assimilate 0.0.3 → 0.0.4

data/lib/assimilate/batch.rb

@@ -80,6 +80,7 @@ class Assimilate::Batch
       :deletes_count => @deleted_keys.count,
       :deleted_ids => @deleted_keys,
       :updates_count => @changes.count,
+      :updated_ids => @changes.map {|rec| rec[idfield]},
       :unchanged_count => @noops.count,
       :updated_fields => @updated_field_counts
     }

data/lib/assimilate/command.rb

@@ -82,7 +82,7 @@ EOT
         Unchanged records: #{results[:unchanged_count]}
               New records: #{results[:adds_count]} (#{results[:new_ids].take(10).join(',')})
                   Deletes: #{results[:deletes_count]} (#{results[:deleted_ids].take(10).join(',')})
-                  Updates: #{results[:updates_count]}
+                  Updates: #{results[:updates_count]} (#{results[:updated_ids].take(10).join(',')})
 EOT
       if results[:updated_fields].any?
         $stderr.puts <<-EOT

data/lib/assimilate/version.rb

@@ -1,3 +1,3 @@
 module Assimilate
-  VERSION = "0.0.3"
+  VERSION = "0.0.4"
 end

data/lib/assimilate.rb

@@ -16,8 +16,7 @@ module Assimilate
       catalog = Catalog.new(:config => opts[:config])
       batcher = catalog.start_batch(opts.merge(:filename => filename))
 
-      records = CSV.read(filename, :headers => true)
-      records.each do |rec|
+      slurp(filename) do |rec|
         batcher << rec
       end
       if opts[:commit]
@@ -37,8 +36,7 @@ module Assimilate
     begin
       catalog = Catalog.new(:config => opts[:config])
       extender = catalog.extend_data(opts)
-      records = CSV.read(filename, :headers => true)
-      records.each do |rec|
+      slurp(filename) do |rec|
         extender << rec
       end
       if opts[:commit]
@@ -49,4 +47,20 @@ module Assimilate
       extender.stats
     end
   end
+
+  def self.slurp(filename)
+    headers = nil
+    CSV.read(filename).each do |row|
+      if !headers
+        headers = row.to_a
+      else
+        raise "Row count mismatch: #{row} vs #{headers}" if row.count > headers.count
+        hash = {}
+        row.zip(headers) do |v,k|
+          hash[k] = v.strip unless v.blank?
+        end
+        yield hash
+      end
+    end
+  end
 end

data/spec/lib/batch_spec.rb

@@ -16,8 +16,7 @@ describe "importing file" do
   def import_data(datestamp, filename = "batch_input.csv")
     @batcher = @catalog.start_batch(domain: 'testdata', datestamp: datestamp, filename: filename, idfield: 'ID')
 
-    @records = CSV.read(File.dirname(__FILE__) + "/../data/#{filename}", :headers => true)
-    @records.each do |rec|
+    Assimilate.slurp(File.dirname(__FILE__) + "/../data/#{filename}") do |rec|
       @batcher << rec
     end
     @batcher.commit
@@ -38,15 +37,15 @@ describe "importing file" do
         :deletes_count => 0,
         :deleted_ids => [],
         :updates_count => 0,
+        :updated_ids => [],
         :unchanged_count => 0,
         :updated_fields => {}
       }
     end
 
     it "should load the records verbatim" do
-      @catalog.catalog.count.should == @records.count
-      example = @records[rand(@records.count)]
-      @catalog.where('_resource' => 'testdata', 'ID' => example['ID']).should == example.to_hash
+      @catalog.catalog.count.should == 6
+      @catalog.where('_resource' => 'testdata', 'ID' => '3').should == {'ID' => '3', 'name' => 'Benjamin Franklin', 'title' => 'Sage'}
     end
 
     it "should refuse to do a duplicate import" do
@@ -67,10 +66,11 @@ describe "importing file" do
         :deletes_count => 0,
         :deleted_ids => [],
         :updates_count => 0,
+        :updated_ids => [],
         :unchanged_count => 6,
         :updated_fields => {}
       }
-      @catalog.catalog.count.should == @records.count
+      @catalog.catalog.count.should == 6
     end
   end
 
@@ -93,10 +93,11 @@ describe "importing file" do
         :deletes_count => 2,
         :deleted_ids => ['4', '6'],
         :updates_count => 1,
+        :updated_ids => ['3'],
         :unchanged_count => 3,
         :updated_fields => {'title' => 1}
       }
-      @catalog.active_count.should == @records.count
+      @catalog.active_count.should == 5
     end
   end
 end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: assimilate
 version: !ruby/object:Gem::Version
-  version: 0.0.3
+  version: 0.0.4
   prerelease: 
 platform: ruby
 authors:
@@ -13,7 +13,7 @@ date: 2012-05-02 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: mongo
-  requirement: &2152913700 !ruby/object:Gem::Requirement
+  requirement: &2152558540 !ruby/object:Gem::Requirement
     none: false
     requirements:
     - - ~>
@@ -21,10 +21,10 @@ dependencies:
         version: 1.6.0
   type: :runtime
   prerelease: false
-  version_requirements: *2152913700
+  version_requirements: *2152558540
 - !ruby/object:Gem::Dependency
   name: bson_ext
-  requirement: &2152912480 !ruby/object:Gem::Requirement
+  requirement: &2152557720 !ruby/object:Gem::Requirement
     none: false
     requirements:
     - - ~>
@@ -32,10 +32,10 @@ dependencies:
         version: 1.6.0
   type: :runtime
   prerelease: false
-  version_requirements: *2152912480
+  version_requirements: *2152557720
 - !ruby/object:Gem::Dependency
   name: activesupport
-  requirement: &2152911920 !ruby/object:Gem::Requirement
+  requirement: &2152556760 !ruby/object:Gem::Requirement
     none: false
     requirements:
     - - ~>
@@ -43,10 +43,10 @@ dependencies:
         version: 3.2.0
   type: :runtime
   prerelease: false
-  version_requirements: *2152911920
+  version_requirements: *2152556760
 - !ruby/object:Gem::Dependency
   name: rspec
-  requirement: &2152911460 !ruby/object:Gem::Requirement
+  requirement: &2152555800 !ruby/object:Gem::Requirement
     none: false
     requirements:
     - - ~>
@@ -54,10 +54,10 @@ dependencies:
         version: 2.9.0
   type: :development
   prerelease: false
-  version_requirements: *2152911460
+  version_requirements: *2152555800
 - !ruby/object:Gem::Dependency
   name: guard-rspec
-  requirement: &2152910880 !ruby/object:Gem::Requirement
+  requirement: &2152554060 !ruby/object:Gem::Requirement
     none: false
     requirements:
     - - ~>
@@ -65,10 +65,10 @@ dependencies:
         version: 0.7.0
   type: :development
   prerelease: false
-  version_requirements: *2152910880
+  version_requirements: *2152554060
 - !ruby/object:Gem::Dependency
   name: ruby_gntp
-  requirement: &2152910140 !ruby/object:Gem::Requirement
+  requirement: &2152550400 !ruby/object:Gem::Requirement
     none: false
     requirements:
     - - ~>
@@ -76,7 +76,7 @@ dependencies:
         version: 0.3.4
   type: :development
   prerelease: false
-  version_requirements: *2152910140
+  version_requirements: *2152550400
 description: Ingest updates from CSV and apply to set of hashes
 email:
 - jmay@pobox.com