assimilate 0.3.1 → 0.3.2

data/.travis.yml

@@ -0,0 +1,5 @@
+language: ruby
+rvm:
+  - 1.9.3
+  - jruby-19mode # JRuby in 1.9 mode
+  - rbx-19mode

data/assimilate.gemspec

@@ -16,6 +16,8 @@ Gem::Specification.new do |gem|
   gem.require_paths = ["lib"]
   gem.version       = Assimilate::VERSION
 
+  gem.required_ruby_version = '>= 1.9.2'
+
   gem.add_dependency "mongo", "~> 1.6.0"
   gem.add_dependency "bson_ext", "~> 1.6.0"
   gem.add_dependency 'activesupport', "~> 3.2.0"

data/lib/assimilate/extender.rb

@@ -26,7 +26,15 @@ class Assimilate::Extender
       if key
         # ignore records that are missing a key value.
         if h.include?(key)
-          raise Assimilate::CorruptDataError, "Duplicate records for key [#{key}] in #{@domainkey} [#{@domain}]"
+          # conflict with existing record for the same key
+          # HACK - HARD-CODED BEHAVIOR:
+          # * look for a boolean field called '_active'; if the previously-stored value says inactive and
+          # this one is active, then use the new one.
+          if h[key].include?('_active') && rec.include?('_active') && h[key]['_active'] != rec['_active']
+            rec = [h[key],rec].select {|r| r['_active']}.first
+          else
+            raise Assimilate::CorruptDataError, "Duplicate records for key [#{key}] in #{@domainkey} [#{@domain}]"
+          end
         end
         h[key] = rec
       end

data/lib/assimilate/version.rb

@@ -1,3 +1,3 @@
 module Assimilate
-  VERSION = "0.3.1"
+  VERSION = "0.3.2"
 end

data/spec/data/master_records_conflicting.csv

@@ -0,0 +1,8 @@
+ID,name,title,spouse,_active
+1,George Washington,President,Martha,1
+2,John Adams,Vice President,Abigail,1
+3,Benjamin Franklin,Sage,Deborah,1
+4,Aaron Burr,Duelist,Theodosia,1
+5,Alexander Hamilton,Financier,Elizabeth,1
+6,James Madison,Theorist,Dolly,1
+5,Alexander Hamilton,Financier,PretendSecondWife,0

data/spec/lib/extend_spec.rb

@@ -210,6 +210,30 @@ describe "loading extended data" do
     end
   end
 
+  describe "with conflicting source records" do
+    before(:all) do
+      reset_catalog
+      import_base_data("123", "master_records_conflicting.csv")
+    end
+
+    before(:each) do
+      import_extended_data("1001", "dates.csv")
+    end
+
+    it "should capture changes" do
+      @extender.stats.should == {
+        :baseline_count => 6,
+        :final_count => 7,
+        :distinct_ids => 4,
+        :adds_count => 1,
+        :new_ids => ['16'],
+        :updates_count => 3,
+        :updated_fields => {'date' => 4},
+        :unchanged_count => 0
+      }
+    end
+  end
+
   # test handling of multiple records for same ID in the extended-data file
   # test importing data at top level (no keyfield for sub-attributes)
 end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: assimilate
 version: !ruby/object:Gem::Version
-  version: 0.3.1
+  version: 0.3.2
   prerelease: 
 platform: ruby
 authors:
@@ -9,11 +9,11 @@ authors:
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2012-06-06 00:00:00.000000000 Z
+date: 2012-06-27 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: mongo
-  requirement: &2156709640 !ruby/object:Gem::Requirement
+  requirement: &2152542320 !ruby/object:Gem::Requirement
     none: false
     requirements:
     - - ~>
@@ -21,10 +21,10 @@ dependencies:
         version: 1.6.0
   type: :runtime
   prerelease: false
-  version_requirements: *2156709640
+  version_requirements: *2152542320
 - !ruby/object:Gem::Dependency
   name: bson_ext
-  requirement: &2156709100 !ruby/object:Gem::Requirement
+  requirement: &2152541840 !ruby/object:Gem::Requirement
     none: false
     requirements:
     - - ~>
@@ -32,10 +32,10 @@ dependencies:
         version: 1.6.0
   type: :runtime
   prerelease: false
-  version_requirements: *2156709100
+  version_requirements: *2152541840
 - !ruby/object:Gem::Dependency
   name: activesupport
-  requirement: &2154106340 !ruby/object:Gem::Requirement
+  requirement: &2152541360 !ruby/object:Gem::Requirement
     none: false
     requirements:
     - - ~>
@@ -43,10 +43,10 @@ dependencies:
         version: 3.2.0
   type: :runtime
   prerelease: false
-  version_requirements: *2154106340
+  version_requirements: *2152541360
 - !ruby/object:Gem::Dependency
   name: rake
-  requirement: &2154105840 !ruby/object:Gem::Requirement
+  requirement: &2152540900 !ruby/object:Gem::Requirement
     none: false
     requirements:
     - - ~>
@@ -54,10 +54,10 @@ dependencies:
         version: 0.9.2
   type: :development
   prerelease: false
-  version_requirements: *2154105840
+  version_requirements: *2152540900
 - !ruby/object:Gem::Dependency
   name: rspec
-  requirement: &2154105340 !ruby/object:Gem::Requirement
+  requirement: &2152540420 !ruby/object:Gem::Requirement
     none: false
     requirements:
     - - ~>
@@ -65,10 +65,10 @@ dependencies:
         version: 2.9.0
   type: :development
   prerelease: false
-  version_requirements: *2154105340
+  version_requirements: *2152540420
 - !ruby/object:Gem::Dependency
   name: guard-rspec
-  requirement: &2154104840 !ruby/object:Gem::Requirement
+  requirement: &2152539940 !ruby/object:Gem::Requirement
     none: false
     requirements:
     - - ~>
@@ -76,10 +76,10 @@ dependencies:
         version: 0.7.0
   type: :development
   prerelease: false
-  version_requirements: *2154104840
+  version_requirements: *2152539940
 - !ruby/object:Gem::Dependency
   name: ruby_gntp
-  requirement: &2154104360 !ruby/object:Gem::Requirement
+  requirement: &2152539480 !ruby/object:Gem::Requirement
     none: false
     requirements:
     - - ~>
@@ -87,7 +87,7 @@ dependencies:
         version: 0.3.4
   type: :development
   prerelease: false
-  version_requirements: *2154104360
+  version_requirements: *2152539480
 description: Ingest updates from CSV and apply to set of hashes
 email:
 - jmay@pobox.com
@@ -97,6 +97,7 @@ extensions: []
 extra_rdoc_files: []
 files:
 - .gitignore
+- .travis.yml
 - Gemfile
 - Guardfile
 - LICENSE
@@ -116,6 +117,7 @@ files:
 - spec/data/duplicate_input.csv
 - spec/data/logs1.csv
 - spec/data/logs2.csv
+- spec/data/master_records_conflicting.csv
 - spec/data/test.yml
 - spec/data/updates.csv
 - spec/lib/batch_spec.rb
@@ -132,7 +134,7 @@ required_ruby_version: !ruby/object:Gem::Requirement
   requirements:
   - - ! '>='
     - !ruby/object:Gem::Version
-      version: '0'
+      version: 1.9.2
 required_rubygems_version: !ruby/object:Gem::Requirement
   none: false
   requirements:
@@ -152,6 +154,7 @@ test_files:
 - spec/data/duplicate_input.csv
 - spec/data/logs1.csv
 - spec/data/logs2.csv
+- spec/data/master_records_conflicting.csv
 - spec/data/test.yml
 - spec/data/updates.csv
 - spec/lib/batch_spec.rb