353
|
#
|
285
|
-
|
354
|
+
# @param [String] host the searchd host name or UNIX socket absolute path.
|
355
|
+
# @param [Integer] port the searchd port name (could be any if UNIX
|
356
|
+
# socket path specified).
|
357
|
+
# @return [Sphinx::Client] self.
|
358
|
+
#
|
359
|
+
# @example
|
360
|
+
# sphinx.set_server('localhost', 9312)
|
361
|
+
# sphinx.set_server('/opt/sphinx/var/run/sphinx.sock')
|
362
|
+
#
|
363
|
+
# @raise [ArgumentError] Occurred when parameters are invalid.
|
364
|
+
# @see #set_servers
|
365
|
+
# @see #set_connect_timeout
|
366
|
+
# @see #set_request_timeout
|
367
|
+
#
|
368
|
+
def set_server(host, port = 9312)
|
286
369
|
raise ArgumentError, '"host" argument must be String' unless host.kind_of?(String)
|
287
|
-
|
370
|
+
|
288
371
|
path = nil
|
289
372
|
# Check if UNIX socket should be used
|
290
373
|
if host[0] == ?/
|
@@ -298,25 +381,47 @@ module Sphinx
|
|
298
381
|
host = port = nil unless path.nil?
|
299
382
|
|
300
383
|
@servers = [Sphinx::Server.new(self, host, port, path)].freeze
|
384
|
+
self
|
301
385
|
end
|
386
|
+
alias :SetServer :set_server
|
302
387
|
|
303
388
|
# Sets the list of searchd servers. Each subsequent request will use next
|
304
389
|
# server in list (round-robin). In case of one server failure, request could
|
305
|
-
# be retried on another server (see
|
306
|
-
#
|
307
|
-
#
|
308
|
-
#
|
309
|
-
# to
|
310
|
-
#
|
311
|
-
|
390
|
+
# be retried on another server (see {#set_connect_timeout} and
|
391
|
+
# {#set_request_timeout}).
|
392
|
+
#
|
393
|
+
# Method accepts an +Array+ of +Hash+es, each of them should have <tt>:host</tt>
|
394
|
+
# and <tt>:port</tt> (to connect to searchd through network) or <tt>:path</tt>
|
395
|
+
# (an absolute path to UNIX socket) specified.
|
396
|
+
#
|
397
|
+
# @param [Array<Hash>] servers an +Array+ of +Hash+ objects with servers parameters.
|
398
|
+
# @option servers [String] :host the searchd host name or UNIX socket absolute path.
|
399
|
+
# @option servers [String] :path the searchd UNIX socket absolute path.
|
400
|
+
# @option servers [Integer] :port (9312) the searchd port name (skiped when UNIX
|
401
|
+
# socket path specified)
|
402
|
+
# @return [Sphinx::Client] self.
|
403
|
+
#
|
404
|
+
# @example
|
405
|
+
# sphinx.set_servers([
|
406
|
+
# { :host => 'browse01.local' }, # default port is 9312
|
407
|
+
# { :host => 'browse02.local', :port => 9312 },
|
408
|
+
# { :path => '/opt/sphinx/var/run/sphinx.sock' }
|
409
|
+
# ])
|
410
|
+
#
|
411
|
+
# @raise [ArgumentError] Occurred when parameters are invalid.
|
412
|
+
# @see #set_server
|
413
|
+
# @see #set_connect_timeout
|
414
|
+
# @see #set_request_timeout
|
415
|
+
#
|
416
|
+
def set_servers(servers)
|
312
417
|
raise ArgumentError, '"servers" argument must be Array' unless servers.kind_of?(Array)
|
313
418
|
raise ArgumentError, '"servers" argument must be not empty' if servers.empty?
|
314
|
-
|
419
|
+
|
315
420
|
@servers = servers.map do |server|
|
316
421
|
raise ArgumentError, '"servers" argument must be Array of Hashes' unless server.kind_of?(Hash)
|
317
422
|
|
318
423
|
host = server[:path] || server['path'] || server[:host] || server['host']
|
319
|
-
port = server[:port] || server['port']
|
424
|
+
port = server[:port] || server['port'] || 9312
|
320
425
|
path = nil
|
321
426
|
raise ArgumentError, '"host" argument must be String' unless host.kind_of?(String)
|
322
427
|
|
@@ -330,11 +435,13 @@ module Sphinx
|
|
330
435
|
end
|
331
436
|
|
332
437
|
host = port = nil unless path.nil?
|
333
|
-
|
438
|
+
|
334
439
|
Sphinx::Server.new(self, host, port, path)
|
335
440
|
end.freeze
|
441
|
+
self
|
336
442
|
end
|
337
|
-
|
443
|
+
alias :SetServers :set_servers
|
444
|
+
|
338
445
|
# Sets the time allowed to spend connecting to the server before giving up
|
339
446
|
# and number of retries to perform.
|
340
447
|
#
|
@@ -342,7 +449,7 @@ module Sphinx
|
|
342
449
|
# be returned back to the application in order for application-level error
|
343
450
|
# handling to advise the user.
|
344
451
|
#
|
345
|
-
# When multiple servers configured through
|
452
|
+
# When multiple servers configured through {#set_servers} method, and +retries+
|
346
453
|
# number is greater than 1, library will try to connect to another server.
|
347
454
|
# In case of single server configured, it will try to reconnect +retries+
|
348
455
|
# times.
|
@@ -350,15 +457,29 @@ module Sphinx
|
|
350
457
|
# Please note, this timeout will only be used for connection establishing, not
|
351
458
|
# for regular API requests.
|
352
459
|
#
|
353
|
-
|
460
|
+
# @param [Integer] timeout a connection timeout in seconds.
|
461
|
+
# @param [Integer] retries number of connect retries.
|
462
|
+
# @return [Sphinx::Client] self.
|
463
|
+
#
|
464
|
+
# @example Set connection timeout to 1 second and number of retries to 5
|
465
|
+
# sphinx.set_connect_timeout(1, 5)
|
466
|
+
#
|
467
|
+
# @raise [ArgumentError] Occurred when parameters are invalid.
|
468
|
+
# @see #set_server
|
469
|
+
# @see #set_servers
|
470
|
+
# @see #set_request_timeout
|
471
|
+
#
|
472
|
+
def set_connect_timeout(timeout, retries = 1)
|
354
473
|
raise ArgumentError, '"timeout" argument must be Integer' unless timeout.respond_to?(:integer?) and timeout.integer?
|
355
474
|
raise ArgumentError, '"retries" argument must be Integer' unless retries.respond_to?(:integer?) and retries.integer?
|
356
475
|
raise ArgumentError, '"retries" argument must be greater than 0' unless retries > 0
|
357
|
-
|
476
|
+
|
358
477
|
@timeout = timeout
|
359
478
|
@retries = retries
|
479
|
+
self
|
360
480
|
end
|
361
|
-
|
481
|
+
alias :SetConnectTimeout :set_connect_timeout
|
482
|
+
|
362
483
|
# Sets the time allowed to spend performing request to the server before giving up
|
363
484
|
# and number of retries to perform.
|
364
485
|
#
|
@@ -366,34 +487,82 @@ module Sphinx
|
|
366
487
|
# be returned back to the application in order for application-level error
|
367
488
|
# handling to advise the user.
|
368
489
|
#
|
369
|
-
# When multiple servers configured through
|
490
|
+
# When multiple servers configured through {#set_servers} method, and +retries+
|
370
491
|
# number is greater than 1, library will try to do another try with this server
|
371
492
|
# (with full reconnect). If connection would fail, behavior depends on
|
372
|
-
#
|
493
|
+
# {#set_connect_timeout} settings.
|
373
494
|
#
|
374
495
|
# Please note, this timeout will only be used for request performing, not
|
375
496
|
# for connection establishing.
|
376
497
|
#
|
377
|
-
|
498
|
+
# @param [Integer] timeout a request timeout in seconds.
|
499
|
+
# @param [Integer] retries number of request retries.
|
500
|
+
# @return [Sphinx::Client] self.
|
501
|
+
#
|
502
|
+
# @example Set request timeout to 1 second and number of retries to 5
|
503
|
+
# sphinx.set_request_timeout(1, 5)
|
504
|
+
#
|
505
|
+
# @raise [ArgumentError] Occurred when parameters are invalid.
|
506
|
+
# @see #set_server
|
507
|
+
# @see #set_servers
|
508
|
+
# @see #set_connect_timeout
|
509
|
+
#
|
510
|
+
def set_request_timeout(timeout, retries = 1)
|
378
511
|
raise ArgumentError, '"timeout" argument must be Integer' unless timeout.respond_to?(:integer?) and timeout.integer?
|
379
512
|
raise ArgumentError, '"retries" argument must be Integer' unless retries.respond_to?(:integer?) and retries.integer?
|
380
513
|
raise ArgumentError, '"retries" argument must be greater than 0' unless retries > 0
|
381
|
-
|
514
|
+
|
382
515
|
@reqtimeout = timeout
|
383
516
|
@reqretries = retries
|
517
|
+
self
|
518
|
+
end
|
519
|
+
alias :SetRequestTimeout :set_request_timeout
|
520
|
+
|
521
|
+
# Sets distributed retry count and delay.
|
522
|
+
#
|
523
|
+
# On temporary failures searchd will attempt up to +count+ retries
|
524
|
+
# per agent. +delay+ is the delay between the retries, in milliseconds.
|
525
|
+
# Retries are disabled by default. Note that this call will not make
|
526
|
+
# the API itself retry on temporary failure; it only tells searchd
|
527
|
+
# to do so. Currently, the list of temporary failures includes all
|
528
|
+
# kinds of connection failures and maxed out (too busy) remote agents.
|
529
|
+
#
|
530
|
+
# @param [Integer] count a number of retries to perform.
|
531
|
+
# @param [Integer] delay a delay between the retries.
|
532
|
+
# @return [Sphinx::Client] self.
|
533
|
+
#
|
534
|
+
# @example Perform 5 retries with 200 ms between them
|
535
|
+
# sphinx.set_retries(5, 200)
|
536
|
+
#
|
537
|
+
# @raise [ArgumentError] Occurred when parameters are invalid.
|
538
|
+
# @see #set_connect_timeout
|
539
|
+
# @see #set_request_timeout
|
540
|
+
#
|
541
|
+
def set_retries(count, delay = 0)
|
542
|
+
raise ArgumentError, '"count" argument must be Integer' unless count.respond_to?(:integer?) and count.integer?
|
543
|
+
raise ArgumentError, '"delay" argument must be Integer' unless delay.respond_to?(:integer?) and delay.integer?
|
544
|
+
|
545
|
+
@retrycount = count
|
546
|
+
@retrydelay = delay
|
547
|
+
self
|
384
548
|
end
|
385
|
-
|
549
|
+
alias :SetRetries :set_retries
|
550
|
+
|
551
|
+
#=================================================================
|
552
|
+
# General query settings
|
553
|
+
#=================================================================
|
554
|
+
|
386
555
|
# Sets offset into server-side result set (+offset+) and amount of matches to
|
387
556
|
# return to client starting from that offset (+limit+). Can additionally control
|
388
557
|
# maximum server-side result set size for current query (+max_matches+) and the
|
389
558
|
# threshold amount of matches to stop searching at (+cutoff+). All parameters
|
390
559
|
# must be non-negative integers.
|
391
560
|
#
|
392
|
-
# First two parameters to
|
561
|
+
# First two parameters to {#set_limits} are identical in behavior to MySQL LIMIT
|
393
562
|
# clause. They instruct searchd to return at most +limit+ matches starting from
|
394
563
|
# match number +offset+. The default offset and limit settings are +0+ and +20+,
|
395
564
|
# that is, to return first +20+ matches.
|
396
|
-
#
|
565
|
+
#
|
397
566
|
# +max_matches+ setting controls how much matches searchd will keep in RAM
|
398
567
|
# while searching. All matching documents will be normally processed, ranked,
|
399
568
|
# filtered, and sorted even if max_matches is set to +1+. But only best +N+
|
@@ -415,12 +584,23 @@ module Sphinx
|
|
415
584
|
# searchd to forcibly stop search query once $cutoff matches had been found
|
416
585
|
# and processed.
|
417
586
|
#
|
418
|
-
|
587
|
+
# @param [Integer] offset an offset into server-side result set.
|
588
|
+
# @param [Integer] limit an amount of matches to return.
|
589
|
+
# @param [Integer] max a maximum server-side result set size.
|
590
|
+
# @param [Integer] cutoff a threshold amount of matches to stop searching at.
|
591
|
+
# @return [Sphinx::Client] self.
|
592
|
+
#
|
593
|
+
# @example
|
594
|
+
# sphinx.set_limits(100, 50, 1000, 5000)
|
595
|
+
#
|
596
|
+
# @raise [ArgumentError] Occurred when parameters are invalid.
|
597
|
+
#
|
598
|
+
def set_limits(offset, limit, max = 0, cutoff = 0)
|
419
599
|
raise ArgumentError, '"offset" argument must be Integer' unless offset.respond_to?(:integer?) and offset.integer?
|
420
600
|
raise ArgumentError, '"limit" argument must be Integer' unless limit.respond_to?(:integer?) and limit.integer?
|
421
601
|
raise ArgumentError, '"max" argument must be Integer' unless max.respond_to?(:integer?) and max.integer?
|
422
602
|
raise ArgumentError, '"cutoff" argument must be Integer' unless cutoff.respond_to?(:integer?) and cutoff.integer?
|
423
|
-
|
603
|
+
|
424
604
|
raise ArgumentError, '"offset" argument should be greater or equal to zero' unless offset >= 0
|
425
605
|
raise ArgumentError, '"limit" argument should be greater to zero' unless limit > 0
|
426
606
|
raise ArgumentError, '"max" argument should be greater or equal to zero' unless max >= 0
|
@@ -430,35 +610,176 @@ module Sphinx
|
|
430
610
|
@limit = limit
|
431
611
|
@maxmatches = max if max > 0
|
432
612
|
@cutoff = cutoff if cutoff > 0
|
613
|
+
self
|
433
614
|
end
|
434
|
-
|
615
|
+
alias :SetLimits :set_limits
|
616
|
+
|
435
617
|
# Sets maximum search query time, in milliseconds. Parameter must be a
|
436
618
|
# non-negative integer. Default valus is +0+ which means "do not limit".
|
437
619
|
#
|
438
|
-
# Similar to +cutoff+ setting from
|
620
|
+
# Similar to +cutoff+ setting from {#set_limits}, but limits elapsed query
|
439
621
|
# time instead of processed matches count. Local search queries will be
|
440
622
|
# stopped once that much time has elapsed. Note that if you're performing
|
441
623
|
# a search which queries several local indexes, this limit applies to each
|
442
624
|
# index separately.
|
443
625
|
#
|
444
|
-
|
626
|
+
# @param [Integer] max maximum search query time in milliseconds.
|
627
|
+
# @return [Sphinx::Client] self.
|
628
|
+
#
|
629
|
+
# @example
|
630
|
+
# sphinx.set_max_query_time(200)
|
631
|
+
#
|
632
|
+
# @raise [ArgumentError] Occurred when parameters are invalid.
|
633
|
+
#
|
634
|
+
def set_max_query_time(max)
|
445
635
|
raise ArgumentError, '"max" argument must be Integer' unless max.respond_to?(:integer?) and max.integer?
|
446
636
|
raise ArgumentError, '"max" argument should be greater or equal to zero' unless max >= 0
|
447
637
|
|
448
638
|
@maxquerytime = max
|
639
|
+
self
|
449
640
|
end
|
450
|
-
|
641
|
+
alias :SetMaxQueryTime :set_max_query_time
|
642
|
+
|
643
|
+
# Sets temporary (per-query) per-document attribute value overrides. Only
|
644
|
+
# supports scalar attributes. +values+ must be a +Hash+ that maps document
|
645
|
+
# IDs to overridden attribute values.
|
646
|
+
#
|
647
|
+
# Override feature lets you "temporary" update attribute values for some
|
648
|
+
# documents within a single query, leaving all other queries unaffected.
|
649
|
+
# This might be useful for personalized data. For example, assume you're
|
650
|
+
# implementing a personalized search function that wants to boost the posts
|
651
|
+
# that the user's friends recommend. Such data is not just dynamic, but
|
652
|
+
# also personal; so you can't simply put it in the index because you don't
|
653
|
+
# want everyone's searches affected. Overrides, on the other hand, are local
|
654
|
+
# to a single query and invisible to everyone else. So you can, say, setup
|
655
|
+
# a "friends_weight" value for every document, defaulting to 0, then
|
656
|
+
# temporary override it with 1 for documents 123, 456 and 789 (recommended
|
657
|
+
# by exactly the friends of current user), and use that value when ranking.
|
658
|
+
#
|
659
|
+
# You can specify attribute type as String ("integer", "float", etc),
|
660
|
+
# Symbol (:integer, :float, etc), or
|
661
|
+
# Fixnum constant (SPH_ATTR_INTEGER, SPH_ATTR_FLOAT, etc).
|
662
|
+
#
|
663
|
+
# @param [String, Symbol] attribute an attribute name to override values of.
|
664
|
+
# @param [Integer, String, Symbol] attrtype attribute type.
|
665
|
+
# @param [Hash] values a +Hash+ that maps document IDs to overridden attribute values.
|
666
|
+
# @return [Sphinx::Client] self.
|
667
|
+
#
|
668
|
+
# @example
|
669
|
+
# sphinx.set_override(:friends_weight, :integer, {123 => 1, 456 => 1, 789 => 1})
|
670
|
+
#
|
671
|
+
# @raise [ArgumentError] Occurred when parameters are invalid.
|
672
|
+
#
|
673
|
+
# @see http://www.sphinxsearch.com/docs/current.html#api-func-setoverride Section 6.2.3, "SetOverride"
|
674
|
+
#
|
675
|
+
def set_override(attribute, attrtype, values)
|
676
|
+
raise ArgumentError, '"attribute" argument must be String or Symbol' unless attribute.kind_of?(String) or attribute.kind_of?(Symbol)
|
677
|
+
|
678
|
+
case attrtype
|
679
|
+
when String, Symbol
|
680
|
+
begin
|
681
|
+
attrtype = self.class.const_get("SPH_ATTR_#{attrtype.to_s.upcase}")
|
682
|
+
rescue NameError
|
683
|
+
raise ArgumentError, "\"attrtype\" argument value \"#{attrtype}\" is invalid"
|
684
|
+
end
|
685
|
+
when Fixnum
|
686
|
+
raise ArgumentError, "\"attrtype\" argument value \"#{attrtype}\" is invalid" unless (SPH_ATTR_INTEGER..SPH_ATTR_BIGINT).include?(attrtype)
|
687
|
+
else
|
688
|
+
raise ArgumentError, '"attrtype" argument must be Fixnum, String, or Symbol'
|
689
|
+
end
|
690
|
+
|
691
|
+
raise ArgumentError, '"values" argument must be Hash' unless values.kind_of?(Hash)
|
692
|
+
|
693
|
+
values.each do |id, value|
|
694
|
+
raise ArgumentError, '"values" argument must be Hash map of Integer to Integer or Time' unless id.respond_to?(:integer?) and id.integer?
|
695
|
+
case attrtype
|
696
|
+
when SPH_ATTR_TIMESTAMP
|
697
|
+
raise ArgumentError, '"values" argument must be Hash map of Integer to Integer or Time' unless (value.respond_to?(:integer?) and value.integer?) or value.kind_of?(Time)
|
698
|
+
when SPH_ATTR_FLOAT
|
699
|
+
raise ArgumentError, '"values" argument must be Hash map of Integer to Float or Integer' unless value.kind_of?(Float) or (value.respond_to?(:integer?) and value.integer?)
|
700
|
+
else
|
701
|
+
# SPH_ATTR_INTEGER, SPH_ATTR_ORDINAL, SPH_ATTR_BOOL, SPH_ATTR_BIGINT
|
702
|
+
raise ArgumentError, '"values" argument must be Hash map of Integer to Integer' unless value.respond_to?(:integer?) and value.integer?
|
703
|
+
end
|
704
|
+
end
|
705
|
+
|
706
|
+
@overrides << { 'attr' => attribute.to_s, 'type' => attrtype, 'values' => values }
|
707
|
+
self
|
708
|
+
end
|
709
|
+
alias :SetOverride :set_override
|
710
|
+
|
711
|
+
# Sets the select clause, listing specific attributes to fetch, and
|
712
|
+
# expressions to compute and fetch. Clause syntax mimics SQL.
|
713
|
+
#
|
714
|
+
# {#set_select} is very similar to the part of a typical SQL query between
|
715
|
+
# +SELECT+ and +FROM+. It lets you choose what attributes (columns) to
|
716
|
+
# fetch, and also what expressions over the columns to compute and fetch.
|
717
|
+
# A certain difference from SQL is that expressions must always be aliased
|
718
|
+
# to a correct identifier (consisting of letters and digits) using +AS+
|
719
|
+
# keyword. SQL also lets you do that but does not require to. Sphinx enforces
|
720
|
+
# aliases so that the computation results can always be returned under a
|
721
|
+
# "normal" name in the result set, used in other clauses, etc.
|
722
|
+
#
|
723
|
+
# Everything else is basically identical to SQL. Star ('*') is supported.
|
724
|
+
# Functions are supported. Arbitrary amount of expressions is supported.
|
725
|
+
# Computed expressions can be used for sorting, filtering, and grouping,
|
726
|
+
# just as the regular attributes.
|
727
|
+
#
|
728
|
+
# Starting with version 0.9.9-rc2, aggregate functions (<tt>AVG()</tt>,
|
729
|
+
# <tt>MIN()</tt>, <tt>MAX()</tt>, <tt>SUM()</tt>) are supported when using
|
730
|
+
# <tt>GROUP BY</tt>.
|
731
|
+
#
|
732
|
+
# Expression sorting (Section 4.5, “SPH_SORT_EXPR mode”) and geodistance
|
733
|
+
# functions ({#set_geo_anchor}) are now internally implemented
|
734
|
+
# using this computed expressions mechanism, using magic names '<tt>@expr</tt>'
|
735
|
+
# and '<tt>@geodist</tt>' respectively.
|
736
|
+
#
|
737
|
+
# @param [String] select a select clause, listing specific attributes to fetch.
|
738
|
+
# @return [Sphinx::Client] self.
|
739
|
+
#
|
740
|
+
# @example
|
741
|
+
# sphinx.set_select('*, @weight+(user_karma+ln(pageviews))*0.1 AS myweight')
|
742
|
+
# sphinx.set_select('exp_years, salary_gbp*{$gbp_usd_rate} AS salary_usd, IF(age>40,1,0) AS over40')
|
743
|
+
# sphinx.set_select('*, AVG(price) AS avgprice')
|
744
|
+
#
|
745
|
+
# @raise [ArgumentError] Occurred when parameters are invalid.
|
746
|
+
#
|
747
|
+
# @see http://www.sphinxsearch.com/docs/current.html#sort-expr Section 4.5, "SPH_SORT_EXPR mode"
|
748
|
+
# @see http://www.sphinxsearch.com/docs/current.html#api-func-setgeoanchor Section 6.4.5, "SetGeoAnchor"
|
749
|
+
# @see http://www.sphinxsearch.com/docs/current.html#api-func-setselect Section 6.2.4, "SetSelect"
|
750
|
+
#
|
751
|
+
def set_select(select)
|
752
|
+
raise ArgumentError, '"select" argument must be String' unless select.kind_of?(String)
|
753
|
+
|
754
|
+
@select = select
|
755
|
+
self
|
756
|
+
end
|
757
|
+
alias :SetSelect :set_select
|
758
|
+
|
759
|
+
#=================================================================
|
760
|
+
# Full-text search query settings
|
761
|
+
#=================================================================
|
762
|
+
|
451
763
|
# Sets full-text query matching mode.
|
452
764
|
#
|
453
765
|
# Parameter must be a +Fixnum+ constant specifying one of the known modes
|
454
766
|
# (+SPH_MATCH_ALL+, +SPH_MATCH_ANY+, etc), +String+ with identifier (<tt>"all"</tt>,
|
455
767
|
# <tt>"any"</tt>, etc), or a +Symbol+ (<tt>:all</tt>, <tt>:any</tt>, etc).
|
456
768
|
#
|
457
|
-
#
|
458
|
-
#
|
459
|
-
# * {Section 6.3.1, "SetMatchMode"}[http://www.sphinxsearch.com/docs/current.html#api-func-setmatchmode] for details.
|
769
|
+
# @param [Integer, String, Symbol] mode full-text query matching mode.
|
770
|
+
# @return [Sphinx::Client] self.
|
460
771
|
#
|
461
|
-
|
772
|
+
# @example
|
773
|
+
# sphinx.set_match_mode(Sphinx::Client::SPH_MATCH_ALL)
|
774
|
+
# sphinx.set_match_mode(:all)
|
775
|
+
# sphinx.set_match_mode('all')
|
776
|
+
#
|
777
|
+
# @raise [ArgumentError] Occurred when parameters are invalid.
|
778
|
+
#
|
779
|
+
# @see http://www.sphinxsearch.com/docs/current.html#matching-modes Section 4.1, "Matching modes"
|
780
|
+
# @see http://www.sphinxsearch.com/docs/current.html#api-func-setmatchmode Section 6.3.1, "SetMatchMode"
|
781
|
+
#
|
782
|
+
def set_match_mode(mode)
|
462
783
|
case mode
|
463
784
|
when String, Symbol
|
464
785
|
begin
|
@@ -473,14 +794,33 @@ module Sphinx
|
|
473
794
|
end
|
474
795
|
|
475
796
|
@mode = mode
|
797
|
+
self
|
476
798
|
end
|
477
|
-
|
478
|
-
|
799
|
+
alias :SetMatchMode :set_match_mode
|
800
|
+
|
801
|
+
# Sets ranking mode. Only available in +SPH_MATCH_EXTENDED2+
|
802
|
+
# matching mode at the time of this writing. Parameter must be a
|
803
|
+
# constant specifying one of the known modes.
|
479
804
|
#
|
480
805
|
# You can specify ranking mode as String ("proximity_bm25", "bm25", etc),
|
481
806
|
# Symbol (:proximity_bm25, :bm25, etc), or
|
482
807
|
# Fixnum constant (SPH_RANK_PROXIMITY_BM25, SPH_RANK_BM25, etc).
|
483
|
-
|
808
|
+
#
|
809
|
+
# @param [Integer, String, Symbol] ranker ranking mode.
|
810
|
+
# @return [Sphinx::Client] self.
|
811
|
+
#
|
812
|
+
# @example
|
813
|
+
# sphinx.set_ranking_mode(Sphinx::Client::SPH_RANK_BM25)
|
814
|
+
# sphinx.set_ranking_mode(:bm25)
|
815
|
+
# sphinx.set_ranking_mode('bm25')
|
816
|
+
#
|
817
|
+
# @raise [ArgumentError] Occurred when parameters are invalid.
|
818
|
+
#
|
819
|
+
# @see http://www.sphinxsearch.com/docs/current.html#matching-modes Section 4.1, "Matching modes"
|
820
|
+
# @see http://www.sphinxsearch.com/docs/current.html#api-func-setmatchmode Section 6.3.1, "SetMatchMode"
|
821
|
+
# @see http://www.sphinxsearch.com/docs/current.html#api-func-setrankingmode Section 6.3.2, "SetRankingMode"
|
822
|
+
#
|
823
|
+
def set_ranking_mode(ranker)
|
484
824
|
case ranker
|
485
825
|
when String, Symbol
|
486
826
|
begin
|
@@ -495,14 +835,33 @@ module Sphinx
|
|
495
835
|
end
|
496
836
|
|
497
837
|
@ranker = ranker
|
838
|
+
self
|
498
839
|
end
|
499
|
-
|
840
|
+
alias :SetRankingMode :set_ranking_mode
|
841
|
+
|
500
842
|
# Set matches sorting mode.
|
501
843
|
#
|
502
844
|
# You can specify sorting mode as String ("relevance", "attr_desc", etc),
|
503
845
|
# Symbol (:relevance, :attr_desc, etc), or
|
504
846
|
# Fixnum constant (SPH_SORT_RELEVANCE, SPH_SORT_ATTR_DESC, etc).
|
505
|
-
|
847
|
+
#
|
848
|
+
# @param [Integer, String, Symbol] mode matches sorting mode.
|
849
|
+
# @param [String] sortby sorting clause, with the syntax depending on
|
850
|
+
# specific mode. Should be specified unless sorting mode is
|
851
|
+
# +SPH_SORT_RELEVANCE+.
|
852
|
+
# @return [Sphinx::Client] self.
|
853
|
+
#
|
854
|
+
# @example
|
855
|
+
# sphinx.set_sort_mode(Sphinx::Client::SPH_SORT_ATTR_ASC, 'attr')
|
856
|
+
# sphinx.set_sort_mode(:attr_asc, 'attr')
|
857
|
+
# sphinx.set_sort_mode('attr_asc', 'attr')
|
858
|
+
#
|
859
|
+
# @raise [ArgumentError] Occurred when parameters are invalid.
|
860
|
+
#
|
861
|
+
# @see http://www.sphinxsearch.com/docs/current.html#sorting-modes Section 4.5, "Sorting modes"
|
862
|
+
# @see http://www.sphinxsearch.com/docs/current.html#api-func-setsortmode Section 6.3.3, "SetSortMode"
|
863
|
+
#
|
864
|
+
def set_sort_mode(mode, sortby = '')
|
506
865
|
case mode
|
507
866
|
when String, Symbol
|
508
867
|
begin
|
@@ -521,27 +880,69 @@ module Sphinx
|
|
521
880
|
|
522
881
|
@sort = mode
|
523
882
|
@sortby = sortby
|
883
|
+
self
|
524
884
|
end
|
525
|
-
|
526
|
-
|
885
|
+
alias :SetSortMode :set_sort_mode
|
886
|
+
|
887
|
+
# Binds per-field weights in the order of appearance in the index.
|
888
|
+
#
|
889
|
+
# @param [Array<Integer>] weights an +Array+ of integer per-field weights.
|
890
|
+
# @return [Sphinx::Client] self.
|
527
891
|
#
|
528
|
-
#
|
529
|
-
|
892
|
+
# @example
|
893
|
+
# sphinx.set_weights([1, 3, 5])
|
894
|
+
#
|
895
|
+
# @raise [ArgumentError] Occurred when parameters are invalid.
|
896
|
+
#
|
897
|
+
# @deprecated Use {#set_field_weights} instead.
|
898
|
+
# @see #set_field_weights
|
899
|
+
#
|
900
|
+
def set_weights(weights)
|
530
901
|
raise ArgumentError, '"weights" argument must be Array' unless weights.kind_of?(Array)
|
531
902
|
weights.each do |weight|
|
532
903
|
raise ArgumentError, '"weights" argument must be Array of integers' unless weight.respond_to?(:integer?) and weight.integer?
|
533
904
|
end
|
534
905
|
|
535
906
|
@weights = weights
|
907
|
+
self
|
536
908
|
end
|
909
|
+
alias :SetWeights :set_weights
|
537
910
|
|
538
|
-
#
|
911
|
+
# Binds per-field weights by name. Parameter must be a +Hash+
|
912
|
+
# mapping string field names to integer weights.
|
913
|
+
#
|
914
|
+
# Match ranking can be affected by per-field weights. For instance,
|
915
|
+
# see Section 4.4, "Weighting" for an explanation how phrase
|
916
|
+
# proximity ranking is affected. This call lets you specify what
|
917
|
+
# non-default weights to assign to different full-text fields.
|
918
|
+
#
|
919
|
+
# The weights must be positive 32-bit integers. The final weight
|
920
|
+
# will be a 32-bit integer too. Default weight value is 1. Unknown
|
921
|
+
# field names will be silently ignored.
|
922
|
+
#
|
923
|
+
# There is no enforced limit on the maximum weight value at the
|
924
|
+
# moment. However, beware that if you set it too high you can
|
925
|
+
# start hitting 32-bit wraparound issues. For instance, if
|
926
|
+
# you set a weight of 10,000,000 and search in extended mode,
|
927
|
+
# then maximum possible weight will be equal to 10 million (your
|
928
|
+
# weight) by 1 thousand (internal BM25 scaling factor, see
|
929
|
+
# Section 4.4, “Weighting”) by 1 or more (phrase proximity rank).
|
930
|
+
# The result is at least 10 billion that does not fit in 32 bits
|
931
|
+
# and will be wrapped around, producing unexpected results.
|
932
|
+
#
|
933
|
+
# @param [Hash] weights a +Hash+ mapping string field names to
|
934
|
+
# integer weights.
|
935
|
+
# @return [Sphinx::Client] self.
|
936
|
+
#
|
937
|
+
# @example
|
938
|
+
# sphinx.set_field_weights(:title => 20, :text => 10)
|
539
939
|
#
|
540
|
-
#
|
541
|
-
#
|
542
|
-
#
|
543
|
-
#
|
544
|
-
|
940
|
+
# @raise [ArgumentError] Occurred when parameters are invalid.
|
941
|
+
#
|
942
|
+
# @see http://www.sphinxsearch.com/docs/current.html#weighting Section 4.4, "Weighting"
|
943
|
+
# @see http://www.sphinxsearch.com/docs/current.html#api-func-setfieldweights Section 6.3.5, "SetFieldWeights"
|
944
|
+
#
|
945
|
+
def set_field_weights(weights)
|
545
946
|
raise ArgumentError, '"weights" argument must be Hash' unless weights.kind_of?(Hash)
|
546
947
|
weights.each do |name, weight|
|
547
948
|
unless (name.kind_of?(String) or name.kind_of?(Symbol)) and (weight.respond_to?(:integer?) and weight.integer?)
|
@@ -550,37 +951,119 @@ module Sphinx
|
|
550
951
|
end
|
551
952
|
|
552
953
|
@fieldweights = weights
|
954
|
+
self
|
553
955
|
end
|
554
|
-
|
555
|
-
|
556
|
-
|
956
|
+
alias :SetFieldWeights :set_field_weights
|
957
|
+
|
958
|
+
# Sets per-index weights, and enables weighted summing of match
|
959
|
+
# weights across different indexes. Parameter must be a hash
|
960
|
+
# (associative array) mapping string index names to integer
|
961
|
+
# weights. Default is empty array that means to disable weighting
|
962
|
+
# summing.
|
963
|
+
#
|
964
|
+
# When a match with the same document ID is found in several
|
965
|
+
# different local indexes, by default Sphinx simply chooses the
|
966
|
+
# match from the index specified last in the query. This is to
|
967
|
+
# support searching through partially overlapping index partitions.
|
968
|
+
#
|
969
|
+
# However in some cases the indexes are not just partitions,
|
970
|
+
# and you might want to sum the weights across the indexes
|
971
|
+
# instead of picking one. {#set_index_weights} lets you do that.
|
972
|
+
# With summing enabled, final match weight in result set will be
|
973
|
+
# computed as a sum of match weight coming from the given index
|
974
|
+
# multiplied by respective per-index weight specified in this
|
975
|
+
# call. Ie. if the document 123 is found in index A with the
|
976
|
+
# weight of 2, and also in index B with the weight of 3, and
|
977
|
+
# you called {#set_index_weights} with <tt>{"A"=>100, "B"=>10}</tt>,
|
978
|
+
# the final weight return to the client will be 2*100+3*10 = 230.
|
979
|
+
#
|
980
|
+
# @param [Hash] weights a +Hash+ mapping string index names to
|
981
|
+
# integer weights.
|
982
|
+
# @return [Sphinx::Client] self.
|
983
|
+
#
|
984
|
+
# @example
|
985
|
+
# sphinx.set_field_weights(:fresh => 20, :archived => 10)
|
986
|
+
#
|
987
|
+
# @raise [ArgumentError] Occurred when parameters are invalid.
|
988
|
+
#
|
989
|
+
# @see http://www.sphinxsearch.com/docs/current.html#api-func-setindexweights Section 6.3.6, "SetIndexWeights"
|
990
|
+
#
|
991
|
+
def set_index_weights(weights)
|
557
992
|
raise ArgumentError, '"weights" argument must be Hash' unless weights.kind_of?(Hash)
|
558
993
|
weights.each do |index, weight|
|
559
994
|
unless (index.kind_of?(String) or index.kind_of?(Symbol)) and (weight.respond_to?(:integer?) and weight.integer?)
|
560
995
|
raise ArgumentError, '"weights" argument must be Hash map of strings to integers'
|
561
996
|
end
|
562
997
|
end
|
563
|
-
|
998
|
+
|
564
999
|
@indexweights = weights
|
1000
|
+
self
|
565
1001
|
end
|
1002
|
+
alias :SetIndexWeights :set_index_weights
|
566
1003
|
|
567
|
-
|
568
|
-
#
|
569
|
-
|
570
|
-
|
1004
|
+
#=================================================================
|
1005
|
+
# Result set filtering settings
|
1006
|
+
#=================================================================
|
1007
|
+
|
1008
|
+
# Sets an accepted range of document IDs. Parameters must be integers.
|
1009
|
+
# Defaults are 0 and 0; that combination means to not limit by range.
|
1010
|
+
#
|
1011
|
+
# After this call, only those records that have document ID between
|
1012
|
+
# +min+ and +max+ (including IDs exactly equal to +min+ or +max+)
|
1013
|
+
# will be matched.
|
1014
|
+
#
|
1015
|
+
# @param [Integer] min min document ID.
|
1016
|
+
# @param [Integer] min max document ID.
|
1017
|
+
# @return [Sphinx::Client] self.
|
1018
|
+
#
|
1019
|
+
# @example
|
1020
|
+
# sphinx.set_id_range(10, 1000)
|
1021
|
+
#
|
1022
|
+
# @raise [ArgumentError] Occurred when parameters are invalid.
|
1023
|
+
#
|
1024
|
+
# @see http://www.sphinxsearch.com/docs/current.html#api-func-setidrange Section 6.4.1, "SetIDRange"
|
1025
|
+
#
|
1026
|
+
def set_id_range(min, max)
|
571
1027
|
raise ArgumentError, '"min" argument must be Integer' unless min.respond_to?(:integer?) and min.integer?
|
572
1028
|
raise ArgumentError, '"max" argument must be Integer' unless max.respond_to?(:integer?) and max.integer?
|
573
1029
|
raise ArgumentError, '"max" argument greater or equal to "min"' unless min <= max
|
574
1030
|
|
575
1031
|
@min_id = min
|
576
1032
|
@max_id = max
|
1033
|
+
self
|
577
1034
|
end
|
578
|
-
|
579
|
-
|
580
|
-
#
|
581
|
-
#
|
582
|
-
#
|
583
|
-
|
1035
|
+
alias :SetIDRange :set_id_range
|
1036
|
+
|
1037
|
+
# Adds new integer values set filter.
|
1038
|
+
#
|
1039
|
+
# On this call, additional new filter is added to the existing
|
1040
|
+
# list of filters. $attribute must be a string with attribute
|
1041
|
+
# name. +values+ must be a plain array containing integer
|
1042
|
+
# values. +exclude+ must be a boolean value; it controls
|
1043
|
+
# whether to accept the matching documents (default mode, when
|
1044
|
+
# +exclude+ is +false+) or reject them.
|
1045
|
+
#
|
1046
|
+
# Only those documents where +attribute+ column value stored in
|
1047
|
+
# the index matches any of the values from +values+ array will
|
1048
|
+
# be matched (or rejected, if +exclude+ is +true+).
|
1049
|
+
#
|
1050
|
+
# @param [String, Symbol] attribute an attribute name to filter by.
|
1051
|
+
# @param [Array<Integer>] values an +Array+ of integers with given attribute values.
|
1052
|
+
# @param [Boolean] exclude indicating whether documents with given attribute
|
1053
|
+
# matching specified values should be excluded from search results.
|
1054
|
+
# @return [Sphinx::Client] self.
|
1055
|
+
#
|
1056
|
+
# @example
|
1057
|
+
# sphinx.set_filter(:group_id, [10, 15, 20])
|
1058
|
+
# sphinx.set_filter(:group_id, [10, 15, 20], true)
|
1059
|
+
#
|
1060
|
+
# @raise [ArgumentError] Occurred when parameters are invalid.
|
1061
|
+
#
|
1062
|
+
# @see http://www.sphinxsearch.com/docs/current.html#api-func-setfilter Section 6.4.2, "SetFilter"
|
1063
|
+
# @see #set_filter_range
|
1064
|
+
# @see #set_filter_float_range
|
1065
|
+
#
|
1066
|
+
def set_filter(attribute, values, exclude = false)
|
584
1067
|
raise ArgumentError, '"attribute" argument must be String or Symbol' unless attribute.kind_of?(String) or attribute.kind_of?(Symbol)
|
585
1068
|
raise ArgumentError, '"values" argument must be Array' unless values.kind_of?(Array)
|
586
1069
|
raise ArgumentError, '"values" argument must not be empty' if values.empty?
|
@@ -589,94 +1072,194 @@ module Sphinx
|
|
589
1072
|
values.each do |value|
|
590
1073
|
raise ArgumentError, '"values" argument must be Array of Integer' unless value.respond_to?(:integer?) and value.integer?
|
591
1074
|
end
|
592
|
-
|
1075
|
+
|
593
1076
|
@filters << { 'type' => SPH_FILTER_VALUES, 'attr' => attribute.to_s, 'exclude' => exclude, 'values' => values }
|
1077
|
+
self
|
594
1078
|
end
|
595
|
-
|
596
|
-
|
597
|
-
#
|
598
|
-
#
|
599
|
-
#
|
600
|
-
|
1079
|
+
alias :SetFilter :set_filter
|
1080
|
+
|
1081
|
+
# Adds new integer range filter.
|
1082
|
+
#
|
1083
|
+
# On this call, additional new filter is added to the existing
|
1084
|
+
# list of filters. +attribute+ must be a string with attribute
|
1085
|
+
# name. +min+ and +max+ must be integers that define the acceptable
|
1086
|
+
# attribute values range (including the boundaries). +exclude+
|
1087
|
+
# must be a boolean value; it controls whether to accept the
|
1088
|
+
# matching documents (default mode, when +exclude+ is false) or
|
1089
|
+
# reject them.
|
1090
|
+
#
|
1091
|
+
# Only those documents where +attribute+ column value stored
|
1092
|
+
# in the index is between +min+ and +max+ (including values
|
1093
|
+
# that are exactly equal to +min+ or +max+) will be matched
|
1094
|
+
# (or rejected, if +exclude+ is true).
|
1095
|
+
#
|
1096
|
+
# @param [String, Symbol] attribute an attribute name to filter by.
|
1097
|
+
# @param [Integer] min min value of the given attribute.
|
1098
|
+
# @param [Integer] max max value of the given attribute.
|
1099
|
+
# @param [Boolean] exclude indicating whether documents with given attribute
|
1100
|
+
# matching specified boundaries should be excluded from search results.
|
1101
|
+
# @return [Sphinx::Client] self.
|
1102
|
+
#
|
1103
|
+
# @example
|
1104
|
+
# sphinx.set_filter_range(:group_id, 10, 20)
|
1105
|
+
# sphinx.set_filter_range(:group_id, 10, 20, true)
|
1106
|
+
#
|
1107
|
+
# @raise [ArgumentError] Occurred when parameters are invalid.
|
1108
|
+
#
|
1109
|
+
# @see http://www.sphinxsearch.com/docs/current.html#api-func-setfilterrange Section 6.4.3, "SetFilterRange"
|
1110
|
+
# @see #set_filter
|
1111
|
+
# @see #set_filter_float_range
|
1112
|
+
#
|
1113
|
+
def set_filter_range(attribute, min, max, exclude = false)
|
601
1114
|
raise ArgumentError, '"attribute" argument must be String or Symbol' unless attribute.kind_of?(String) or attribute.kind_of?(Symbol)
|
602
1115
|
raise ArgumentError, '"min" argument must be Integer' unless min.respond_to?(:integer?) and min.integer?
|
603
1116
|
raise ArgumentError, '"max" argument must be Integer' unless max.respond_to?(:integer?) and max.integer?
|
604
1117
|
raise ArgumentError, '"max" argument greater or equal to "min"' unless min <= max
|
605
1118
|
raise ArgumentError, '"exclude" argument must be Boolean' unless exclude.kind_of?(TrueClass) or exclude.kind_of?(FalseClass)
|
606
|
-
|
1119
|
+
|
607
1120
|
@filters << { 'type' => SPH_FILTER_RANGE, 'attr' => attribute.to_s, 'exclude' => exclude, 'min' => min, 'max' => max }
|
1121
|
+
self
|
608
1122
|
end
|
609
|
-
|
610
|
-
|
611
|
-
#
|
612
|
-
#
|
613
|
-
#
|
614
|
-
|
1123
|
+
alias :SetFilterRange :set_filter_range
|
1124
|
+
|
1125
|
+
# Adds new float range filter.
|
1126
|
+
#
|
1127
|
+
# On this call, additional new filter is added to the existing
|
1128
|
+
# list of filters. +attribute+ must be a string with attribute name.
|
1129
|
+
# +min+ and +max+ must be floats that define the acceptable
|
1130
|
+
# attribute values range (including the boundaries). +exclude+ must
|
1131
|
+
# be a boolean value; it controls whether to accept the matching
|
1132
|
+
# documents (default mode, when +exclude+ is false) or reject them.
|
1133
|
+
#
|
1134
|
+
# Only those documents where +attribute+ column value stored in
|
1135
|
+
# the index is between +min+ and +max+ (including values that are
|
1136
|
+
# exactly equal to +min+ or +max+) will be matched (or rejected,
|
1137
|
+
# if +exclude+ is true).
|
1138
|
+
#
|
1139
|
+
# @param [String, Symbol] attribute an attribute name to filter by.
|
1140
|
+
# @param [Integer, Float] min min value of the given attribute.
|
1141
|
+
# @param [Integer, Float] max max value of the given attribute.
|
1142
|
+
# @param [Boolean] exclude indicating whether documents with given attribute
|
1143
|
+
# matching specified boundaries should be excluded from search results.
|
1144
|
+
# @return [Sphinx::Client] self.
|
1145
|
+
#
|
1146
|
+
# @example
|
1147
|
+
# sphinx.set_filter_float_range(:group_id, 10.5, 20)
|
1148
|
+
# sphinx.set_filter_float_range(:group_id, 10.5, 20, true)
|
1149
|
+
#
|
1150
|
+
# @raise [ArgumentError] Occurred when parameters are invalid.
|
1151
|
+
#
|
1152
|
+
# @see http://www.sphinxsearch.com/docs/current.html#api-func-setfilterfloatrange Section 6.4.4, "SetFilterFloatRange"
|
1153
|
+
# @see #set_filter
|
1154
|
+
# @see #set_filter_range
|
1155
|
+
#
|
1156
|
+
def set_filter_float_range(attribute, min, max, exclude = false)
|
615
1157
|
raise ArgumentError, '"attribute" argument must be String or Symbol' unless attribute.kind_of?(String) or attribute.kind_of?(Symbol)
|
616
1158
|
raise ArgumentError, '"min" argument must be Float or Integer' unless min.kind_of?(Float) or (min.respond_to?(:integer?) and min.integer?)
|
617
1159
|
raise ArgumentError, '"max" argument must be Float or Integer' unless max.kind_of?(Float) or (max.respond_to?(:integer?) and max.integer?)
|
618
1160
|
raise ArgumentError, '"max" argument greater or equal to "min"' unless min <= max
|
619
1161
|
raise ArgumentError, '"exclude" argument must be Boolean' unless exclude.kind_of?(TrueClass) or exclude.kind_of?(FalseClass)
|
620
|
-
|
1162
|
+
|
621
1163
|
@filters << { 'type' => SPH_FILTER_FLOATRANGE, 'attr' => attribute.to_s, 'exclude' => exclude, 'min' => min.to_f, 'max' => max.to_f }
|
1164
|
+
self
|
622
1165
|
end
|
623
|
-
|
624
|
-
|
625
|
-
#
|
626
|
-
#
|
627
|
-
#
|
628
|
-
# must be
|
629
|
-
#
|
630
|
-
#
|
631
|
-
#
|
632
|
-
#
|
633
|
-
#
|
634
|
-
|
1166
|
+
alias :SetFilterFloatRange :set_filter_float_range
|
1167
|
+
|
1168
|
+
# Sets anchor point for and geosphere distance (geodistance)
|
1169
|
+
# calculations, and enable them.
|
1170
|
+
#
|
1171
|
+
# +attrlat+ and +attrlong+ must be strings that contain the names
|
1172
|
+
# of latitude and longitude attributes, respectively. +lat+ and
|
1173
|
+
# +long+ are floats that specify anchor point latitude and
|
1174
|
+
# longitude, in radians.
|
1175
|
+
#
|
1176
|
+
# Once an anchor point is set, you can use magic <tt>"@geodist"</tt>
|
1177
|
+
# attribute name in your filters and/or sorting expressions.
|
1178
|
+
# Sphinx will compute geosphere distance between the given anchor
|
1179
|
+
# point and a point specified by latitude and lognitude attributes
|
1180
|
+
# from each full-text match, and attach this value to the resulting
|
1181
|
+
# match. The latitude and longitude values both in {#set_geo_anchor}
|
1182
|
+
# and the index attribute data are expected to be in radians.
|
1183
|
+
# The result will be returned in meters, so geodistance value of
|
1184
|
+
# 1000.0 means 1 km. 1 mile is approximately 1609.344 meters.
|
1185
|
+
#
|
1186
|
+
# @param [String, Symbol] attrlat a name of latitude attribute.
|
1187
|
+
# @param [String, Symbol] attrlong a name of longitude attribute.
|
1188
|
+
# @param [Integer, Float] lat an anchor point latitude, in radians.
|
1189
|
+
# @param [Integer, Float] long an anchor point longitude, in radians.
|
1190
|
+
# @return [Sphinx::Client] self.
|
1191
|
+
#
|
1192
|
+
# @example
|
1193
|
+
# sphinx.set_geo_anchor(:latitude, :longitude, 192.5, 143.5)
|
1194
|
+
#
|
1195
|
+
# @raise [ArgumentError] Occurred when parameters are invalid.
|
1196
|
+
#
|
1197
|
+
# @see http://www.sphinxsearch.com/docs/current.html#api-func-setgeoanchor Section 6.4.5, "SetGeoAnchor"
|
1198
|
+
#
|
1199
|
+
def set_geo_anchor(attrlat, attrlong, lat, long)
|
635
1200
|
raise ArgumentError, '"attrlat" argument must be String or Symbol' unless attrlat.kind_of?(String) or attrlat.kind_of?(Symbol)
|
636
1201
|
raise ArgumentError, '"attrlong" argument must be String or Symbol' unless attrlong.kind_of?(String) or attrlong.kind_of?(Symbol)
|
637
1202
|
raise ArgumentError, '"lat" argument must be Float or Integer' unless lat.kind_of?(Float) or (lat.respond_to?(:integer?) and lat.integer?)
|
638
1203
|
raise ArgumentError, '"long" argument must be Float or Integer' unless long.kind_of?(Float) or (long.respond_to?(:integer?) and long.integer?)
|
639
1204
|
|
640
1205
|
@anchor = { 'attrlat' => attrlat.to_s, 'attrlong' => attrlong.to_s, 'lat' => lat.to_f, 'long' => long.to_f }
|
1206
|
+
self
|
641
1207
|
end
|
1208
|
+
alias :SetGeoAnchor :set_geo_anchor
|
642
1209
|
|
643
|
-
|
1210
|
+
#=================================================================
|
1211
|
+
# GROUP BY settings
|
1212
|
+
#=================================================================
|
1213
|
+
|
1214
|
+
# Sets grouping attribute, function, and groups sorting mode; and
|
1215
|
+
# enables grouping (as described in Section 4.6, "Grouping (clustering) search results").
|
644
1216
|
#
|
645
|
-
#
|
646
|
-
#
|
1217
|
+
# +attribute+ is a string that contains group-by attribute name.
|
1218
|
+
# +func+ is a constant that chooses a function applied to the
|
1219
|
+
# attribute value in order to compute group-by key. +groupsort+
|
1220
|
+
# is a clause that controls how the groups will be sorted. Its
|
1221
|
+
# syntax is similar to that described in Section 4.5,
|
1222
|
+
# "SPH_SORT_EXTENDED mode".
|
647
1223
|
#
|
648
|
-
#
|
649
|
-
#
|
1224
|
+
# Grouping feature is very similar in nature to <tt>GROUP BY</tt> clause
|
1225
|
+
# from SQL. Results produces by this function call are going to
|
1226
|
+
# be the same as produced by the following pseudo code:
|
650
1227
|
#
|
651
|
-
#
|
652
|
-
# grouping function value and matches count attached.
|
1228
|
+
# SELECT ... GROUP BY func(attribute) ORDER BY groupsort
|
653
1229
|
#
|
654
|
-
#
|
655
|
-
#
|
656
|
-
#
|
1230
|
+
# Note that it's +groupsort+ that affects the order of matches in
|
1231
|
+
# the final result set. Sorting mode (see {#set_sort_mode}) affect
|
1232
|
+
# the ordering of matches within group, ie. what match will be
|
1233
|
+
# selected as the best one from the group. So you can for instance
|
1234
|
+
# order the groups by matches count and select the most relevant
|
1235
|
+
# match within each group at the same time.
|
657
1236
|
#
|
658
|
-
#
|
659
|
-
#
|
660
|
-
#
|
661
|
-
#
|
1237
|
+
# Starting with version 0.9.9-rc2, aggregate functions (<tt>AVG()</tt>,
|
1238
|
+
# <tt>MIN()</tt>, <tt>MAX()</tt>, <tt>SUM()</tt>) are supported
|
1239
|
+
# through {#set_select} API call when using <tt>GROUP BY</tt>.
|
1240
|
+
#
|
1241
|
+
# You can specify group function and attribute as String
|
1242
|
+
# ("attr", "day", etc), Symbol (:attr, :day, etc), or
|
1243
|
+
# Fixnum constant (SPH_GROUPBY_ATTR, SPH_GROUPBY_DAY, etc).
|
1244
|
+
#
|
1245
|
+
# @param [String, Symbol] attribute an attribute name to group by.
|
1246
|
+
# @param [Integer, String, Symbol] func a grouping function.
|
1247
|
+
# @param [String] groupsort a groups sorting mode.
|
1248
|
+
# @return [Sphinx::Client] self.
|
662
1249
|
#
|
663
|
-
#
|
664
|
-
#
|
1250
|
+
# @example
|
1251
|
+
# sphinx.set_group_by(:tag_id, :attr)
|
665
1252
|
#
|
666
|
-
#
|
667
|
-
# the whole index.
|
1253
|
+
# @raise [ArgumentError] Occurred when parameters are invalid.
|
668
1254
|
#
|
669
|
-
#
|
670
|
-
#
|
671
|
-
#
|
672
|
-
#
|
1255
|
+
# @see http://www.sphinxsearch.com/docs/current.html#clustering Section 4.6, "Grouping (clustering) search results"
|
1256
|
+
# @see http://www.sphinxsearch.com/docs/current.html#sort-extended Section 4.5, "SPH_SORT_EXTENDED mode"
|
1257
|
+
# @see http://www.sphinxsearch.com/docs/current.html#api-func-setgroupby Section 6.5.1, "SetGroupBy"
|
1258
|
+
# @see #set_sort_mode
|
1259
|
+
# @see #set_select
|
1260
|
+
# @see #set_group_distinct
|
673
1261
|
#
|
674
|
-
|
675
|
-
# attribute with SPH_GROUPBY_DAY function, then the result set will
|
676
|
-
# contain one most relevant match per each day when there were any
|
677
|
-
# matches published, with day number and per-day match count attached,
|
678
|
-
# and sorted by day number in descending order (ie. recent days first).
|
679
|
-
def SetGroupBy(attribute, func, groupsort = '@group desc')
|
1262
|
+
def set_group_by(attribute, func, groupsort = '@group desc')
|
680
1263
|
raise ArgumentError, '"attribute" argument must be String or Symbol' unless attribute.kind_of?(String) or attribute.kind_of?(Symbol)
|
681
1264
|
raise ArgumentError, '"groupsort" argument must be String' unless groupsort.kind_of?(String)
|
682
1265
|
|
@@ -696,217 +1279,311 @@ module Sphinx
|
|
696
1279
|
@groupby = attribute.to_s
|
697
1280
|
@groupfunc = func
|
698
1281
|
@groupsort = groupsort
|
1282
|
+
self
|
699
1283
|
end
|
700
|
-
|
701
|
-
# Set count-distinct attribute for group-by queries.
|
702
|
-
def SetGroupDistinct(attribute)
|
703
|
-
raise ArgumentError, '"attribute" argument must be String or Symbol' unless attribute.kind_of?(String) or attribute.kind_of?(Symbol)
|
1284
|
+
alias :SetGroupBy :set_group_by
|
704
1285
|
|
705
|
-
|
706
|
-
|
707
|
-
|
708
|
-
# Sets distributed retry count and delay.
|
1286
|
+
# Sets attribute name for per-group distinct values count
|
1287
|
+
# calculations. Only available for grouping queries.
|
709
1288
|
#
|
710
|
-
#
|
711
|
-
#
|
712
|
-
#
|
713
|
-
#
|
714
|
-
#
|
715
|
-
#
|
1289
|
+
# +attribute+ is a string that contains the attribute name. For
|
1290
|
+
# each group, all values of this attribute will be stored (as
|
1291
|
+
# RAM limits permit), then the amount of distinct values will
|
1292
|
+
# be calculated and returned to the client. This feature is
|
1293
|
+
# similar to <tt>COUNT(DISTINCT)</tt> clause in standard SQL;
|
1294
|
+
# so these Sphinx calls:
|
716
1295
|
#
|
717
|
-
|
718
|
-
|
719
|
-
raise ArgumentError, '"delay" argument must be Integer' unless delay.respond_to?(:integer?) and delay.integer?
|
720
|
-
|
721
|
-
@retrycount = count
|
722
|
-
@retrydelay = delay
|
723
|
-
end
|
724
|
-
|
725
|
-
# Sets temporary (per-query) per-document attribute value overrides. Only
|
726
|
-
# supports scalar attributes. +values+ must be a +Hash+ that maps document
|
727
|
-
# IDs to overridden attribute values.
|
1296
|
+
# sphinx.set_group_by(:category, :attr, '@count desc')
|
1297
|
+
# sphinx.set_group_distinct(:vendor)
|
728
1298
|
#
|
729
|
-
#
|
730
|
-
# documents within a single query, leaving all other queries unaffected.
|
731
|
-
# This might be useful for personalized data. For example, assume you're
|
732
|
-
# implementing a personalized search function that wants to boost the posts
|
733
|
-
# that the user's friends recommend. Such data is not just dynamic, but
|
734
|
-
# also personal; so you can't simply put it in the index because you don't
|
735
|
-
# want everyone's searches affected. Overrides, on the other hand, are local
|
736
|
-
# to a single query and invisible to everyone else. So you can, say, setup
|
737
|
-
# a "friends_weight" value for every document, defaulting to 0, then
|
738
|
-
# temporary override it with 1 for documents 123, 456 and 789 (recommended
|
739
|
-
# by exactly the friends of current user), and use that value when ranking.
|
1299
|
+
# can be expressed using the following SQL clauses:
|
740
1300
|
#
|
741
|
-
|
742
|
-
|
743
|
-
|
744
|
-
|
745
|
-
|
746
|
-
|
747
|
-
attrtype = self.class.const_get("SPH_ATTR_#{attrtype.to_s.upcase}")
|
748
|
-
rescue NameError
|
749
|
-
raise ArgumentError, "\"attrtype\" argument value \"#{attrtype}\" is invalid"
|
750
|
-
end
|
751
|
-
when Fixnum
|
752
|
-
raise ArgumentError, "\"attrtype\" argument value \"#{attrtype}\" is invalid" unless (SPH_ATTR_INTEGER..SPH_ATTR_BIGINT).include?(attrtype)
|
753
|
-
else
|
754
|
-
raise ArgumentError, '"attrtype" argument must be Fixnum, String, or Symbol'
|
755
|
-
end
|
756
|
-
|
757
|
-
raise ArgumentError, '"values" argument must be Hash' unless values.kind_of?(Hash)
|
758
|
-
|
759
|
-
values.each do |id, value|
|
760
|
-
raise ArgumentError, '"values" argument must be Hash map of Integer to Integer or Time' unless id.respond_to?(:integer?) and id.integer?
|
761
|
-
case attrtype
|
762
|
-
when SPH_ATTR_TIMESTAMP
|
763
|
-
raise ArgumentError, '"values" argument must be Hash map of Integer to Integer or Time' unless (value.respond_to?(:integer?) and value.integer?) or value.kind_of?(Time)
|
764
|
-
when SPH_ATTR_FLOAT
|
765
|
-
raise ArgumentError, '"values" argument must be Hash map of Integer to Float or Integer' unless value.kind_of?(Float) or (value.respond_to?(:integer?) and value.integer?)
|
766
|
-
else
|
767
|
-
# SPH_ATTR_INTEGER, SPH_ATTR_ORDINAL, SPH_ATTR_BOOL, SPH_ATTR_BIGINT
|
768
|
-
raise ArgumentError, '"values" argument must be Hash map of Integer to Integer' unless value.respond_to?(:integer?) and value.integer?
|
769
|
-
end
|
770
|
-
end
|
771
|
-
|
772
|
-
@overrides << { 'attr' => attrname.to_s, 'type' => attrtype, 'values' => values }
|
773
|
-
end
|
774
|
-
|
775
|
-
# Sets the select clause, listing specific attributes to fetch, and
|
776
|
-
# expressions to compute and fetch. Clause syntax mimics SQL.
|
1301
|
+
# SELECT id, weight, all-attributes,
|
1302
|
+
# COUNT(DISTINCT vendor) AS @distinct,
|
1303
|
+
# COUNT(*) AS @count
|
1304
|
+
# FROM products
|
1305
|
+
# GROUP BY category
|
1306
|
+
# ORDER BY @count DESC
|
777
1307
|
#
|
778
|
-
#
|
779
|
-
#
|
780
|
-
#
|
781
|
-
#
|
782
|
-
#
|
783
|
-
#
|
784
|
-
#
|
785
|
-
#
|
1308
|
+
# In the sample pseudo code shown just above, {#set_group_distinct}
|
1309
|
+
# call corresponds to <tt>COUNT(DISINCT vendor)</tt> clause only.
|
1310
|
+
# <tt>GROUP BY</tt>, <tt>ORDER BY</tt>, and <tt>COUNT(*)</tt>
|
1311
|
+
# clauses are all an equivalent of {#set_group_by} settings. Both
|
1312
|
+
# queries will return one matching row for each category. In
|
1313
|
+
# addition to indexed attributes, matches will also contain
|
1314
|
+
# total per-category matches count, and the count of distinct
|
1315
|
+
# vendor IDs within each category.
|
786
1316
|
#
|
787
|
-
#
|
788
|
-
#
|
789
|
-
# Computed expressions can be used for sorting, filtering, and grouping,
|
790
|
-
# just as the regular attributes.
|
1317
|
+
# @param [String, Symbol] attribute an attribute name.
|
1318
|
+
# @return [Sphinx::Client] self.
|
791
1319
|
#
|
792
|
-
#
|
793
|
-
#
|
794
|
-
# <tt>GROUP BY</tt>.
|
1320
|
+
# @example
|
1321
|
+
# sphinx.set_group_distinct(:category_id)
|
795
1322
|
#
|
796
|
-
#
|
797
|
-
# functions (+SetGeoAnchor+) are now internally implemented
|
798
|
-
# using this computed expressions mechanism, using magic names '<tt>@expr</tt>'
|
799
|
-
# and '<tt>@geodist</tt>' respectively.
|
1323
|
+
# @raise [ArgumentError] Occurred when parameters are invalid.
|
800
1324
|
#
|
801
|
-
#
|
1325
|
+
# @see http://www.sphinxsearch.com/docs/current.html#api-func-setgroupdistinct Section 6.5.2, "SetGroupDistinct"
|
1326
|
+
# @see #set_group_by
|
802
1327
|
#
|
803
|
-
|
804
|
-
|
805
|
-
# sphinx.SetSelect('*, AVG(price) AS avgprice')
|
806
|
-
#
|
807
|
-
def SetSelect(select)
|
808
|
-
raise ArgumentError, '"select" argument must be String' unless select.kind_of?(String)
|
1328
|
+
def set_group_distinct(attribute)
|
1329
|
+
raise ArgumentError, '"attribute" argument must be String or Symbol' unless attribute.kind_of?(String) or attribute.kind_of?(Symbol)
|
809
1330
|
|
810
|
-
@
|
1331
|
+
@groupdistinct = attribute.to_s
|
1332
|
+
self
|
811
1333
|
end
|
812
|
-
|
1334
|
+
alias :SetGroupDistinct :set_group_distinct
|
1335
|
+
|
1336
|
+
#=================================================================
|
1337
|
+
# Querying
|
1338
|
+
#=================================================================
|
1339
|
+
|
813
1340
|
# Clears all currently set filters.
|
814
1341
|
#
|
815
1342
|
# This call is only normally required when using multi-queries. You might want
|
816
1343
|
# to set different filters for different queries in the batch. To do that,
|
817
|
-
# you should call
|
1344
|
+
# you should call {#reset_filters} and add new filters using the respective calls.
|
1345
|
+
#
|
1346
|
+
# @return [Sphinx::Client] self.
|
818
1347
|
#
|
819
|
-
#
|
1348
|
+
# @example
|
1349
|
+
# sphinx.reset_filters
|
820
1350
|
#
|
821
|
-
#
|
1351
|
+
# @see #set_filter
|
1352
|
+
# @see #set_filter_range
|
1353
|
+
# @see #set_filter_float_range
|
1354
|
+
# @see #set_geo_anchor
|
822
1355
|
#
|
823
|
-
def
|
1356
|
+
def reset_filters
|
824
1357
|
@filters = []
|
825
1358
|
@anchor = []
|
1359
|
+
self
|
826
1360
|
end
|
827
|
-
|
1361
|
+
alias :ResetFilters :reset_filters
|
1362
|
+
|
828
1363
|
# Clears all currently group-by settings, and disables group-by.
|
829
1364
|
#
|
830
1365
|
# This call is only normally required when using multi-queries. You can
|
831
|
-
# change individual group-by settings using
|
832
|
-
# calls, but you can not disable group-by using those calls.
|
1366
|
+
# change individual group-by settings using {#set_group_by} and {#set_group_distinct}
|
1367
|
+
# calls, but you can not disable group-by using those calls. {#reset_group_by}
|
833
1368
|
# fully resets previous group-by settings and disables group-by mode in the
|
834
|
-
# current state, so that subsequent
|
1369
|
+
# current state, so that subsequent {#add_query} calls can perform non-grouping
|
835
1370
|
# searches.
|
836
1371
|
#
|
837
|
-
#
|
1372
|
+
# @return [Sphinx::Client] self.
|
838
1373
|
#
|
839
|
-
#
|
1374
|
+
# @example
|
1375
|
+
# sphinx.reset_group_by
|
840
1376
|
#
|
841
|
-
|
1377
|
+
# @see #set_group_by
|
1378
|
+
# @see #set_group_distinct
|
1379
|
+
#
|
1380
|
+
def reset_group_by
|
842
1381
|
@groupby = ''
|
843
1382
|
@groupfunc = SPH_GROUPBY_DAY
|
844
1383
|
@groupsort = '@group desc'
|
845
1384
|
@groupdistinct = ''
|
1385
|
+
self
|
846
1386
|
end
|
847
|
-
|
1387
|
+
alias :ResetGroupBy :reset_group_by
|
1388
|
+
|
848
1389
|
# Clear all attribute value overrides (for multi-queries).
|
849
|
-
def ResetOverrides
|
850
|
-
@overrides = []
|
851
|
-
end
|
852
|
-
|
853
|
-
# Connect to searchd server and run given search query.
|
854
1390
|
#
|
855
|
-
#
|
856
|
-
|
857
|
-
#
|
858
|
-
#
|
859
|
-
# dash, and underscore; everything else is considered a separator. Therefore,
|
860
|
-
# all the following calls are valid and will search two indexes:
|
1391
|
+
# This call is only normally required when using multi-queries. You might want
|
1392
|
+
# to set field overrides for different queries in the batch. To do that,
|
1393
|
+
# you should call {#reset_overrides} and add new overrides using the
|
1394
|
+
# respective calls.
|
861
1395
|
#
|
862
|
-
#
|
863
|
-
# sphinx.Query('test query', 'main;delta')
|
864
|
-
# sphinx.Query('test query', 'main, delta')
|
1396
|
+
# @return [Sphinx::Client] self.
|
865
1397
|
#
|
866
|
-
#
|
867
|
-
#
|
868
|
-
# for sorting and returning to client. Therefore, in the example above,
|
869
|
-
# matches from "delta" index will always "win" over matches from "main".
|
1398
|
+
# @example
|
1399
|
+
# sphinx.reset_overrides
|
870
1400
|
#
|
871
|
-
#
|
872
|
-
#
|
873
|
-
|
874
|
-
|
875
|
-
|
876
|
-
|
877
|
-
|
878
|
-
|
879
|
-
|
1401
|
+
# @see #set_override
|
1402
|
+
#
|
1403
|
+
def reset_overrides
|
1404
|
+
@overrides = []
|
1405
|
+
self
|
1406
|
+
end
|
1407
|
+
alias :ResetOverrides :reset_overrides
|
1408
|
+
|
1409
|
+
# Connects to searchd server, runs given search query with
|
1410
|
+
# current settings, obtains and returns the result set.
|
1411
|
+
#
|
1412
|
+
# +query+ is a query string. +index+ is an index name (or names)
|
1413
|
+
# string. Returns false and sets {#last_error} message on general
|
1414
|
+
# error. Returns search result set on success. Additionally,
|
1415
|
+
# the contents of +comment+ are sent to the query log, marked in
|
1416
|
+
# square brackets, just before the search terms, which can be very
|
1417
|
+
# useful for debugging. Currently, the comment is limited to 128
|
1418
|
+
# characters.
|
1419
|
+
#
|
1420
|
+
# Default value for +index+ is <tt>"*"</tt> that means to query
|
1421
|
+
# all local indexes. Characters allowed in index names include
|
1422
|
+
# Latin letters (a-z), numbers (0-9), minus sign (-), and
|
1423
|
+
# underscore (_); everything else is considered a separator.
|
1424
|
+
# Therefore, all of the following samples calls are valid and
|
1425
|
+
# will search the same two indexes:
|
1426
|
+
#
|
1427
|
+
# sphinx.query('test query', 'main delta')
|
1428
|
+
# sphinx.query('test query', 'main;delta')
|
1429
|
+
# sphinx.query('test query', 'main, delta');
|
1430
|
+
#
|
1431
|
+
# Index specification order matters. If document with identical
|
1432
|
+
# IDs are found in two or more indexes, weight and attribute
|
1433
|
+
# values from the very last matching index will be used for
|
1434
|
+
# sorting and returning to client (unless explicitly overridden
|
1435
|
+
# with {#set_index_weights}). Therefore, in the example above,
|
1436
|
+
# matches from "delta" index will always win over matches
|
1437
|
+
# from "main".
|
1438
|
+
#
|
1439
|
+
# On success, {#query} returns a result set that contains some
|
1440
|
+
# of the found matches (as requested by {#set_limits}) and
|
1441
|
+
# additional general per-query statistics. The result set
|
1442
|
+
# is an +Hash+ with the following keys and values:
|
1443
|
+
#
|
1444
|
+
# <tt>"matches"</tt>::
|
1445
|
+
# Array with small +Hash+es containing document weight and
|
1446
|
+
# attribute values.
|
1447
|
+
# <tt>"total"</tt>::
|
1448
|
+
# Total amount of matches retrieved on server (ie. to the server
|
1449
|
+
# side result set) by this query. You can retrieve up to this
|
1450
|
+
# amount of matches from server for this query text with current
|
1451
|
+
# query settings.
|
1452
|
+
# <tt>"total_found"</tt>::
|
1453
|
+
# Total amount of matching documents in index (that were found
|
1454
|
+
# and procesed on server).
|
1455
|
+
# <tt>"words"</tt>::
|
1456
|
+
# Hash which maps query keywords (case-folded, stemmed, and
|
1457
|
+
# otherwise processed) to a small Hash with per-keyword statitics
|
1458
|
+
# ("docs", "hits").
|
1459
|
+
# <tt>"error"</tt>::
|
1460
|
+
# Query error message reported by searchd (string, human readable).
|
1461
|
+
# Empty if there were no errors.
|
1462
|
+
# <tt>"warning"</tt>::
|
1463
|
+
# Query warning message reported by searchd (string, human readable).
|
1464
|
+
# Empty if there were no warnings.
|
1465
|
+
#
|
1466
|
+
# It should be noted that {#query} carries out the same actions as
|
1467
|
+
# {#add_query} and {#run_queries} without the intermediate steps; it
|
1468
|
+
# is analoguous to a single {#add_query} call, followed by a
|
1469
|
+
# corresponding {#run_queries}, then returning the first array
|
1470
|
+
# element of matches (from the first, and only, query.)
|
1471
|
+
#
|
1472
|
+
# @param [String] query a query string.
|
1473
|
+
# @param [String] index an index name (or names).
|
1474
|
+
# @param [String] comment a comment to be sent to the query log.
|
1475
|
+
# @return [Hash, false] result set described above or +false+ on error.
|
1476
|
+
#
|
1477
|
+
# @example
|
1478
|
+
# sphinx.query('some search text', '*', 'search page')
|
1479
|
+
#
|
1480
|
+
# @see http://www.sphinxsearch.com/docs/current.html#api-func-query Section 6.6.1, "Query"
|
1481
|
+
# @see #add_query
|
1482
|
+
# @see #run_queries
|
1483
|
+
#
|
1484
|
+
def query(query, index = '*', comment = '')
|
880
1485
|
@reqs = []
|
881
|
-
|
882
|
-
self.
|
883
|
-
results = self.
|
884
|
-
|
1486
|
+
|
1487
|
+
self.add_query(query, index, comment)
|
1488
|
+
results = self.run_queries
|
1489
|
+
|
885
1490
|
# probably network error; error message should be already filled
|
886
1491
|
return false unless results.instance_of?(Array)
|
887
|
-
|
1492
|
+
|
888
1493
|
@error = results[0]['error']
|
889
1494
|
@warning = results[0]['warning']
|
890
|
-
|
1495
|
+
|
891
1496
|
return false if results[0]['status'] == SEARCHD_ERROR
|
892
1497
|
return results[0]
|
893
1498
|
end
|
894
|
-
|
895
|
-
|
896
|
-
#
|
897
|
-
#
|
898
|
-
#
|
899
|
-
#
|
900
|
-
#
|
901
|
-
#
|
902
|
-
#
|
903
|
-
#
|
1499
|
+
alias :Query :query
|
1500
|
+
|
1501
|
+
# Adds additional query with current settings to multi-query batch.
|
1502
|
+
# +query+ is a query string. +index+ is an index name (or names)
|
1503
|
+
# string. Additionally if provided, the contents of +comment+ are
|
1504
|
+
# sent to the query log, marked in square brackets, just before
|
1505
|
+
# the search terms, which can be very useful for debugging.
|
1506
|
+
# Currently, this is limited to 128 characters. Returns index
|
1507
|
+
# to results array returned from {#run_queries}.
|
1508
|
+
#
|
1509
|
+
# Batch queries (or multi-queries) enable searchd to perform
|
1510
|
+
# internal optimizations if possible. They also reduce network
|
1511
|
+
# connection overheads and search process creation overheads in all
|
1512
|
+
# cases. They do not result in any additional overheads compared
|
1513
|
+
# to simple queries. Thus, if you run several different queries
|
1514
|
+
# from your web page, you should always consider using multi-queries.
|
1515
|
+
#
|
1516
|
+
# For instance, running the same full-text query but with different
|
1517
|
+
# sorting or group-by settings will enable searchd to perform
|
1518
|
+
# expensive full-text search and ranking operation only once, but
|
1519
|
+
# compute multiple group-by results from its output.
|
1520
|
+
#
|
1521
|
+
# This can be a big saver when you need to display not just plain
|
1522
|
+
# search results but also some per-category counts, such as the
|
1523
|
+
# amount of products grouped by vendor. Without multi-query, you
|
1524
|
+
# would have to run several queries which perform essentially the
|
1525
|
+
# same search and retrieve the same matches, but create result
|
1526
|
+
# sets differently. With multi-query, you simply pass all these
|
1527
|
+
# queries in a single batch and Sphinx optimizes the redundant
|
1528
|
+
# full-text search internally.
|
1529
|
+
#
|
1530
|
+
# {#add_query} internally saves full current settings state along
|
1531
|
+
# with the query, and you can safely change them afterwards for
|
1532
|
+
# subsequent {#add_query} calls. Already added queries will not
|
1533
|
+
# be affected; there's actually no way to change them at all.
|
1534
|
+
# Here's an example:
|
1535
|
+
#
|
1536
|
+
# sphinx.set_sort_mode(:relevance)
|
1537
|
+
# sphinx.add_query("hello world", "documents")
|
1538
|
+
#
|
1539
|
+
# sphinx.set_sort_mode(:attr_desc, :price)
|
1540
|
+
# sphinx.add_query("ipod", "products")
|
904
1541
|
#
|
905
|
-
#
|
906
|
-
#
|
907
|
-
|
1542
|
+
# sphinx.add_query("harry potter", "books")
|
1543
|
+
#
|
1544
|
+
# results = sphinx.run_queries
|
1545
|
+
#
|
1546
|
+
# With the code above, 1st query will search for "hello world"
|
1547
|
+
# in "documents" index and sort results by relevance, 2nd query
|
1548
|
+
# will search for "ipod" in "products" index and sort results
|
1549
|
+
# by price, and 3rd query will search for "harry potter" in
|
1550
|
+
# "books" index while still sorting by price. Note that 2nd
|
1551
|
+
# {#set_sort_mode} call does not affect the first query (because
|
1552
|
+
# it's already added) but affects both other subsequent queries.
|
1553
|
+
#
|
1554
|
+
# Additionally, any filters set up before an {#add_query} will
|
1555
|
+
# fall through to subsequent queries. So, if {#set_filter} is
|
1556
|
+
# called before the first query, the same filter will be in
|
1557
|
+
# place for the second (and subsequent) queries batched through
|
1558
|
+
# {#add_query} unless you call {#reset_filters} first. Alternatively,
|
1559
|
+
# you can add additional filters as well.
|
1560
|
+
#
|
1561
|
+
# This would also be true for grouping options and sorting options;
|
1562
|
+
# no current sorting, filtering, and grouping settings are affected
|
1563
|
+
# by this call; so subsequent queries will reuse current query settings.
|
1564
|
+
#
|
1565
|
+
# {#add_query} returns an index into an array of results that will
|
1566
|
+
# be returned from {#run_queries} call. It is simply a sequentially
|
1567
|
+
# increasing 0-based integer, ie. first call will return 0, second
|
1568
|
+
# will return 1, and so on. Just a small helper so you won't have
|
1569
|
+
# to track the indexes manualy if you need then.
|
1570
|
+
#
|
1571
|
+
# @param [String] query a query string.
|
1572
|
+
# @param [String] index an index name (or names).
|
1573
|
+
# @param [String] comment a comment to be sent to the query log.
|
1574
|
+
# @return [Integer] an index into an array of results that will
|
1575
|
+
# be returned from {#run_queries} call.
|
1576
|
+
#
|
1577
|
+
# @example
|
1578
|
+
# sphinx.add_query('some search text', '*', 'search page')
|
1579
|
+
#
|
1580
|
+
# @see http://www.sphinxsearch.com/docs/current.html#api-func-addquery Section 6.6.2, "AddQuery"
|
1581
|
+
# @see #query
|
1582
|
+
# @see #run_queries
|
1583
|
+
#
|
1584
|
+
def add_query(query, index = '*', comment = '')
|
908
1585
|
# build request
|
909
|
-
|
1586
|
+
|
910
1587
|
# mode and limits
|
911
1588
|
request = Request.new
|
912
1589
|
request.put_int @offset, @limit, @mode, @ranker, @sort
|
@@ -920,8 +1597,8 @@ module Sphinx
|
|
920
1597
|
# id64 range marker
|
921
1598
|
request.put_int 1
|
922
1599
|
# id64 range
|
923
|
-
request.put_int64 @min_id.to_i, @max_id.to_i
|
924
|
-
|
1600
|
+
request.put_int64 @min_id.to_i, @max_id.to_i
|
1601
|
+
|
925
1602
|
# filters
|
926
1603
|
request.put_int @filters.length
|
927
1604
|
@filters.each do |filter|
|
@@ -940,7 +1617,7 @@ module Sphinx
|
|
940
1617
|
end
|
941
1618
|
request.put_int filter['exclude'] ? 1 : 0
|
942
1619
|
end
|
943
|
-
|
1620
|
+
|
944
1621
|
# group-by clause, max-matches count, group-sort clause, cutoff count
|
945
1622
|
request.put_int @groupfunc
|
946
1623
|
request.put_string @groupby
|
@@ -948,7 +1625,7 @@ module Sphinx
|
|
948
1625
|
request.put_string @groupsort
|
949
1626
|
request.put_int @cutoff, @retrycount, @retrydelay
|
950
1627
|
request.put_string @groupdistinct
|
951
|
-
|
1628
|
+
|
952
1629
|
# anchor point
|
953
1630
|
if @anchor.empty?
|
954
1631
|
request.put_int 0
|
@@ -957,27 +1634,27 @@ module Sphinx
|
|
957
1634
|
request.put_string @anchor['attrlat'], @anchor['attrlong']
|
958
1635
|
request.put_float @anchor['lat'], @anchor['long']
|
959
1636
|
end
|
960
|
-
|
1637
|
+
|
961
1638
|
# per-index weights
|
962
1639
|
request.put_int @indexweights.length
|
963
1640
|
@indexweights.each do |idx, weight|
|
964
1641
|
request.put_string idx.to_s
|
965
1642
|
request.put_int weight
|
966
1643
|
end
|
967
|
-
|
1644
|
+
|
968
1645
|
# max query time
|
969
1646
|
request.put_int @maxquerytime
|
970
|
-
|
1647
|
+
|
971
1648
|
# per-field weights
|
972
1649
|
request.put_int @fieldweights.length
|
973
1650
|
@fieldweights.each do |field, weight|
|
974
1651
|
request.put_string field.to_s
|
975
1652
|
request.put_int weight
|
976
1653
|
end
|
977
|
-
|
1654
|
+
|
978
1655
|
# comment
|
979
1656
|
request.put_string comment
|
980
|
-
|
1657
|
+
|
981
1658
|
# attribute overrides
|
982
1659
|
request.put_int @overrides.length
|
983
1660
|
for entry in @overrides do
|
@@ -995,173 +1672,196 @@ module Sphinx
|
|
995
1672
|
end
|
996
1673
|
end
|
997
1674
|
end
|
998
|
-
|
1675
|
+
|
999
1676
|
# select-list
|
1000
1677
|
request.put_string @select
|
1001
|
-
|
1678
|
+
|
1002
1679
|
# store request to requests array
|
1003
1680
|
@reqs << request.to_s;
|
1004
1681
|
return @reqs.length - 1
|
1005
1682
|
end
|
1006
|
-
|
1007
|
-
|
1008
|
-
#
|
1009
|
-
#
|
1010
|
-
#
|
1011
|
-
#
|
1012
|
-
#
|
1013
|
-
#
|
1014
|
-
#
|
1015
|
-
#
|
1016
|
-
#
|
1017
|
-
#
|
1018
|
-
|
1683
|
+
alias :AddQuery :add_query
|
1684
|
+
|
1685
|
+
# Connect to searchd, runs a batch of all queries added using
|
1686
|
+
# {#add_query}, obtains and returns the result sets. Returns
|
1687
|
+
# +false+ and sets {#last_error} message on general error
|
1688
|
+
# (such as network I/O failure). Returns a plain array of
|
1689
|
+
# result sets on success.
|
1690
|
+
#
|
1691
|
+
# Each result set in the returned array is exactly the same as
|
1692
|
+
# the result set returned from {#query}.
|
1693
|
+
#
|
1694
|
+
# Note that the batch query request itself almost always succeds —
|
1695
|
+
# unless there's a network error, blocking index rotation in
|
1696
|
+
# progress, or another general failure which prevents the whole
|
1697
|
+
# request from being processed.
|
1698
|
+
#
|
1699
|
+
# However individual queries within the batch might very well
|
1700
|
+
# fail. In this case their respective result sets will contain
|
1701
|
+
# non-empty "error" message, but no matches or query statistics.
|
1702
|
+
# In the extreme case all queries within the batch could fail.
|
1703
|
+
# There still will be no general error reported, because API
|
1704
|
+
# was able to succesfully connect to searchd, submit the batch,
|
1705
|
+
# and receive the results — but every result set will have a
|
1706
|
+
# specific error message.
|
1707
|
+
#
|
1708
|
+
# @return [Array<Hash>] an +Array+ of +Hash+es which are exactly
|
1709
|
+
# the same as the result set returned from {#query}.
|
1710
|
+
#
|
1711
|
+
# @example
|
1712
|
+
# sphinx.add_query('some search text', '*', 'search page')
|
1713
|
+
# results = sphinx.run_queries
|
1714
|
+
#
|
1715
|
+
# @see http://www.sphinxsearch.com/docs/current.html#api-func-runqueries Section 6.6.3, "RunQueries"
|
1716
|
+
# @see #add_query
|
1717
|
+
#
|
1718
|
+
def run_queries
|
1019
1719
|
if @reqs.empty?
|
1020
|
-
@error = 'No queries defined, issue
|
1720
|
+
@error = 'No queries defined, issue add_query() first'
|
1021
1721
|
return false
|
1022
1722
|
end
|
1023
1723
|
|
1024
|
-
|
1025
|
-
nreqs = @reqs.length
|
1724
|
+
reqs, nreqs = @reqs.join(''), @reqs.length
|
1026
1725
|
@reqs = []
|
1027
|
-
response = perform_request(:search,
|
1028
|
-
|
1726
|
+
response = perform_request(:search, reqs, nreqs)
|
1727
|
+
|
1029
1728
|
# parse response
|
1030
|
-
|
1031
|
-
|
1032
|
-
|
1033
|
-
|
1034
|
-
|
1035
|
-
|
1036
|
-
|
1037
|
-
|
1038
|
-
|
1039
|
-
|
1040
|
-
|
1041
|
-
|
1042
|
-
if status != SEARCHD_OK
|
1043
|
-
message = response.get_string
|
1044
|
-
if status == SEARCHD_WARNING
|
1045
|
-
result['warning'] = message
|
1046
|
-
else
|
1047
|
-
result['error'] = message
|
1048
|
-
results << result
|
1049
|
-
next
|
1050
|
-
end
|
1051
|
-
end
|
1052
|
-
|
1053
|
-
# read schema
|
1054
|
-
fields = []
|
1055
|
-
attrs = {}
|
1056
|
-
attrs_names_in_order = []
|
1057
|
-
|
1058
|
-
nfields = response.get_int
|
1059
|
-
while nfields > 0
|
1060
|
-
nfields -= 1
|
1061
|
-
fields << response.get_string
|
1729
|
+
(1..nreqs).map do
|
1730
|
+
result = { 'error' => '', 'warning' => '' }
|
1731
|
+
|
1732
|
+
# extract status
|
1733
|
+
status = result['status'] = response.get_int
|
1734
|
+
if status != SEARCHD_OK
|
1735
|
+
message = response.get_string
|
1736
|
+
if status == SEARCHD_WARNING
|
1737
|
+
result['warning'] = message
|
1738
|
+
else
|
1739
|
+
result['error'] = message
|
1740
|
+
next result
|
1062
1741
|
end
|
1063
|
-
|
1064
|
-
|
1065
|
-
|
1066
|
-
|
1067
|
-
|
1068
|
-
|
1069
|
-
|
1070
|
-
|
1071
|
-
|
1742
|
+
end
|
1743
|
+
|
1744
|
+
# read schema
|
1745
|
+
nfields = response.get_int
|
1746
|
+
result['fields'] = (1..nfields).map { response.get_string }
|
1747
|
+
|
1748
|
+
attrs_names_in_order = []
|
1749
|
+
nattrs = response.get_int
|
1750
|
+
attrs = (1..nattrs).inject({}) do |hash, idx|
|
1751
|
+
name, type = response.get_string, response.get_int
|
1752
|
+
hash[name] = type
|
1753
|
+
attrs_names_in_order << name
|
1754
|
+
hash
|
1755
|
+
end
|
1756
|
+
result['attrs'] = attrs
|
1757
|
+
|
1758
|
+
# read match count
|
1759
|
+
count, id64 = response.get_ints(2)
|
1760
|
+
|
1761
|
+
# read matches
|
1762
|
+
result['matches'] = (1..count).map do
|
1763
|
+
doc, weight = if id64 == 0
|
1764
|
+
response.get_ints(2)
|
1765
|
+
else
|
1766
|
+
[response.get_int64, response.get_int]
|
1072
1767
|
end
|
1073
|
-
|
1074
|
-
|
1075
|
-
|
1076
|
-
|
1077
|
-
|
1078
|
-
|
1079
|
-
|
1080
|
-
|
1081
|
-
|
1082
|
-
|
1083
|
-
|
1084
|
-
|
1085
|
-
|
1086
|
-
|
1087
|
-
|
1088
|
-
|
1089
|
-
|
1090
|
-
|
1091
|
-
r = {} # This is a single result put in the result['matches'] array
|
1092
|
-
r['id'] = doc
|
1093
|
-
r['weight'] = weight
|
1094
|
-
attrs_names_in_order.each do |a|
|
1095
|
-
r['attrs'] ||= {}
|
1096
|
-
|
1097
|
-
case attrs[a]
|
1098
|
-
when SPH_ATTR_BIGINT
|
1099
|
-
# handle 64-bit ints
|
1100
|
-
r['attrs'][a] = response.get_int64
|
1101
|
-
when SPH_ATTR_FLOAT
|
1102
|
-
# handle floats
|
1103
|
-
r['attrs'][a] = response.get_float
|
1104
|
-
when SPH_ATTR_STRING
|
1105
|
-
r['attrs'][a] = response.get_string
|
1768
|
+
|
1769
|
+
# This is a single result put in the result['matches'] array
|
1770
|
+
match = { 'id' => doc, 'weight' => weight }
|
1771
|
+
match['attrs'] = attrs_names_in_order.inject({}) do |hash, name|
|
1772
|
+
hash[name] = case attrs[name]
|
1773
|
+
when SPH_ATTR_BIGINT
|
1774
|
+
# handle 64-bit ints
|
1775
|
+
response.get_int64
|
1776
|
+
when SPH_ATTR_FLOAT
|
1777
|
+
# handle floats
|
1778
|
+
response.get_float
|
1779
|
+
when SPH_ATTR_STRING
|
1780
|
+
response.get_string
|
1781
|
+
else
|
1782
|
+
# handle everything else as unsigned ints
|
1783
|
+
val = response.get_int
|
1784
|
+
if (attrs[name] & SPH_ATTR_MULTI) != 0
|
1785
|
+
(1..val).map { response.get_int }
|
1106
1786
|
else
|
1107
|
-
|
1108
|
-
|
1109
|
-
if (attrs[a] & SPH_ATTR_MULTI) != 0
|
1110
|
-
r['attrs'][a] = []
|
1111
|
-
1.upto(val) do
|
1112
|
-
r['attrs'][a] << response.get_int
|
1113
|
-
end
|
1114
|
-
else
|
1115
|
-
r['attrs'][a] = val
|
1116
|
-
end
|
1117
|
-
end
|
1787
|
+
val
|
1788
|
+
end
|
1118
1789
|
end
|
1119
|
-
|
1790
|
+
hash
|
1120
1791
|
end
|
1121
|
-
|
1122
|
-
result['time'] = '%.3f' % (msecs / 1000.0)
|
1123
|
-
|
1124
|
-
result['words'] = {}
|
1125
|
-
while words > 0
|
1126
|
-
words -= 1
|
1127
|
-
word = response.get_string
|
1128
|
-
docs, hits = response.get_ints(2)
|
1129
|
-
result['words'][word] = { 'docs' => docs, 'hits' => hits }
|
1130
|
-
end
|
1131
|
-
|
1132
|
-
results << result
|
1792
|
+
match
|
1133
1793
|
end
|
1134
|
-
|
1135
|
-
|
1136
|
-
|
1794
|
+
result['total'], result['total_found'], msecs = response.get_ints(3)
|
1795
|
+
result['time'] = '%.3f' % (msecs / 1000.0)
|
1796
|
+
|
1797
|
+
nwords = response.get_int
|
1798
|
+
result['words'] = (1..nwords).inject({}) do |hash, idx|
|
1799
|
+
word = response.get_string
|
1800
|
+
docs, hits = response.get_ints(2)
|
1801
|
+
hash[word] = { 'docs' => docs, 'hits' => hits }
|
1802
|
+
hash
|
1803
|
+
end
|
1804
|
+
|
1805
|
+
result
|
1137
1806
|
end
|
1138
|
-
|
1139
|
-
return results
|
1140
1807
|
end
|
1141
|
-
|
1142
|
-
|
1143
|
-
|
1144
|
-
#
|
1145
|
-
|
1146
|
-
|
1147
|
-
#
|
1148
|
-
#
|
1149
|
-
#
|
1150
|
-
#
|
1151
|
-
#
|
1152
|
-
#
|
1153
|
-
#
|
1154
|
-
#
|
1155
|
-
#
|
1156
|
-
#
|
1157
|
-
#
|
1158
|
-
#
|
1159
|
-
#
|
1160
|
-
#
|
1161
|
-
#
|
1162
|
-
#
|
1163
|
-
#
|
1164
|
-
|
1808
|
+
alias :RunQueries :run_queries
|
1809
|
+
|
1810
|
+
#=================================================================
|
1811
|
+
# Additional functionality
|
1812
|
+
#=================================================================
|
1813
|
+
|
1814
|
+
# Excerpts (snippets) builder function. Connects to searchd, asks
|
1815
|
+
# it to generate excerpts (snippets) from given documents, and
|
1816
|
+
# returns the results.
|
1817
|
+
#
|
1818
|
+
# +docs+ is a plain array of strings that carry the documents'
|
1819
|
+
# contents. +index+ is an index name string. Different settings
|
1820
|
+
# (such as charset, morphology, wordforms) from given index will
|
1821
|
+
# be used. +words+ is a string that contains the keywords to
|
1822
|
+
# highlight. They will be processed with respect to index settings.
|
1823
|
+
# For instance, if English stemming is enabled in the index,
|
1824
|
+
# "shoes" will be highlighted even if keyword is "shoe". Starting
|
1825
|
+
# with version 0.9.9-rc1, keywords can contain wildcards, that
|
1826
|
+
# work similarly to star-syntax available in queries.
|
1827
|
+
#
|
1828
|
+
# @param [Array<String>] docs an array of strings which represent
|
1829
|
+
# the documents' contents.
|
1830
|
+
# @param [String] index an index which settings will be used for
|
1831
|
+
# stemming, lexing and case folding.
|
1832
|
+
# @param [String] words a string which contains the words to highlight.
|
1833
|
+
# @param [Hash] opts a +Hash+ which contains additional optional
|
1834
|
+
# highlighting parameters.
|
1835
|
+
# @option opts [String] 'before_match' ("<b>") a string to insert before a
|
1836
|
+
# keyword match.
|
1837
|
+
# @option opts [String] 'after_match' ("</b>") a string to insert after a
|
1838
|
+
# keyword match.
|
1839
|
+
# @option opts [String] 'chunk_separator' (" ... ") a string to insert
|
1840
|
+
# between snippet chunks (passages).
|
1841
|
+
# @option opts [Integer] 'limit' (256) maximum snippet size, in symbols
|
1842
|
+
# (codepoints).
|
1843
|
+
# @option opts [Integer] 'around' (5) how many words to pick around
|
1844
|
+
# each matching keywords block.
|
1845
|
+
# @option opts [Boolean] 'exact_phrase' (false) whether to highlight exact
|
1846
|
+
# query phrase matches only instead of individual keywords.
|
1847
|
+
# @option opts [Boolean] 'single_passage' (false) whether to extract single
|
1848
|
+
# best passage only.
|
1849
|
+
# @option opts [Boolean] 'use_boundaries' (false) whether to extract
|
1850
|
+
# passages by phrase boundaries setup in tokenizer.
|
1851
|
+
# @option opts [Boolean] 'weight_order' (false) whether to sort the
|
1852
|
+
# extracted passages in order of relevance (decreasing weight),
|
1853
|
+
# or in order of appearance in the document (increasing position).
|
1854
|
+
# @return [Array<String>, false] a plain array of strings with
|
1855
|
+
# excerpts (snippets) on success; otherwise, +false+.
|
1856
|
+
#
|
1857
|
+
# @raise [ArgumentError] Occurred when parameters are invalid.
|
1858
|
+
#
|
1859
|
+
# @example
|
1860
|
+
# sphinx.build_excerpts(['hello world', 'hello me'], 'idx', 'hello')
|
1861
|
+
#
|
1862
|
+
# @see http://www.sphinxsearch.com/docs/current.html#api-func-buildexcerpts Section 6.7.1, "BuildExcerpts"
|
1863
|
+
#
|
1864
|
+
def build_excerpts(docs, index, words, opts = {})
|
1165
1865
|
raise ArgumentError, '"docs" argument must be Array' unless docs.kind_of?(Array)
|
1166
1866
|
raise ArgumentError, '"index" argument must be String' unless index.kind_of?(String) or index.kind_of?(Symbol)
|
1167
1867
|
raise ArgumentError, '"words" argument must be String' unless words.kind_of?(String)
|
@@ -1182,9 +1882,9 @@ module Sphinx
|
|
1182
1882
|
opts['use_boundaries'] ||= opts[:use_boundaries] || false
|
1183
1883
|
opts['weight_order'] ||= opts[:weight_order] || false
|
1184
1884
|
opts['query_mode'] ||= opts[:query_mode] || false
|
1185
|
-
|
1885
|
+
|
1186
1886
|
# build request
|
1187
|
-
|
1887
|
+
|
1188
1888
|
# v.1.0 req
|
1189
1889
|
flags = 1
|
1190
1890
|
flags |= 2 if opts['exact_phrase']
|
@@ -1192,47 +1892,71 @@ module Sphinx
|
|
1192
1892
|
flags |= 8 if opts['use_boundaries']
|
1193
1893
|
flags |= 16 if opts['weight_order']
|
1194
1894
|
flags |= 32 if opts['query_mode']
|
1195
|
-
|
1895
|
+
|
1196
1896
|
request = Request.new
|
1197
1897
|
request.put_int 0, flags # mode=0, flags=1 (remove spaces)
|
1198
1898
|
# req index
|
1199
1899
|
request.put_string index.to_s
|
1200
1900
|
# req words
|
1201
1901
|
request.put_string words
|
1202
|
-
|
1902
|
+
|
1203
1903
|
# options
|
1204
1904
|
request.put_string opts['before_match']
|
1205
1905
|
request.put_string opts['after_match']
|
1206
1906
|
request.put_string opts['chunk_separator']
|
1207
1907
|
request.put_int opts['limit'].to_i, opts['around'].to_i
|
1208
|
-
|
1908
|
+
|
1209
1909
|
# documents
|
1210
1910
|
request.put_int docs.size
|
1211
1911
|
request.put_string(*docs)
|
1212
|
-
|
1912
|
+
|
1213
1913
|
response = perform_request(:excerpt, request)
|
1214
|
-
|
1914
|
+
|
1215
1915
|
# parse response
|
1216
|
-
|
1217
|
-
res = []
|
1218
|
-
docs.each do |doc|
|
1219
|
-
res << response.get_string
|
1220
|
-
end
|
1221
|
-
rescue EOFError
|
1222
|
-
@error = 'incomplete reply'
|
1223
|
-
raise SphinxResponseError, @error
|
1224
|
-
end
|
1225
|
-
return res
|
1916
|
+
docs.map { response.get_string }
|
1226
1917
|
end
|
1227
|
-
|
1228
|
-
|
1229
|
-
#
|
1230
|
-
#
|
1231
|
-
|
1918
|
+
alias :BuildExcerpts :build_excerpts
|
1919
|
+
|
1920
|
+
# Extracts keywords from query using tokenizer settings for given
|
1921
|
+
# index, optionally with per-keyword occurrence statistics.
|
1922
|
+
# Returns an array of hashes with per-keyword information.
|
1923
|
+
#
|
1924
|
+
# +query+ is a query to extract keywords from. +index+ is a name of
|
1925
|
+
# the index to get tokenizing settings and keyword occurrence
|
1926
|
+
# statistics from. +hits+ is a boolean flag that indicates whether
|
1927
|
+
# keyword occurrence statistics are required.
|
1928
|
+
#
|
1929
|
+
# The result set consists of +Hash+es with the following keys and values:
|
1930
|
+
#
|
1931
|
+
# <tt>'tokenized'</tt>::
|
1932
|
+
# Tokenized keyword.
|
1933
|
+
# <tt>'normalized'</tt>::
|
1934
|
+
# Normalized keyword.
|
1935
|
+
# <tt>'docs'</tt>::
|
1936
|
+
# A number of documents where keyword is found (if +hits+ param is +true+).
|
1937
|
+
# <tt>'hits'</tt>::
|
1938
|
+
# A number of keywords occurrences among all documents (if +hits+ param is +true+).
|
1939
|
+
#
|
1940
|
+
# @param [String] query a query string.
|
1941
|
+
# @param [String] index an index to get tokenizing settings and
|
1942
|
+
# keyword occurrence statistics from.
|
1943
|
+
# @param [Boolean] hits indicates whether keyword occurrence
|
1944
|
+
# statistics are required.
|
1945
|
+
# @return [Array<Hash>] an +Array+ of +Hash+es in format specified
|
1946
|
+
# above.
|
1947
|
+
#
|
1948
|
+
# @raise [ArgumentError] Occurred when parameters are invalid.
|
1949
|
+
#
|
1950
|
+
# @example
|
1951
|
+
# keywords = sphinx.build_keywords("this.is.my query", "test1", false)
|
1952
|
+
#
|
1953
|
+
# @see http://www.sphinxsearch.com/docs/current.html#api-func-buildkeywords Section 6.7.3, "BuildKeywords"
|
1954
|
+
#
|
1955
|
+
def build_keywords(query, index, hits)
|
1232
1956
|
raise ArgumentError, '"query" argument must be String' unless query.kind_of?(String)
|
1233
1957
|
raise ArgumentError, '"index" argument must be String' unless index.kind_of?(String) or index.kind_of?(Symbol)
|
1234
1958
|
raise ArgumentError, '"hits" argument must be Boolean' unless hits.kind_of?(TrueClass) or hits.kind_of?(FalseClass)
|
1235
|
-
|
1959
|
+
|
1236
1960
|
# build request
|
1237
1961
|
request = Request.new
|
1238
1962
|
# v.1.0 req
|
@@ -1241,53 +1965,79 @@ module Sphinx
|
|
1241
1965
|
request.put_int hits ? 1 : 0
|
1242
1966
|
|
1243
1967
|
response = perform_request(:keywords, request)
|
1244
|
-
|
1968
|
+
|
1245
1969
|
# parse response
|
1246
|
-
|
1247
|
-
|
1248
|
-
|
1249
|
-
|
1250
|
-
|
1251
|
-
|
1252
|
-
|
1253
|
-
|
1254
|
-
|
1255
|
-
|
1256
|
-
res << entry
|
1257
|
-
end
|
1258
|
-
rescue EOFError
|
1259
|
-
@error = 'incomplete reply'
|
1260
|
-
raise SphinxResponseError, @error
|
1970
|
+
nwords = response.get_int
|
1971
|
+
(0...nwords).map do
|
1972
|
+
tokenized = response.get_string
|
1973
|
+
normalized = response.get_string
|
1974
|
+
|
1975
|
+
entry = { 'tokenized' => tokenized, 'normalized' => normalized }
|
1976
|
+
entry['docs'], entry['hits'] = response.get_ints(2) if hits
|
1977
|
+
|
1978
|
+
entry
|
1261
1979
|
end
|
1262
|
-
|
1263
|
-
return res
|
1264
1980
|
end
|
1981
|
+
alias :BuildKeywords :build_keywords
|
1265
1982
|
|
1266
|
-
#
|
1983
|
+
# Instantly updates given attribute values in given documents.
|
1984
|
+
# Returns number of actually updated documents (0 or more) on
|
1985
|
+
# success, or -1 on failure.
|
1986
|
+
#
|
1987
|
+
# +index+ is a name of the index (or indexes) to be updated.
|
1988
|
+
# +attrs+ is a plain array with string attribute names, listing
|
1989
|
+
# attributes that are updated. +values+ is a Hash where key is
|
1990
|
+
# document ID, and value is a plain array of new attribute values.
|
1991
|
+
#
|
1992
|
+
# +index+ can be either a single index name or a list, like in
|
1993
|
+
# {#query}. Unlike {#query}, wildcard is not allowed and all the
|
1994
|
+
# indexes to update must be specified explicitly. The list of
|
1995
|
+
# indexes can include distributed index names. Updates on
|
1996
|
+
# distributed indexes will be pushed to all agents.
|
1997
|
+
#
|
1998
|
+
# The updates only work with docinfo=extern storage strategy.
|
1999
|
+
# They are very fast because they're working fully in RAM, but
|
2000
|
+
# they can also be made persistent: updates are saved on disk
|
2001
|
+
# on clean searchd shutdown initiated by SIGTERM signal. With
|
2002
|
+
# additional restrictions, updates are also possible on MVA
|
2003
|
+
# attributes; refer to mva_updates_pool directive for details.
|
2004
|
+
#
|
2005
|
+
# The first sample statement will update document 1 in index
|
2006
|
+
# "test1", setting "group_id" to 456. The second one will update
|
2007
|
+
# documents 1001, 1002 and 1003 in index "products". For document
|
2008
|
+
# 1001, the new price will be set to 123 and the new amount in
|
2009
|
+
# stock to 5; for document 1002, the new price will be 37 and the
|
2010
|
+
# new amount will be 11; etc. The third one updates document 1
|
2011
|
+
# in index "test2", setting MVA attribute "group_id" to [456, 789].
|
2012
|
+
#
|
2013
|
+
# @example
|
2014
|
+
# sphinx.update_attributes("test1", ["group_id"], { 1 => [456] });
|
2015
|
+
# sphinx.update_attributes("products", ["price", "amount_in_stock"],
|
2016
|
+
# { 1001 => [123, 5], 1002 => [37, 11], 1003 => [25, 129] });
|
2017
|
+
# sphinx.update_attributes('test2', ['group_id'], { 1 => [[456, 789]] }, true)
|
1267
2018
|
#
|
1268
|
-
#
|
1269
|
-
#
|
1270
|
-
#
|
1271
|
-
#
|
1272
|
-
#
|
2019
|
+
# @param [String] index a name of the index to be updated.
|
2020
|
+
# @param [Array<String>] attrs an array of attribute name strings.
|
2021
|
+
# @param [Hash] values is a hash where key is document id, and
|
2022
|
+
# value is an array of new attribute values.
|
2023
|
+
# @param [Boolean] mva indicating whether to update MVA.
|
2024
|
+
# @return [Integer] number of actually updated documents (0 or more) on success,
|
2025
|
+
# -1 on failure.
|
1273
2026
|
#
|
1274
|
-
#
|
1275
|
-
# Returns -1 on failure.
|
2027
|
+
# @raise [ArgumentError] Occurred when parameters are invalid.
|
1276
2028
|
#
|
1277
|
-
#
|
1278
|
-
# sphinx.UpdateAttributes('test1', ['group_id'], { 1 => [456] })
|
1279
|
-
# sphinx.UpdateAttributes('test1', ['group_id'], { 1 => [[456, 789]] }, true)
|
2029
|
+
# @see http://www.sphinxsearch.com/docs/current.html#api-func-updateatttributes Section 6.7.2, "UpdateAttributes"
|
1280
2030
|
#
|
1281
|
-
def
|
2031
|
+
def update_attributes(index, attrs, values, mva = false)
|
1282
2032
|
# verify everything
|
1283
2033
|
raise ArgumentError, '"index" argument must be String' unless index.kind_of?(String) or index.kind_of?(Symbol)
|
1284
2034
|
raise ArgumentError, '"mva" argument must be Boolean' unless mva.kind_of?(TrueClass) or mva.kind_of?(FalseClass)
|
1285
|
-
|
2035
|
+
|
1286
2036
|
raise ArgumentError, '"attrs" argument must be Array' unless attrs.kind_of?(Array)
|
1287
2037
|
attrs.each do |attr|
|
1288
2038
|
raise ArgumentError, '"attrs" argument must be Array of Strings' unless attr.kind_of?(String) or attr.kind_of?(Symbol)
|
1289
2039
|
end
|
1290
|
-
|
2040
|
+
|
1291
2041
|
raise ArgumentError, '"values" argument must be Hash' unless values.kind_of?(Hash)
|
1292
2042
|
values.each do |id, entry|
|
1293
2043
|
raise ArgumentError, '"values" argument must be Hash map of Integer to Array' unless id.respond_to?(:integer?) and id.integer?
|
@@ -1304,17 +2054,17 @@ module Sphinx
|
|
1304
2054
|
end
|
1305
2055
|
end
|
1306
2056
|
end
|
1307
|
-
|
2057
|
+
|
1308
2058
|
# build request
|
1309
2059
|
request = Request.new
|
1310
2060
|
request.put_string index
|
1311
|
-
|
2061
|
+
|
1312
2062
|
request.put_int attrs.length
|
1313
2063
|
for attr in attrs
|
1314
2064
|
request.put_string attr
|
1315
2065
|
request.put_int mva ? 1 : 0
|
1316
2066
|
end
|
1317
|
-
|
2067
|
+
|
1318
2068
|
request.put_int values.length
|
1319
2069
|
values.each do |id, entry|
|
1320
2070
|
request.put_int64 id
|
@@ -1324,33 +2074,89 @@ module Sphinx
|
|
1324
2074
|
request.put_int(*entry)
|
1325
2075
|
end
|
1326
2076
|
end
|
1327
|
-
|
2077
|
+
|
1328
2078
|
response = perform_request(:update, request)
|
1329
|
-
|
2079
|
+
|
2080
|
+
# parse response
|
2081
|
+
response.get_int
|
2082
|
+
end
|
2083
|
+
alias :UpdateAttributes :update_attributes
|
2084
|
+
|
2085
|
+
# Queries searchd status, and returns an array of status variable name
|
2086
|
+
# and value pairs.
|
2087
|
+
#
|
2088
|
+
# @return [Array<Array>] a table containing searchd status information.
|
2089
|
+
#
|
2090
|
+
# @example
|
2091
|
+
# status = sphinx.status
|
2092
|
+
# puts status.map { |key, value| "#{key.rjust(20)}: #{value}" }
|
2093
|
+
#
|
2094
|
+
def status
|
2095
|
+
request = Request.new
|
2096
|
+
request.put_int(1)
|
2097
|
+
response = perform_request(:status, request)
|
2098
|
+
|
2099
|
+
# parse response
|
2100
|
+
rows, cols = response.get_ints(2)
|
2101
|
+
(0...rows).map do
|
2102
|
+
(0...cols).map { response.get_string }
|
2103
|
+
end
|
2104
|
+
end
|
2105
|
+
alias :Status :status
|
2106
|
+
|
2107
|
+
# Force attribute flush, and block until it completes.
|
2108
|
+
#
|
2109
|
+
# @return [Integer] current internal flush tag on success, -1 on failure.
|
2110
|
+
#
|
2111
|
+
# @example
|
2112
|
+
# sphinx.flush_attrs
|
2113
|
+
#
|
2114
|
+
def flush_attrs
|
2115
|
+
request = Request.new
|
2116
|
+
response = perform_request(:flushattrs, request)
|
2117
|
+
|
1330
2118
|
# parse response
|
1331
2119
|
begin
|
1332
|
-
|
2120
|
+
response.get_int
|
1333
2121
|
rescue EOFError
|
1334
|
-
|
1335
|
-
raise SphinxResponseError, @error
|
2122
|
+
-1
|
1336
2123
|
end
|
1337
2124
|
end
|
1338
|
-
|
1339
|
-
|
1340
|
-
|
2125
|
+
alias :FlushAttrs :flush_attrs
|
2126
|
+
|
2127
|
+
#=================================================================
|
2128
|
+
# Persistent connections
|
2129
|
+
#=================================================================
|
2130
|
+
|
1341
2131
|
# Opens persistent connection to the server.
|
1342
2132
|
#
|
1343
|
-
|
2133
|
+
# This method could be used only when a single searchd server
|
2134
|
+
# configured.
|
2135
|
+
#
|
2136
|
+
# @return [Boolean] +true+ when persistent connection has been
|
2137
|
+
# established; otherwise, +false+.
|
2138
|
+
#
|
2139
|
+
# @example
|
2140
|
+
# begin
|
2141
|
+
# sphinx.open
|
2142
|
+
# # perform several requests
|
2143
|
+
# ensure
|
2144
|
+
# sphinx.close
|
2145
|
+
# end
|
2146
|
+
#
|
2147
|
+
# @see #close
|
2148
|
+
#
|
2149
|
+
def open
|
1344
2150
|
if @servers.size > 1
|
1345
2151
|
@error = 'too many servers. persistent socket allowed only for a single server.'
|
1346
2152
|
return false
|
1347
2153
|
end
|
1348
|
-
|
2154
|
+
|
1349
2155
|
if @servers.first.persistent?
|
1350
2156
|
@error = 'already connected'
|
1351
2157
|
return false;
|
1352
2158
|
end
|
1353
|
-
|
2159
|
+
|
1354
2160
|
request = Request.new
|
1355
2161
|
request.put_int(1)
|
1356
2162
|
|
@@ -1360,85 +2166,64 @@ module Sphinx
|
|
1360
2166
|
|
1361
2167
|
true
|
1362
2168
|
end
|
1363
|
-
|
2169
|
+
alias :Open :open
|
2170
|
+
|
1364
2171
|
# Closes previously opened persistent connection.
|
1365
2172
|
#
|
1366
|
-
|
2173
|
+
# This method could be used only when a single searchd server
|
2174
|
+
# configured.
|
2175
|
+
#
|
2176
|
+
# @return [Boolean] +true+ when persistent connection has been
|
2177
|
+
# closed; otherwise, +false+.
|
2178
|
+
#
|
2179
|
+
# @example
|
2180
|
+
# begin
|
2181
|
+
# sphinx.open
|
2182
|
+
# # perform several requests
|
2183
|
+
# ensure
|
2184
|
+
# sphinx.close
|
2185
|
+
# end
|
2186
|
+
#
|
2187
|
+
# @see #open
|
2188
|
+
#
|
2189
|
+
def close
|
1367
2190
|
if @servers.size > 1
|
1368
2191
|
@error = 'too many servers. persistent socket allowed only for a single server.'
|
1369
2192
|
return false
|
1370
2193
|
end
|
1371
|
-
|
2194
|
+
|
1372
2195
|
unless @servers.first.persistent?
|
1373
2196
|
@error = 'not connected'
|
1374
2197
|
return false;
|
1375
2198
|
end
|
1376
|
-
|
1377
|
-
@servers.first.close_persistent!
|
1378
|
-
end
|
1379
|
-
|
1380
|
-
# Queries searchd status, and returns an array of status variable name
|
1381
|
-
# and value pairs.
|
1382
|
-
#
|
1383
|
-
# Usage example:
|
1384
|
-
#
|
1385
|
-
# status = sphinx.Status
|
1386
|
-
# puts status.map { |key, value| "#{key.rjust(20)}: #{value}" }
|
1387
|
-
#
|
1388
|
-
def Status
|
1389
|
-
request = Request.new
|
1390
|
-
request.put_int(1)
|
1391
|
-
response = perform_request(:status, request)
|
1392
2199
|
|
1393
|
-
|
1394
|
-
begin
|
1395
|
-
rows, cols = response.get_ints(2)
|
1396
|
-
|
1397
|
-
res = []
|
1398
|
-
0.upto(rows - 1) do |i|
|
1399
|
-
res[i] = []
|
1400
|
-
0.upto(cols - 1) do |j|
|
1401
|
-
res[i] << response.get_string
|
1402
|
-
end
|
1403
|
-
end
|
1404
|
-
rescue EOFError
|
1405
|
-
@error = 'incomplete reply'
|
1406
|
-
raise SphinxResponseError, @error
|
1407
|
-
end
|
1408
|
-
|
1409
|
-
res
|
2200
|
+
@servers.first.close_persistent!
|
1410
2201
|
end
|
1411
|
-
|
1412
|
-
def FlushAttrs
|
1413
|
-
request = Request.new
|
1414
|
-
response = perform_request(:flushattrs, request)
|
2202
|
+
alias :Close :close
|
1415
2203
|
|
1416
|
-
# parse response
|
1417
|
-
begin
|
1418
|
-
response.get_int
|
1419
|
-
rescue EOFError
|
1420
|
-
-1
|
1421
|
-
end
|
1422
|
-
end
|
1423
|
-
|
1424
2204
|
protected
|
1425
|
-
|
2205
|
+
|
1426
2206
|
# Connect, send query, get response.
|
1427
2207
|
#
|
1428
2208
|
# Use this method to communicate with Sphinx server. It ensures connection
|
1429
2209
|
# will be instantiated properly, all headers will be generated properly, etc.
|
1430
2210
|
#
|
1431
|
-
#
|
1432
|
-
# * +command+ -- searchd command to perform (<tt>:search</tt>, <tt>:excerpt</tt>,
|
2211
|
+
# @param [Symbol, String] command searchd command to perform (<tt>:search</tt>, <tt>:excerpt</tt>,
|
1433
2212
|
# <tt>:update</tt>, <tt>:keywords</tt>, <tt>:persist</tt>, <tt>:status</tt>,
|
1434
2213
|
# <tt>:query</tt>, <tt>:flushattrs</tt>. See <tt>SEARCHD_COMMAND_*</tt> for details).
|
1435
|
-
#
|
1436
|
-
#
|
1437
|
-
#
|
1438
|
-
#
|
1439
|
-
#
|
2214
|
+
# @param [Sphinx::Request] request contains request body.
|
2215
|
+
# @param [nil, Integer] additional additional integer data to be placed between header and body.
|
2216
|
+
#
|
2217
|
+
# @yield if block given, response will not be parsed, plain socket
|
2218
|
+
# will be yielded instead. This is special mode used for
|
2219
|
+
# persistent connections, do not use for other tasks.
|
2220
|
+
# @yieldparam [Sphinx::Server] server a server where request was performed on.
|
2221
|
+
# @yieldparam [Sphinx::BufferedIO] socket a socket used to perform the request.
|
2222
|
+
# @return [Sphinx::Response] contains response body.
|
1440
2223
|
#
|
1441
|
-
|
2224
|
+
# @see #parse_response
|
2225
|
+
#
|
2226
|
+
def perform_request(command, request, additional = nil)
|
1442
2227
|
with_server do |server|
|
1443
2228
|
cmd = command.to_s.upcase
|
1444
2229
|
command_id = Sphinx::Client.const_get("SEARCHD_COMMAND_#{cmd}")
|
@@ -1465,26 +2250,31 @@ module Sphinx
|
|
1465
2250
|
#
|
1466
2251
|
# There are several exceptions which could be thrown in this method:
|
1467
2252
|
#
|
1468
|
-
#
|
1469
|
-
#
|
1470
|
-
#
|
1471
|
-
#
|
1472
|
-
#
|
2253
|
+
# @param [Sphinx::BufferedIO] socket an input stream object.
|
2254
|
+
# @param [Integer] client_version a command version which client supports.
|
2255
|
+
# @return [Sphinx::Response] could be used for context-based
|
2256
|
+
# parsing of reply from the server.
|
2257
|
+
#
|
2258
|
+
# @raise [SystemCallError, SocketError] should be handled by caller (see {#with_socket}).
|
2259
|
+
# @raise [SphinxResponseError] incomplete reply from searchd.
|
2260
|
+
# @raise [SphinxInternalError] searchd internal error.
|
2261
|
+
# @raise [SphinxTemporaryError] searchd temporary error.
|
2262
|
+
# @raise [SphinxUnknownError] searchd unknown error.
|
1473
2263
|
#
|
1474
|
-
#
|
1475
|
-
#
|
2264
|
+
# @see #with_socket
|
2265
|
+
# @private
|
1476
2266
|
#
|
1477
2267
|
def parse_response(socket, client_version)
|
1478
2268
|
response = ''
|
1479
2269
|
status = ver = len = 0
|
1480
|
-
|
1481
|
-
# Read server reply from server. All exceptions are handled by
|
2270
|
+
|
2271
|
+
# Read server reply from server. All exceptions are handled by {#with_socket}.
|
1482
2272
|
header = socket.read(8)
|
1483
2273
|
if header.length == 8
|
1484
2274
|
status, ver, len = header.unpack('n2N')
|
1485
2275
|
response = socket.read(len) if len > 0
|
1486
2276
|
end
|
1487
|
-
|
2277
|
+
|
1488
2278
|
# check response
|
1489
2279
|
read = response.length
|
1490
2280
|
if response.empty? or read != len.to_i
|
@@ -1493,7 +2283,7 @@ module Sphinx
|
|
1493
2283
|
: 'received zero-sized searchd response'
|
1494
2284
|
raise SphinxResponseError, error
|
1495
2285
|
end
|
1496
|
-
|
2286
|
+
|
1497
2287
|
# check status
|
1498
2288
|
if (status == SEARCHD_WARNING)
|
1499
2289
|
wlen = response[0, 4].unpack('N*').first
|
@@ -1505,34 +2295,40 @@ module Sphinx
|
|
1505
2295
|
error = 'searchd error: ' + response[4, response.length - 4]
|
1506
2296
|
raise SphinxInternalError, error
|
1507
2297
|
end
|
1508
|
-
|
2298
|
+
|
1509
2299
|
if status == SEARCHD_RETRY
|
1510
2300
|
error = 'temporary searchd error: ' + response[4, response.length - 4]
|
1511
2301
|
raise SphinxTemporaryError, error
|
1512
2302
|
end
|
1513
|
-
|
2303
|
+
|
1514
2304
|
unless status == SEARCHD_OK
|
1515
2305
|
error = "unknown status code: '#{status}'"
|
1516
2306
|
raise SphinxUnknownError, error
|
1517
2307
|
end
|
1518
|
-
|
2308
|
+
|
1519
2309
|
# check version
|
1520
2310
|
if ver < client_version
|
1521
2311
|
@warning = "searchd command v.#{ver >> 8}.#{ver & 0xff} older than client's " +
|
1522
2312
|
"v.#{client_version >> 8}.#{client_version & 0xff}, some options might not work"
|
1523
2313
|
end
|
1524
|
-
|
2314
|
+
|
1525
2315
|
Response.new(response)
|
1526
2316
|
end
|
1527
|
-
|
2317
|
+
|
1528
2318
|
# This is internal method which selects next server (round-robin)
|
1529
2319
|
# and yields it to the block passed.
|
1530
2320
|
#
|
1531
2321
|
# In case of connection error, it will try next server several times
|
1532
|
-
# (see
|
1533
|
-
# it will set
|
1534
|
-
#
|
1535
|
-
#
|
2322
|
+
# (see {#set_connect_timeout} method details). If all servers are down,
|
2323
|
+
# it will set error attribute (could be retrieved with {#last_error}
|
2324
|
+
# method) with the last exception message, and {#connect_error?}
|
2325
|
+
# method will return true. Also, {SphinxConnectError} exception
|
2326
|
+
# will be raised.
|
2327
|
+
#
|
2328
|
+
# @yield a block which performs request on a given server.
|
2329
|
+
# @yieldparam [Sphinx::Server] server contains information
|
2330
|
+
# about the server to perform request on.
|
2331
|
+
# @raise [SphinxConnectError] on any connection error.
|
1536
2332
|
#
|
1537
2333
|
def with_server
|
1538
2334
|
attempts = @retries
|
@@ -1552,29 +2348,39 @@ module Sphinx
|
|
1552
2348
|
raise
|
1553
2349
|
end
|
1554
2350
|
end
|
1555
|
-
|
2351
|
+
|
1556
2352
|
# This is internal method which retrieves socket for a given server,
|
1557
2353
|
# initiates Sphinx session, and yields this socket to a block passed.
|
1558
2354
|
#
|
1559
|
-
# In case of any problems with session initiation,
|
2355
|
+
# In case of any problems with session initiation, {SphinxConnectError}
|
1560
2356
|
# will be raised, because this is part of connection establishing. See
|
1561
|
-
#
|
2357
|
+
# {#with_server} method details to get more infromation about how this
|
1562
2358
|
# exception is handled.
|
1563
2359
|
#
|
1564
2360
|
# Socket retrieving routine is wrapped in a block with it's own
|
1565
|
-
# timeout value (see
|
1566
|
-
#
|
2361
|
+
# timeout value (see {#set_connect_timeout}). This is done in
|
2362
|
+
# {Server#get_socket} method, so check it for details.
|
1567
2363
|
#
|
1568
2364
|
# Request execution is wrapped with block with another timeout
|
1569
|
-
# (see
|
2365
|
+
# (see {#set_request_timeout}). This ensures no Sphinx request will
|
1570
2366
|
# take unreasonable time.
|
1571
2367
|
#
|
1572
2368
|
# In case of any Sphinx error (incomplete reply, internal or temporary
|
1573
2369
|
# error), connection to the server will be re-established, and request
|
1574
|
-
# will be retried (see
|
2370
|
+
# will be retried (see {#set_request_timeout}). Of course, if connection
|
1575
2371
|
# could not be established, next server will be selected (see explanation
|
1576
2372
|
# above).
|
1577
2373
|
#
|
2374
|
+
# @param [Sphinx::Server] server contains information
|
2375
|
+
# about the server to perform request on.
|
2376
|
+
# @yield a block which will actually perform the request.
|
2377
|
+
# @yieldparam [Sphinx::BufferedIO] socket a socket used to
|
2378
|
+
# perform the request.
|
2379
|
+
#
|
2380
|
+
# @raise [SphinxResponseError, SphinxInternalError, SphinxTemporaryError, SphinxUnknownError]
|
2381
|
+
# on any response error.
|
2382
|
+
# @raise [SphinxConnectError] on any connection error.
|
2383
|
+
#
|
1578
2384
|
def with_socket(server)
|
1579
2385
|
attempts = @reqretries
|
1580
2386
|
socket = nil
|
@@ -1612,14 +2418,14 @@ module Sphinx
|
|
1612
2418
|
new_e.set_backtrace(e.backtrace)
|
1613
2419
|
e = new_e
|
1614
2420
|
end
|
1615
|
-
|
2421
|
+
|
1616
2422
|
# Close previously opened socket (in case of it has been really opened)
|
1617
2423
|
server.free_socket(socket)
|
1618
2424
|
|
1619
2425
|
# Request error! Do we need to try it again?
|
1620
2426
|
attempts -= 1
|
1621
2427
|
retry if attempts > 0
|
1622
|
-
|
2428
|
+
|
1623
2429
|
# Re-raise original exception
|
1624
2430
|
@error = e.message
|
1625
2431
|
raise e
|