cubicle 0.1.20 → 0.1.21
Sign up to get free protection for your applications and to get access to all the features.
- data/CHANGELOG.rdoc +5 -0
- data/cubicle.gemspec +8 -2
- data/lib/cubicle.rb +2 -0
- data/lib/cubicle/aggregation/aggregation_manager.rb +38 -70
- data/lib/cubicle/aggregation/aggregation_metadata.rb +121 -0
- data/lib/cubicle/aggregation/cubicle_metadata.rb +30 -0
- data/lib/cubicle/aggregation/dsl.rb +1 -1
- data/lib/cubicle/version.rb +1 -1
- data/test/cubicle/aggregation/aggregation_metadata_test.rb +89 -0
- data/test/cubicle/aggregation/cubicle_metadata_test.rb +9 -0
- data/test/cubicle/cubicle_aggregation_test.rb +13 -11
- data/test/cubicle/cubicle_query_test.rb +1 -1
- data/test/cubicle/mongo_mapper/aggregate_plugin_test.rb +2 -6
- data/test/log/test.log +84495 -1228
- metadata +9 -3
data/CHANGELOG.rdoc
CHANGED
@@ -1,3 +1,8 @@
|
|
1
|
+
==0.1.21
|
2
|
+
*Added metadata tables in the database for cubicle to manage aggregation info. This was necessary because previously
|
3
|
+
I was trying to overload the collection name with metadata, which was making the names longer than MongoDb could support
|
4
|
+
and causing errors. This change will enable richer monitoring and profiling and optimization in the near future.
|
5
|
+
|
1
6
|
==0.1.20
|
2
7
|
*Updated to work with mongo driver 1.0 (and therefore latest versions of MongoMapper)
|
3
8
|
|
data/cubicle.gemspec
CHANGED
@@ -5,11 +5,11 @@
|
|
5
5
|
|
6
6
|
Gem::Specification.new do |s|
|
7
7
|
s.name = %q{cubicle}
|
8
|
-
s.version = "0.1.
|
8
|
+
s.version = "0.1.21"
|
9
9
|
|
10
10
|
s.required_rubygems_version = Gem::Requirement.new(">= 0") if s.respond_to? :required_rubygems_version=
|
11
11
|
s.authors = ["Nathan Stults"]
|
12
|
-
s.date = %q{2010-05-
|
12
|
+
s.date = %q{2010-05-05}
|
13
13
|
s.description = %q{Cubicle provides a dsl and aggregation caching framework for automating the generation, execution and caching of map reduce queries when using MongoDB in Ruby. Cubicle also includes a MongoMapper plugin for quickly performing ad-hoc, multi-level group-by queries against a MongoMapper model.}
|
14
14
|
s.email = %q{hereiam@sonic.net}
|
15
15
|
s.extra_rdoc_files = [
|
@@ -27,7 +27,9 @@ Gem::Specification.new do |s|
|
|
27
27
|
"lib/cubicle/aggregation.rb",
|
28
28
|
"lib/cubicle/aggregation/ad_hoc.rb",
|
29
29
|
"lib/cubicle/aggregation/aggregation_manager.rb",
|
30
|
+
"lib/cubicle/aggregation/aggregation_metadata.rb",
|
30
31
|
"lib/cubicle/aggregation/aggregation_view.rb",
|
32
|
+
"lib/cubicle/aggregation/cubicle_metadata.rb",
|
31
33
|
"lib/cubicle/aggregation/dsl.rb",
|
32
34
|
"lib/cubicle/aggregation/map_reduce_helper.rb",
|
33
35
|
"lib/cubicle/bucketized_dimension.rb",
|
@@ -55,6 +57,8 @@ Gem::Specification.new do |s|
|
|
55
57
|
"lib/cubicle/version.rb",
|
56
58
|
"test/config/database.yml",
|
57
59
|
"test/cubicle/aggregation/ad_hoc_test.rb",
|
60
|
+
"test/cubicle/aggregation/aggregation_metadata_test.rb",
|
61
|
+
"test/cubicle/aggregation/cubicle_metadata_test.rb",
|
58
62
|
"test/cubicle/bucketized_dimension_test.rb",
|
59
63
|
"test/cubicle/cubicle_aggregation_test.rb",
|
60
64
|
"test/cubicle/cubicle_query_test.rb",
|
@@ -77,6 +81,8 @@ Gem::Specification.new do |s|
|
|
77
81
|
s.summary = %q{Pseudo-Multi Dimensional analysis / simplified aggregation for MongoDB in Ruby (NOLAP ;))}
|
78
82
|
s.test_files = [
|
79
83
|
"test/cubicle/aggregation/ad_hoc_test.rb",
|
84
|
+
"test/cubicle/aggregation/aggregation_metadata_test.rb",
|
85
|
+
"test/cubicle/aggregation/cubicle_metadata_test.rb",
|
80
86
|
"test/cubicle/bucketized_dimension_test.rb",
|
81
87
|
"test/cubicle/cubicle_aggregation_test.rb",
|
82
88
|
"test/cubicle/cubicle_query_test.rb",
|
data/lib/cubicle.rb
CHANGED
@@ -25,6 +25,8 @@ dir = File.dirname(__FILE__)
|
|
25
25
|
"data/level",
|
26
26
|
"data/hierarchy",
|
27
27
|
"data/table",
|
28
|
+
"aggregation/aggregation_metadata",
|
29
|
+
"aggregation/cubicle_metadata",
|
28
30
|
"aggregation/aggregation_view",
|
29
31
|
"aggregation/aggregation_manager",
|
30
32
|
"aggregation/map_reduce_helper",
|
@@ -2,10 +2,11 @@ module Cubicle
|
|
2
2
|
module Aggregation
|
3
3
|
class AggregationManager
|
4
4
|
|
5
|
-
attr_reader :aggregation
|
5
|
+
attr_reader :aggregation, :metadata
|
6
6
|
|
7
7
|
def initialize(aggregation)
|
8
8
|
@aggregation = aggregation
|
9
|
+
@metadata = Cubicle::Aggregation::CubicleMetadata.new(aggregation)
|
9
10
|
end
|
10
11
|
|
11
12
|
def database
|
@@ -32,26 +33,28 @@ module Cubicle
|
|
32
33
|
|
33
34
|
find_options[:sort] = prepare_order_by(query)
|
34
35
|
filter = {}
|
36
|
+
|
35
37
|
if query == aggregation || query.transient?
|
36
|
-
|
38
|
+
reduction = aggregate(query,options)
|
37
39
|
else
|
38
40
|
process_if_required
|
39
|
-
|
41
|
+
agg_data = aggregation_for(query)
|
42
|
+
reduction = agg_data.collection
|
40
43
|
#if the query exactly matches the aggregation in terms of requested members, we can issue a simple find
|
41
44
|
#otherwise, a second map reduce is required to reduce the data set one last time
|
42
|
-
if query.all_dimensions? || (
|
45
|
+
if query.all_dimensions? || (agg_data.member_names - query.member_names - [:all_measures]).blank?
|
43
46
|
filter = prepare_filter(query,options[:where] || {})
|
44
47
|
else
|
45
|
-
|
48
|
+
reduction = aggregate(query,:source_collection=>agg_data.target_collection_name)
|
46
49
|
end
|
47
50
|
end
|
48
51
|
|
49
|
-
if
|
50
|
-
Cubicle::Data::Table.new(query,[],0)
|
52
|
+
if reduction.blank?
|
53
|
+
Cubicle::Data::Table.new(query,[],0)
|
51
54
|
else
|
52
|
-
count =
|
53
|
-
results =
|
54
|
-
|
55
|
+
count = reduction.count
|
56
|
+
results = reduction.find(filter,find_options).to_a
|
57
|
+
reduction.drop if reduction.name =~ /^tmp.mr.*/
|
55
58
|
Cubicle::Data::Table.new(query, results, count)
|
56
59
|
end
|
57
60
|
|
@@ -77,44 +80,42 @@ module Cubicle
|
|
77
80
|
|
78
81
|
def expire!
|
79
82
|
collection.drop
|
80
|
-
|
83
|
+
@metadata.expire!
|
81
84
|
end
|
82
85
|
|
83
|
-
|
86
|
+
def aggregate(query,options={})
|
87
|
+
view = AggregationView.new(aggregation,query)
|
84
88
|
|
85
|
-
|
86
|
-
database.collection_names.select {|col_name|col_name=~/#{aggregation.target_collection_name}_aggregation_(.*)/}
|
87
|
-
end
|
89
|
+
map, reduce = MapReduceHelper.generate_map_function(query), MapReduceHelper.generate_reduce_function
|
88
90
|
|
89
|
-
|
90
|
-
|
91
|
-
|
91
|
+
options[:finalize] = MapReduceHelper.generate_finalize_function(query)
|
92
|
+
options["query"] = expand_template(prepare_filter(query,options[:where] || {}),view)
|
93
|
+
|
94
|
+
query.source_collection_name = options.delete(:source_collection) || query.source_collection_name || aggregation.source_collection_name
|
92
95
|
|
93
|
-
|
94
|
-
|
95
|
-
|
96
|
-
|
97
|
-
|
96
|
+
target_collection = options.delete(:target_collection)
|
97
|
+
target_collection ||= query.target_collection_name if query.respond_to?(:target_collection_name)
|
98
|
+
|
99
|
+
options[:out] = target_collection unless target_collection.blank? || query.transient?
|
100
|
+
|
101
|
+
#This is defensive - some tests run without ever initializing any collections
|
102
|
+
unless database.collection_names.include?(query.source_collection_name)
|
103
|
+
Cubicle.logger.info "No collection was found in the database with a name of #{query.source_collection_name}"
|
104
|
+
return []
|
98
105
|
end
|
99
106
|
|
100
|
-
|
101
|
-
#we are sorting by length because the aggregation with the least number of members
|
102
|
-
#is likely to be the most efficient data source as it will likely contain the smallest number of rows.
|
103
|
-
#this will not always be true, and situations may exist where it is rarely true, however the alternative
|
104
|
-
#is to actually count rows of candidates, which seems a bit wasteful. Of course only the profiler knows,
|
105
|
-
#but until there is some reason to believe the aggregation caching process needs be highly performant,
|
106
|
-
#this should do for now.
|
107
|
-
candidates = existing.select {|candidate|(dimension_names - candidate).blank?}.sort {|a,b|a.length <=> b.length}
|
107
|
+
result = database[query.source_collection_name].map_reduce(expand_template(map, view),reduce,options)
|
108
108
|
|
109
|
-
|
110
|
-
#we'll just use the base cubes aggregation collection
|
111
|
-
return target_collection_name if candidates.blank?
|
112
|
-
"#{target_collection_name}_aggregation_#{candidates[0].join('.')}"
|
109
|
+
ensure_indexes(target_collection,query.dimension_names) if target_collection
|
113
110
|
|
111
|
+
result
|
114
112
|
end
|
115
113
|
|
114
|
+
protected
|
115
|
+
|
116
|
+
|
116
117
|
def aggregation_for(query)
|
117
|
-
return collection if query.all_dimensions?
|
118
|
+
#return collection if query.all_dimensions?
|
118
119
|
|
119
120
|
aggregation_query = query.clone
|
120
121
|
#If the query needs to filter on a field, it had better be in the aggregation...if it isn't a $where filter...
|
@@ -122,15 +123,7 @@ module Cubicle
|
|
122
123
|
filter.keys.each {|filter_key|aggregation_query.select(filter_key) unless filter_key=~/\$where/} unless filter.blank?
|
123
124
|
|
124
125
|
dimension_names = aggregation_query.dimension_names.sort
|
125
|
-
|
126
|
-
|
127
|
-
unless database.collection_names.include?(agg_col_name)
|
128
|
-
source_col_name = find_best_source_collection(dimension_names)
|
129
|
-
exec_query = aggregation.query(dimension_names + [:all_measures], :source_collection=>source_col_name, :defer=>true)
|
130
|
-
aggregate(exec_query, :target_collection=>agg_col_name)
|
131
|
-
end
|
132
|
-
|
133
|
-
database[agg_col_name]
|
126
|
+
@metadata.aggregation_for(dimension_names)
|
134
127
|
end
|
135
128
|
|
136
129
|
def ensure_indexes(collection_name,dimension_names)
|
@@ -146,31 +139,6 @@ module Cubicle
|
|
146
139
|
#col.create_index(dimension_names.map{|dim|[dim,1]})
|
147
140
|
end
|
148
141
|
|
149
|
-
def aggregate(query,options={})
|
150
|
-
view = AggregationView.new(aggregation,query)
|
151
|
-
|
152
|
-
map, reduce = MapReduceHelper.generate_map_function(query), MapReduceHelper.generate_reduce_function
|
153
|
-
|
154
|
-
options[:finalize] = MapReduceHelper.generate_finalize_function(query)
|
155
|
-
options["query"] = expand_template(prepare_filter(query,options[:where] || {}),view)
|
156
|
-
|
157
|
-
query.source_collection_name = options.delete(:source_collection) || query.source_collection_name || aggregation.source_collection_name
|
158
|
-
|
159
|
-
target_collection = options.delete(:target_collection)
|
160
|
-
target_collection ||= query.target_collection_name if query.respond_to?(:target_collection_name)
|
161
|
-
|
162
|
-
options[:out] = target_collection unless target_collection.blank? || query.transient?
|
163
|
-
|
164
|
-
#This is defensive - some tests run without ever initializing any collections
|
165
|
-
return [] unless database.collection_names.include?(query.source_collection_name)
|
166
|
-
|
167
|
-
result = database[query.source_collection_name].map_reduce(expand_template(map, view),reduce,options)
|
168
|
-
|
169
|
-
ensure_indexes(target_collection,query.dimension_names) if target_collection
|
170
|
-
|
171
|
-
result
|
172
|
-
end
|
173
|
-
|
174
142
|
def expand_template(template,view)
|
175
143
|
return "" unless template
|
176
144
|
return Mustache.render(template,view) if template.is_a?(String)
|
@@ -0,0 +1,121 @@
|
|
1
|
+
module Cubicle
|
2
|
+
module Aggregation
|
3
|
+
class AggregationMetadata
|
4
|
+
class << self
|
5
|
+
|
6
|
+
def collection
|
7
|
+
@@aggregations_collection_name ||= "#{Cubicle::Aggregation::CubicleMetadata.collection.name}.aggregations"
|
8
|
+
Cubicle.mongo.database[@@aggregations_collection_name]
|
9
|
+
end
|
10
|
+
|
11
|
+
def collection=(collection_name)
|
12
|
+
@@aggregations_collection_name = collection_name
|
13
|
+
end
|
14
|
+
|
15
|
+
def min_records_to_reduce
|
16
|
+
@min_records_to_reduce ||= 100
|
17
|
+
end
|
18
|
+
|
19
|
+
def min_records_to_reduce=(min)
|
20
|
+
@min_records_to_reduce = min
|
21
|
+
end
|
22
|
+
|
23
|
+
def expire(aggregation)
|
24
|
+
aggregation_name = case aggregation
|
25
|
+
when String then aggregation
|
26
|
+
when Symbol then aggregation.to_s
|
27
|
+
when Cubicle::Aggregation::CubicleMetadata then aggregation.aggregation.name
|
28
|
+
else aggregation.name
|
29
|
+
end
|
30
|
+
Cubicle.mongo.database.collection_names.each do |col|
|
31
|
+
Cubicle.mongo.database[col].drop if col =~ /cubicle.aggregation.#{aggregation_name}._*/i
|
32
|
+
collection.remove(:aggregation=>aggregation_name)
|
33
|
+
end
|
34
|
+
end
|
35
|
+
end
|
36
|
+
|
37
|
+
def initialize(cubicle_metadata,member_names_or_attribute_hash)
|
38
|
+
@cubicle_metadata = cubicle_metadata
|
39
|
+
if (member_names_or_attribute_hash.kind_of?(Hash))
|
40
|
+
@attributes = member_names_or_attribute_hash
|
41
|
+
else
|
42
|
+
member_names = member_names_or_attribute_hash
|
43
|
+
@candidate_aggregation = self.class.collection.find(
|
44
|
+
:aggregation=>@cubicle_metadata.aggregation.name,
|
45
|
+
:member_names=>{"$all"=>member_names}, :document_count=>{"$gte"=>0}).sort([:document_count, :asc]).limit(1).next_document
|
46
|
+
|
47
|
+
|
48
|
+
#since the operator used in the query was $all, having equal lengths in the original and returned
|
49
|
+
#member array means that they are identical, which means that regardless of the number of documents
|
50
|
+
#in the aggregation, it is the candidate we want. Otherwise, we'll check to see if we
|
51
|
+
#boil down the data further, or just make our soup with what we've got.
|
52
|
+
@attributes = @candidate_aggregation if @candidate_aggregation &&
|
53
|
+
(@candidate_aggregation["member_names"].length == member_names.length ||
|
54
|
+
@candidate_aggregation["document_count"] < self.class.min_records_to_reduce)
|
55
|
+
|
56
|
+
unless @attributes
|
57
|
+
@attributes = HashWithIndifferentAccess.new({:aggregation=>@cubicle_metadata.aggregation.name,
|
58
|
+
:member_names=>member_names,
|
59
|
+
:document_count=>-1})
|
60
|
+
|
61
|
+
#materialize the aggregation, and, if the operation was successful,
|
62
|
+
#register it as available for use by future queries
|
63
|
+
@attributes[:_id] = self.class.collection.insert(@attributes)
|
64
|
+
materialize!
|
65
|
+
end
|
66
|
+
|
67
|
+
end
|
68
|
+
end
|
69
|
+
|
70
|
+
def target_collection_name
|
71
|
+
"cubicle.aggregation.#{@cubicle_metadata.aggregation.name}._#{@attributes["_id"].to_s}"
|
72
|
+
end
|
73
|
+
|
74
|
+
def source_collection_name
|
75
|
+
if @candidate_aggregation
|
76
|
+
candidate = Cubicle::Aggregation::AggregationMetadata.new(@cubicle_metadata,@candidate_aggregation)
|
77
|
+
return candidate.target_collection_name
|
78
|
+
end
|
79
|
+
@cubicle_metadata.aggregation.target_collection_name
|
80
|
+
end
|
81
|
+
|
82
|
+
def member_names; @attributes["member_names"] || []; end
|
83
|
+
|
84
|
+
def materialized?
|
85
|
+
document_count >= 0 &&
|
86
|
+
(!@collection.blank? ||
|
87
|
+
Cubicle.mongo.database.collection_names.include?(target_collection_name))
|
88
|
+
end
|
89
|
+
|
90
|
+
def collection
|
91
|
+
@collection ||= Cubicle.mongo.database[target_collection_name] if materialized?
|
92
|
+
end
|
93
|
+
|
94
|
+
def collection=(collection)
|
95
|
+
@collection = collection
|
96
|
+
end
|
97
|
+
|
98
|
+
def document_count
|
99
|
+
@attributes["document_count"]
|
100
|
+
end
|
101
|
+
|
102
|
+
protected
|
103
|
+
def update_document_count!(new_doc_count)
|
104
|
+
self.class.collection.update({:_id=>@attributes[:_id]}, "$set"=>{:document_count=>new_doc_count})
|
105
|
+
@attributes["document_count"]=new_doc_count
|
106
|
+
end
|
107
|
+
|
108
|
+
def materialize!
|
109
|
+
unless materialized?
|
110
|
+
exec_query = @cubicle_metadata.aggregation.query(member_names + [:all_measures],
|
111
|
+
:source_collection=>source_collection_name,
|
112
|
+
:defer=>true)
|
113
|
+
self.collection = @cubicle_metadata.aggregation.aggregator.aggregate(exec_query,
|
114
|
+
:target_collection=>target_collection_name)
|
115
|
+
end
|
116
|
+
update_document_count!(@collection.count) unless @collection.blank?
|
117
|
+
end
|
118
|
+
|
119
|
+
end
|
120
|
+
end
|
121
|
+
end
|
@@ -0,0 +1,30 @@
|
|
1
|
+
module Cubicle
|
2
|
+
module Aggregation
|
3
|
+
class CubicleMetadata
|
4
|
+
|
5
|
+
class << self
|
6
|
+
|
7
|
+
def collection
|
8
|
+
@@collection_name ||= "cubicle.metadata"
|
9
|
+
Cubicle.mongo.database[@@collection_name]
|
10
|
+
end
|
11
|
+
def collection=(collection_name)
|
12
|
+
@@collection_name = collection_name
|
13
|
+
end
|
14
|
+
end
|
15
|
+
|
16
|
+
attr_reader :aggregation
|
17
|
+
def initialize(aggregation)
|
18
|
+
@aggregation = aggregation
|
19
|
+
end
|
20
|
+
|
21
|
+
def aggregation_for(member_names = [])
|
22
|
+
AggregationMetadata.new(self,member_names)
|
23
|
+
end
|
24
|
+
|
25
|
+
def expire!
|
26
|
+
AggregationMetadata.expire(self)
|
27
|
+
end
|
28
|
+
end
|
29
|
+
end
|
30
|
+
end
|
@@ -11,7 +11,7 @@ module Cubicle
|
|
11
11
|
def target_collection_name(collection_name = nil)
|
12
12
|
return nil if transient?
|
13
13
|
return @target_name = collection_name if collection_name
|
14
|
-
@target_name ||= "
|
14
|
+
@target_name ||= "cubicle.fact.#{name.blank? ? source_collection_name : name.underscore}"
|
15
15
|
end
|
16
16
|
alias target_collection_name= target_collection_name
|
17
17
|
|
data/lib/cubicle/version.rb
CHANGED
@@ -0,0 +1,89 @@
|
|
1
|
+
require "test_helper"
|
2
|
+
|
3
|
+
class AggregationMetadataTest < ActiveSupport::TestCase
|
4
|
+
context "Class level collection names" do
|
5
|
+
should "use appropriate default values for the aggregations collection" do
|
6
|
+
assert_equal "cubicle.metadata.aggregations", Cubicle::Aggregation::AggregationMetadata.collection.name
|
7
|
+
end
|
8
|
+
end
|
9
|
+
|
10
|
+
context "AggregationMetadata.update_document_count" do
|
11
|
+
setup do
|
12
|
+
@cm = Cubicle::Aggregation::CubicleMetadata.new(DefectCubicle)
|
13
|
+
end
|
14
|
+
should "update the document count for a given aggregation instance" do
|
15
|
+
agg_info = Cubicle::Aggregation::AggregationMetadata.new(@cm,[:product])
|
16
|
+
agg_info.send(:update_document_count!,1024)
|
17
|
+
assert_equal 1024, agg_info.document_count
|
18
|
+
assert_equal false,agg_info.materialized?
|
19
|
+
end
|
20
|
+
end
|
21
|
+
|
22
|
+
context "AggregationMetadata#new" do
|
23
|
+
setup do
|
24
|
+
@cm = Cubicle::Aggregation::CubicleMetadata.new(DefectCubicle)
|
25
|
+
end
|
26
|
+
should "create initialize an instance of AggregationMetadata in the database" do
|
27
|
+
agg_info = Cubicle::Aggregation::AggregationMetadata.new(@cm,[:product,:region])
|
28
|
+
assert /cubicle.aggregation.DefectCubicle._+/ =~ agg_info.target_collection_name
|
29
|
+
assert_equal [:product,:region], agg_info.member_names
|
30
|
+
assert_equal false, agg_info.materialized?
|
31
|
+
assert_nil agg_info.collection
|
32
|
+
end
|
33
|
+
should "fetch an existing aggregation from the database" do
|
34
|
+
ag = Cubicle::Aggregation::AggregationMetadata.new(@cm,[:product,:region])
|
35
|
+
ag.send(:update_document_count!,1)
|
36
|
+
col_name = ag.target_collection_name
|
37
|
+
assert_equal col_name, Cubicle::Aggregation::AggregationMetadata.new(@cm,[:product,:region]).target_collection_name
|
38
|
+
end
|
39
|
+
should "ignore an existing aggregation that does not satisfy all fields" do
|
40
|
+
ag = Cubicle::Aggregation::AggregationMetadata.new(@cm,[:product])
|
41
|
+
ag.send(:update_document_count!,1)
|
42
|
+
col_name = ag.target_collection_name
|
43
|
+
assert col_name != Cubicle::Aggregation::AggregationMetadata.new(@cm,[:product,:region]).target_collection_name
|
44
|
+
end
|
45
|
+
should "select an existing aggregation with rows below the minimum threshold instead of creating a new one" do
|
46
|
+
agg_info = Cubicle::Aggregation::AggregationMetadata.new(@cm,[:product,:region,:operator])
|
47
|
+
agg_info.send(:update_document_count!,99)
|
48
|
+
assert_equal agg_info.target_collection_name, Cubicle::Aggregation::AggregationMetadata.new(@cm,[:product]).target_collection_name
|
49
|
+
end
|
50
|
+
|
51
|
+
should "ignore an existing aggregation with too many rows, but store that aggregation as a candidate source for use when materializing the aggregation" do
|
52
|
+
agg_info = Cubicle::Aggregation::AggregationMetadata.new(@cm,[:product,:region,:operator])
|
53
|
+
agg_info.send(:update_document_count!,101)
|
54
|
+
new_agg_info = Cubicle::Aggregation::AggregationMetadata.new(@cm,[:product])
|
55
|
+
assert agg_info.target_collection_name != new_agg_info.target_collection_name
|
56
|
+
assert_equal agg_info.target_collection_name, new_agg_info.source_collection_name
|
57
|
+
end
|
58
|
+
end
|
59
|
+
|
60
|
+
context "AggregationMetadata#materialize!" do
|
61
|
+
should "run a map reduce and produce the resulting collection" do
|
62
|
+
Defect.create_test_data
|
63
|
+
DefectCubicle.process
|
64
|
+
@cm = Cubicle::Aggregation::CubicleMetadata.new(DefectCubicle)
|
65
|
+
agg_info = Cubicle::Aggregation::AggregationMetadata.new(@cm,[:product])
|
66
|
+
aggregation = agg_info.collection
|
67
|
+
assert_not_nil aggregation
|
68
|
+
assert aggregation.count > 0
|
69
|
+
assert_equal aggregation.count, agg_info.document_count
|
70
|
+
end
|
71
|
+
end
|
72
|
+
|
73
|
+
context "AggregationMetadata.expire" do
|
74
|
+
should "drop any aggregation columns and remove metadata rows from the database" do
|
75
|
+
Defect.create_test_data
|
76
|
+
DefectCubicle.process
|
77
|
+
@cm = Cubicle::Aggregation::CubicleMetadata.new(DefectCubicle)
|
78
|
+
agg_info = Cubicle::Aggregation::AggregationMetadata.new(@cm,[:product])
|
79
|
+
|
80
|
+
assert Cubicle.mongo.database.collection_names.include?(agg_info.target_collection_name)
|
81
|
+
assert Cubicle::Aggregation::AggregationMetadata.collection.find(:aggregation=>"DefectCubicle").count > 0
|
82
|
+
|
83
|
+
Cubicle::Aggregation::AggregationMetadata.expire(@cm)
|
84
|
+
|
85
|
+
assert !Cubicle.mongo.database.collection_names.include?(agg_info.target_collection_name)
|
86
|
+
assert_equal 0, Cubicle::Aggregation::AggregationMetadata.collection.find(:aggregation=>"DefectCubicle").count
|
87
|
+
end
|
88
|
+
end
|
89
|
+
end
|