textacular 3.0.0
Sign up to get free protection for your applications and to get access to all the features.
- data/CHANGELOG.md +164 -0
- data/Gemfile +3 -0
- data/README.md +143 -0
- data/Rakefile +154 -0
- data/lib/textacular.rb +227 -0
- data/lib/textacular/full_text_indexer.rb +79 -0
- data/lib/textacular/postgres_module_installer.rb +57 -0
- data/lib/textacular/rails.rb +10 -0
- data/lib/textacular/searchable.rb +20 -0
- data/lib/textacular/tasks.rb +18 -0
- data/lib/textacular/version.rb +7 -0
- data/spec/config.yml.example +4 -0
- data/spec/fixtures/character.rb +0 -0
- data/spec/fixtures/game.rb +9 -0
- data/spec/fixtures/webcomic.rb +9 -0
- data/spec/spec_helper.rb +78 -0
- data/spec/textacular/searchable_spec.rb +92 -0
- data/spec/textacular_spec.rb +216 -0
- metadata +178 -0
data/lib/textacular.rb
ADDED
@@ -0,0 +1,227 @@
|
|
1
|
+
require 'active_record'
|
2
|
+
|
3
|
+
require 'textacular/version'
|
4
|
+
|
5
|
+
module Textacular
|
6
|
+
def self.searchable_language
|
7
|
+
'english'
|
8
|
+
end
|
9
|
+
|
10
|
+
def search(query = "", exclusive = true)
|
11
|
+
basic_search(query, exclusive)
|
12
|
+
end
|
13
|
+
|
14
|
+
def basic_search(query = "", exclusive = true)
|
15
|
+
exclusive, query = munge_exclusive_and_query(exclusive, query)
|
16
|
+
parsed_query_hash = parse_query_hash(query)
|
17
|
+
similarities, conditions = basic_similarities_and_conditions(parsed_query_hash)
|
18
|
+
assemble_query(similarities, conditions, exclusive)
|
19
|
+
end
|
20
|
+
|
21
|
+
def advanced_search(query = "", exclusive = true)
|
22
|
+
exclusive, query = munge_exclusive_and_query(exclusive, query)
|
23
|
+
parsed_query_hash = parse_query_hash(query)
|
24
|
+
similarities, conditions = advanced_similarities_and_conditions(parsed_query_hash)
|
25
|
+
assemble_query(similarities, conditions, exclusive)
|
26
|
+
end
|
27
|
+
|
28
|
+
def fuzzy_search(query = '', exclusive = true)
|
29
|
+
exclusive, query = munge_exclusive_and_query(exclusive, query)
|
30
|
+
parsed_query_hash = parse_query_hash(query)
|
31
|
+
similarities, conditions = fuzzy_similarities_and_conditions(parsed_query_hash)
|
32
|
+
assemble_query(similarities, conditions, exclusive)
|
33
|
+
end
|
34
|
+
|
35
|
+
def method_missing(method, *search_terms)
|
36
|
+
return super if self == ActiveRecord::Base
|
37
|
+
if Helper.dynamic_search_method?(method, self.columns)
|
38
|
+
exclusive = Helper.exclusive_dynamic_search_method?(method, self.columns)
|
39
|
+
columns = exclusive ? Helper.exclusive_dynamic_search_columns(method) : Helper.inclusive_dynamic_search_columns(method)
|
40
|
+
metaclass = class << self; self; end
|
41
|
+
metaclass.__send__(:define_method, method) do |*args|
|
42
|
+
query = columns.inject({}) do |query, column|
|
43
|
+
query.merge column => args.shift
|
44
|
+
end
|
45
|
+
self.send(Helper.search_type(method), query, exclusive)
|
46
|
+
end
|
47
|
+
__send__(method, *search_terms, exclusive)
|
48
|
+
else
|
49
|
+
super
|
50
|
+
end
|
51
|
+
rescue ActiveRecord::StatementInvalid
|
52
|
+
super
|
53
|
+
end
|
54
|
+
|
55
|
+
def respond_to?(method, include_private = false)
|
56
|
+
return super if self == ActiveRecord::Base
|
57
|
+
Helper.dynamic_search_method?(method, self.columns) or super
|
58
|
+
rescue StandardError
|
59
|
+
super
|
60
|
+
end
|
61
|
+
|
62
|
+
private
|
63
|
+
|
64
|
+
def munge_exclusive_and_query(exclusive, query)
|
65
|
+
unless query.is_a?(Hash)
|
66
|
+
exclusive = false
|
67
|
+
query = searchable_columns.inject({}) do |terms, column|
|
68
|
+
terms.merge column => query.to_s
|
69
|
+
end
|
70
|
+
end
|
71
|
+
|
72
|
+
[exclusive, query]
|
73
|
+
end
|
74
|
+
|
75
|
+
def parse_query_hash(query, table_name = quoted_table_name)
|
76
|
+
table_name = connection.quote_table_name(table_name)
|
77
|
+
|
78
|
+
results = []
|
79
|
+
|
80
|
+
query.each do |column_or_table, search_term|
|
81
|
+
if search_term.is_a?(Hash)
|
82
|
+
results += parse_query_hash(search_term, column_or_table)
|
83
|
+
else
|
84
|
+
column = connection.quote_column_name(column_or_table)
|
85
|
+
search_term = connection.quote normalize(Helper.normalize(search_term))
|
86
|
+
|
87
|
+
results << [table_name, column, search_term]
|
88
|
+
end
|
89
|
+
end
|
90
|
+
|
91
|
+
results
|
92
|
+
end
|
93
|
+
|
94
|
+
def basic_similarities_and_conditions(parsed_query_hash)
|
95
|
+
parsed_query_hash.inject([[], []]) do |(similarities, conditions), query_args|
|
96
|
+
similarities << basic_similarity_string(*query_args)
|
97
|
+
conditions << basic_condition_string(*query_args)
|
98
|
+
|
99
|
+
[similarities, conditions]
|
100
|
+
end
|
101
|
+
end
|
102
|
+
|
103
|
+
def basic_similarity_string(table_name, column, search_term)
|
104
|
+
"ts_rank(to_tsvector(#{quoted_language}, #{table_name}.#{column}::text), plainto_tsquery(#{quoted_language}, #{search_term}::text))"
|
105
|
+
end
|
106
|
+
|
107
|
+
def basic_condition_string(table_name, column, search_term)
|
108
|
+
"to_tsvector(#{quoted_language}, #{table_name}.#{column}::text) @@ plainto_tsquery(#{quoted_language}, #{search_term}::text)"
|
109
|
+
end
|
110
|
+
|
111
|
+
def advanced_similarities_and_conditions(parsed_query_hash)
|
112
|
+
parsed_query_hash.inject([[], []]) do |(similarities, conditions), query_args|
|
113
|
+
similarities << advanced_similarity_string(*query_args)
|
114
|
+
conditions << advanced_condition_string(*query_args)
|
115
|
+
|
116
|
+
[similarities, conditions]
|
117
|
+
end
|
118
|
+
end
|
119
|
+
|
120
|
+
def advanced_similarity_string(table_name, column, search_term)
|
121
|
+
"ts_rank(to_tsvector(#{quoted_language}, #{table_name}.#{column}::text), to_tsquery(#{quoted_language}, #{search_term}::text))"
|
122
|
+
end
|
123
|
+
|
124
|
+
def advanced_condition_string(table_name, column, search_term)
|
125
|
+
"to_tsvector(#{quoted_language}, #{table_name}.#{column}::text) @@ to_tsquery(#{quoted_language}, #{search_term}::text)"
|
126
|
+
end
|
127
|
+
|
128
|
+
def fuzzy_similarities_and_conditions(parsed_query_hash)
|
129
|
+
parsed_query_hash.inject([[], []]) do |(similarities, conditions), query_args|
|
130
|
+
similarities << fuzzy_similarity_string(*query_args)
|
131
|
+
conditions << fuzzy_condition_string(*query_args)
|
132
|
+
|
133
|
+
[similarities, conditions]
|
134
|
+
end
|
135
|
+
end
|
136
|
+
|
137
|
+
def fuzzy_similarity_string(table_name, column, search_term)
|
138
|
+
"similarity(#{table_name}.#{column}, #{search_term})"
|
139
|
+
end
|
140
|
+
|
141
|
+
def fuzzy_condition_string(table_name, column, search_term)
|
142
|
+
"(#{table_name}.#{column} % #{search_term})"
|
143
|
+
end
|
144
|
+
|
145
|
+
def assemble_query(similarities, conditions, exclusive)
|
146
|
+
rank = connection.quote_column_name('rank' + rand.to_s)
|
147
|
+
|
148
|
+
select("#{quoted_table_name + '.*,' if scoped.select_values.empty?} #{similarities.join(" + ")} AS #{rank}").
|
149
|
+
where(conditions.join(exclusive ? " AND " : " OR ")).
|
150
|
+
order("#{rank} DESC")
|
151
|
+
end
|
152
|
+
|
153
|
+
def normalize(query)
|
154
|
+
query
|
155
|
+
end
|
156
|
+
|
157
|
+
def searchable_columns
|
158
|
+
columns.select {|column| [:string, :text].include? column.type }.map(&:name)
|
159
|
+
end
|
160
|
+
|
161
|
+
def quoted_language
|
162
|
+
@quoted_language ||= connection.quote(searchable_language)
|
163
|
+
end
|
164
|
+
|
165
|
+
def searchable_language
|
166
|
+
Textacular.searchable_language
|
167
|
+
end
|
168
|
+
|
169
|
+
module Helper
|
170
|
+
class << self
|
171
|
+
def normalize(query)
|
172
|
+
query.to_s.gsub(' ', '\\\\ ')
|
173
|
+
end
|
174
|
+
|
175
|
+
def method_name_regex
|
176
|
+
/^(?<search_type>((basic|advanced|fuzzy)_)?search)_by_(?<columns>[_a-zA-Z]\w*)$/
|
177
|
+
end
|
178
|
+
|
179
|
+
def search_type(method)
|
180
|
+
method.to_s.match(method_name_regex)[:search_type]
|
181
|
+
end
|
182
|
+
|
183
|
+
def exclusive_dynamic_search_columns(method)
|
184
|
+
if match = method.to_s.match(method_name_regex)
|
185
|
+
match[:columns].split('_and_')
|
186
|
+
else
|
187
|
+
[]
|
188
|
+
end
|
189
|
+
end
|
190
|
+
|
191
|
+
def inclusive_dynamic_search_columns(method)
|
192
|
+
if match = method.to_s.match(method_name_regex)
|
193
|
+
match[:columns].split('_or_')
|
194
|
+
else
|
195
|
+
[]
|
196
|
+
end
|
197
|
+
end
|
198
|
+
|
199
|
+
def exclusive_dynamic_search_method?(method, class_columns)
|
200
|
+
string_columns = class_columns.map(&:name)
|
201
|
+
columns = exclusive_dynamic_search_columns(method)
|
202
|
+
unless columns.empty?
|
203
|
+
columns.all? {|column| string_columns.include?(column) }
|
204
|
+
else
|
205
|
+
false
|
206
|
+
end
|
207
|
+
end
|
208
|
+
|
209
|
+
def inclusive_dynamic_search_method?(method, class_columns)
|
210
|
+
string_columns = class_columns.map(&:name)
|
211
|
+
columns = inclusive_dynamic_search_columns(method)
|
212
|
+
unless columns.empty?
|
213
|
+
columns.all? {|column| string_columns.include?(column) }
|
214
|
+
else
|
215
|
+
false
|
216
|
+
end
|
217
|
+
end
|
218
|
+
|
219
|
+
def dynamic_search_method?(method, class_columns)
|
220
|
+
exclusive_dynamic_search_method?(method, class_columns) or
|
221
|
+
inclusive_dynamic_search_method?(method, class_columns)
|
222
|
+
end
|
223
|
+
end
|
224
|
+
end
|
225
|
+
end
|
226
|
+
|
227
|
+
require File.expand_path(File.dirname(__FILE__) + '/textacular/full_text_indexer')
|
@@ -0,0 +1,79 @@
|
|
1
|
+
class Textacular::FullTextIndexer
|
2
|
+
def generate_migration(model_name)
|
3
|
+
stream_output do |io|
|
4
|
+
io.puts(<<-MIGRATION)
|
5
|
+
class #{model_name}FullTextSearch < ActiveRecord::Migration
|
6
|
+
def self.up
|
7
|
+
execute(<<-SQL.strip)
|
8
|
+
#{up_migration(model_name)}
|
9
|
+
SQL
|
10
|
+
end
|
11
|
+
|
12
|
+
def self.down
|
13
|
+
execute(<<-SQL.strip)
|
14
|
+
#{down_migration(model_name)}
|
15
|
+
SQL
|
16
|
+
end
|
17
|
+
end
|
18
|
+
MIGRATION
|
19
|
+
end
|
20
|
+
end
|
21
|
+
|
22
|
+
def stream_output(now = Time.now.utc, &block)
|
23
|
+
if !@output_stream && defined?(Rails)
|
24
|
+
File.open(migration_file_name(now), 'w', &block)
|
25
|
+
else
|
26
|
+
@output_stream ||= $stdout
|
27
|
+
|
28
|
+
yield @output_stream
|
29
|
+
end
|
30
|
+
end
|
31
|
+
|
32
|
+
private
|
33
|
+
|
34
|
+
def migration_file_name(now = Time.now.utc)
|
35
|
+
File.join(Rails.root, 'db', 'migrate',"#{now.strftime('%Y%m%d%H%M%S')}_full_text_search.rb")
|
36
|
+
end
|
37
|
+
|
38
|
+
def up_migration(model_name)
|
39
|
+
migration_with_type(model_name, :up)
|
40
|
+
end
|
41
|
+
|
42
|
+
def down_migration(model_name)
|
43
|
+
migration_with_type(model_name, :down)
|
44
|
+
end
|
45
|
+
|
46
|
+
def migration_with_type(model_name, type)
|
47
|
+
sql_lines = ''
|
48
|
+
|
49
|
+
model = Kernel.const_get(model_name)
|
50
|
+
model.indexable_columns.each do |column|
|
51
|
+
sql_lines << drop_index_sql_for(model, column)
|
52
|
+
sql_lines << create_index_sql_for(model, column) if type == :up
|
53
|
+
end
|
54
|
+
|
55
|
+
sql_lines.strip.gsub("\n","\n ")
|
56
|
+
end
|
57
|
+
|
58
|
+
def drop_index_sql_for(model, column)
|
59
|
+
"DROP index IF EXISTS #{index_name_for(model, column)};\n"
|
60
|
+
end
|
61
|
+
|
62
|
+
def create_index_sql_for(model, column)
|
63
|
+
# The spacing gets sort of wonky in here.
|
64
|
+
|
65
|
+
<<-SQL
|
66
|
+
CREATE index #{index_name_for(model, column)}
|
67
|
+
ON #{model.table_name}
|
68
|
+
USING gin(to_tsvector("#{dictionary}", "#{model.table_name}"."#{column}"::text));
|
69
|
+
SQL
|
70
|
+
end
|
71
|
+
|
72
|
+
def index_name_for(model, column)
|
73
|
+
"#{model.table_name}_#{column}_fts_idx"
|
74
|
+
end
|
75
|
+
|
76
|
+
def dictionary
|
77
|
+
Textacular.searchable_language
|
78
|
+
end
|
79
|
+
end
|
@@ -0,0 +1,57 @@
|
|
1
|
+
module Textacular
|
2
|
+
class PostgresModuleInstaller
|
3
|
+
def install_module(module_name)
|
4
|
+
major, minor, patch = postgres_version.split('.')
|
5
|
+
|
6
|
+
if major.to_i >= 9 && minor.to_i >= 1
|
7
|
+
install_postgres_91_module(module_name)
|
8
|
+
else
|
9
|
+
install_postgres_90_module(module_name)
|
10
|
+
end
|
11
|
+
end
|
12
|
+
|
13
|
+
def db_name
|
14
|
+
@db_name ||= ActiveRecord::Base.connection.current_database
|
15
|
+
end
|
16
|
+
|
17
|
+
private
|
18
|
+
|
19
|
+
def postgres_version
|
20
|
+
@postgres_version ||= ask_pg_config('version').match(/PostgreSQL ([0-9]+(\.[0-9]+)*)/)[1]
|
21
|
+
end
|
22
|
+
|
23
|
+
def postgres_share_dir
|
24
|
+
@share_dir ||= ask_pg_config('sharedir')
|
25
|
+
end
|
26
|
+
|
27
|
+
def ask_pg_config(argument)
|
28
|
+
result = `pg_config --#{argument}`.chomp
|
29
|
+
|
30
|
+
raise RuntimeError, "Cannot find Postgres's #{argument}." unless $?.success?
|
31
|
+
|
32
|
+
result
|
33
|
+
end
|
34
|
+
|
35
|
+
def install_postgres_90_module(module_name)
|
36
|
+
module_location = "#{postgres_share_dir}/contrib/#{module_name}.sql"
|
37
|
+
|
38
|
+
unless system("ls #{module_location}")
|
39
|
+
raise RuntimeError, "Cannot find the #{module_name} module. Was it compiled and installed?"
|
40
|
+
end
|
41
|
+
|
42
|
+
unless system("psql -d #{db_name} -f #{module_location}")
|
43
|
+
raise RuntimeError, "`psql -d #{db_name} -f #{module_location}` cannot complete successfully."
|
44
|
+
end
|
45
|
+
end
|
46
|
+
|
47
|
+
def install_postgres_91_module(module_name)
|
48
|
+
module_location = "#{postgres_share_dir}/extension/#{module_name}.control"
|
49
|
+
|
50
|
+
unless system("ls #{module_location}")
|
51
|
+
raise RuntimeError, "Cannot find the #{module_name} module. Was it compiled and installed?"
|
52
|
+
end
|
53
|
+
|
54
|
+
ActiveRecord::Base.connection.execute("CREATE EXTENSION #{module_name};")
|
55
|
+
end
|
56
|
+
end
|
57
|
+
end
|
@@ -0,0 +1,10 @@
|
|
1
|
+
# Module used to conform to Rails 3 plugin API
|
2
|
+
require File.expand_path(File.dirname(__FILE__) + '/../textacular')
|
3
|
+
|
4
|
+
module Textacular
|
5
|
+
class Railtie < Rails::Railtie
|
6
|
+
initializer "textacular.configure_rails_initialization" do
|
7
|
+
ActiveRecord::Base.extend(Textacular)
|
8
|
+
end
|
9
|
+
end
|
10
|
+
end
|
@@ -0,0 +1,20 @@
|
|
1
|
+
require 'textacular'
|
2
|
+
|
3
|
+
def Searchable(*searchable_columns)
|
4
|
+
Module.new do
|
5
|
+
|
6
|
+
include Textacular
|
7
|
+
|
8
|
+
define_method(:searchable_columns) do
|
9
|
+
searchable_columns.map(&:to_s)
|
10
|
+
end
|
11
|
+
|
12
|
+
private :searchable_columns
|
13
|
+
|
14
|
+
def indexable_columns
|
15
|
+
searchable_columns.to_enum
|
16
|
+
end
|
17
|
+
end
|
18
|
+
end
|
19
|
+
|
20
|
+
Searchable = Textacular
|
@@ -0,0 +1,18 @@
|
|
1
|
+
require 'rake'
|
2
|
+
require 'textacular'
|
3
|
+
|
4
|
+
namespace :textacular do
|
5
|
+
desc 'Create full text search index migration, give the model for which you want to create the indexes'
|
6
|
+
task :create_index_migration, [:model_name] => :environment do |task, args|
|
7
|
+
raise 'A model name is required' unless args[:model_name]
|
8
|
+
Textacular::FullTextIndexer.new.generate_migration(args[:model_name])
|
9
|
+
end
|
10
|
+
|
11
|
+
desc "Install trigram text search module"
|
12
|
+
task :install_trigram => [:environment] do
|
13
|
+
installer = Textacular::PostgresModuleInstaller.new
|
14
|
+
installer.install_module('pg_trgm')
|
15
|
+
|
16
|
+
puts "Trigram text search module successfully installed into '#{installer.db_name}' database."
|
17
|
+
end
|
18
|
+
end
|
File without changes
|