dbd 0.0.14 → 0.0.15

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: a4df1f0e326b407d9c2f0391a7ae7b0980a2deae
-  data.tar.gz: a9724f1c990e1fdeaa3c3159ae059b26de4ddc4c
+  metadata.gz: a631ac1b673b24327126207b5898c6132a08f41f
+  data.tar.gz: 2abdda7d7da624dace39c486381130a813c28e92
 SHA512:
-  metadata.gz: 391388832659492f0c00259140f97c1626820a0473373d5491ef9fb7e1fca6f1dd0badc3be4fca3bbbb2c567b52bdfa37bfef14bc1efe60f40beaa22813bc0ca
-  data.tar.gz: ca0facecd89bca3dfffb24865b8cf14e4676582d8254fed22e26ca268738c61ffc35f4f3fc7b755c8b4e728779e14579a8ae99a91be6f17f306f3c10f2cf490e
+  metadata.gz: 03eac42ce2d9216371065f46a2ab8ab0dc19a88b9270f59cdcd50496d771dc4852fd30f471514028b00da6ed1de71787c8767388c697158e21b73e0e625b636a
+  data.tar.gz: 4baa8c0d598d1982442ee8c31a1e612939b5ae39c0167eb32eab25ebfa919c3a37523bb3a249a35f504d4bd67a08cdd65bdc76a2d9010846420ba7a3ab1fc265

data/HISTORY.txt

@@ -90,4 +90,9 @@
 ======
 
 * in to_CSV the time_stamp now comes first (swapped with id)
-  * this will allow file level 'sort' on the ISO dates 
+  * this will allow file level 'sort' on the ISO dates
+
+0.0.15 (25 Aug 2013)
+======
+
+* better exception message for an OutOfOrder exception

data/README.md

@@ -1,6 +1,8 @@
 # Dbd: A data store that (almost) never forgets
 
-This is facts based data store, inspired by [RDF] concepts, but adding a log based structure and fine-grained provenance.
+This is facts based data store, inspired by [RDF] concepts, but adding a
+log based structure and fine-grained context (with provenance) for each fact.
+I am building a simple demo application in [DbdDemo].
 
 * [Why?][Rationale]
 * <https://github.com/petervandenabeele/dbd>
@@ -150,7 +152,7 @@ This version also has input validation on the strings in the CSV. The time neede
 and memory size (RSS) for reading the file (and populating the in-memory dataset
 was):
 
-Read from CSV (to_CSV) 10 M facts (of 250 Bytes; 2.5 GB netto data):
+Read from CSV (from_CSV) 10 M facts (of 250 Bytes; 2.5 GB netto data):
 
 | ruby	     | time          |  memory (RSS) |
 |------------|---------------|--------------:|
@@ -169,3 +171,4 @@ gsub with a regexp).
 [RDF]:              http://www.w3.org/RDF/
 [Rationale]:        http://github.com/petervandenabeele/dbd/blob/master/docs/rationale.md
 [MIT]:              https://github.com/petervandenabeele/dbd/blob/master/LICENSE.txt
+[DbdDemo]:          https://github.com/petervandenabeele/dbd_demo#readme

data/lib/dbd/context_fact.rb

@@ -32,15 +32,6 @@ module Dbd
       super
     end
 
-    ##
-    # Executes the required update in used_context_subjects.
-    #
-    # For a ContextFact, there is no context_subject, so
-    # pointless to mark it in used_context_subjects hash.
-    def update_used_context_subjects(h)
-      # Do nothing (override the behaviour from super).
-    end
-
     ##
     # Validates the presence or absence of context_subject.
     #

data/lib/dbd/fact.rb

@@ -177,22 +177,6 @@ module Dbd
       "#{object.to_s.truncate_utf8(80).gsub(/\n/, '_')}"
     end
 
-    ##
-    # Executes the required update in used_context_subjects.
-    #
-    # For a Fact, pointing to a Context in it's context_subject,
-    # marks this context_subject in the "used_context_subjects" hash that
-    # is passed in as an argument (DCI). This will avoid further changes to the
-    # Context with this context_subject.
-    #
-    # This is overridden in the ContextFact, since only relevant for a Fact.
-    #
-    # @param[Hash] :h the hash that contains the context_subject index
-    def update_used_context_subjects(h)
-      # using a context_subject sets the key
-      h[context_subject] = true
-    end
-
     ##
     # Checks if a fact has errors for storing in the graph.
     #

data/lib/dbd/fact/collection.rb

@@ -9,7 +9,6 @@ module Dbd
       def initialize
         super
         @hash_by_subject = Hash.new { |h, k| h[k] = [] }
-        @used_context_subjects = {}
       end
 
       def newest_time_stamp
@@ -33,20 +32,14 @@ module Dbd
       #
       # Validates that added fact is newer.
       #
-      # Validates that subject was never used as context_subject [A].
-      #
       # Adds the fact and return the index in the collection.
       #
       # Store this index in the hash_by_subject.
-      #
-      # Mark the fact in the list of used context_subjects (for [A]).
       def <<(fact)
         raise FactError, "#{fact.errors.join(', ')}." unless fact.errors.empty?
-        raise OutOfOrderError if (self.newest_time_stamp && fact.time_stamp <= self.newest_time_stamp)
-        raise OutOfOrderError if (@used_context_subjects[fact.subject])
+        validate_time_stamp(fact)
         index = Helpers::OrderedSetCollection.add_and_return_index(fact, @internal_collection)
         @hash_by_subject[fact.subject] << index
-        fact.update_used_context_subjects(@used_context_subjects)
         self
       end
 
@@ -54,6 +47,14 @@ module Dbd
         @hash_by_subject[fact_subject].map{ |index| @internal_collection[index]}
       end
 
+    private
+
+      def validate_time_stamp(fact)
+        if (newest_time_stamp && fact.time_stamp <= newest_time_stamp)
+          raise OutOfOrderError,  "time_stamp of fact was too old : #{fact.time_stamp}"
+        end
+      end
+
     end
   end
 end

data/lib/dbd/version.rb

@@ -1,3 +1,3 @@
 module Dbd
-  VERSION = "0.0.14"
+  VERSION = "0.0.15"
 end

data/spec/lib/dbd/context_fact/methods_spec.rb

@@ -52,14 +52,6 @@ module Dbd
       end
     end
 
-    describe 'update_used_context_subjects' do
-      it 'does nothing for a context_fact' do
-        h = {}
-        context_fact_1.update_used_context_subjects(h)
-        h.should be_empty
-      end
-    end
-
     describe 'attributes and values' do
       it 'there are 6 attributes' do
         described_class.attributes.size.should == 6

data/spec/lib/dbd/fact/collection/collection_spec.rb

@@ -110,12 +110,16 @@ module Dbd
         it 'adding an element with an older time_stamp fails' do
           fact_2_with_subject # will be older then fact_3_with_subject
           subject << fact_3_with_subject
-          lambda { subject << fact_2_with_subject } . should raise_error OutOfOrderError
+          lambda{ subject << fact_2_with_subject }.should raise_error(
+            OutOfOrderError,
+            "time_stamp of fact was too old : #{fact_2_with_subject.time_stamp}")
         end
 
         it 'adding an element with an equal time_stamp fails' do
           subject << fact_2_with_subject
-          lambda { subject << fact_2_with_subject } . should raise_error OutOfOrderError
+          lambda{ subject << fact_2_with_subject }.should raise_error(
+            OutOfOrderError,
+            "time_stamp of fact was too old : #{fact_2_with_subject.time_stamp}")
         end
       end
 
@@ -136,41 +140,6 @@ module Dbd
         end
       end
 
-      describe 'context_facts must all come before first use by a fact' do
-        it 'adding a context_fact, depending fact, another context_fact with same subject fail' do
-          subject << context_fact_visibility
-          subject << fact_2_with_subject
-          lambda{ subject << context_fact_created_by }.should raise_error OutOfOrderError
-        end
-
-        # testing private functionality (kept temporarily as documentation)
-        # A hash with all the context_subjects that are used by at least one fact.
-        # Needed for the validation that no context_fact may be added that is
-        # referred from a fact that is already in the fact stream.
-        describe 'used_context_subjects' do
-          # testing an internal variable ...
-
-          let(:used_context_subjects) do
-            subject.instance_variable_get(:@used_context_subjects)
-          end
-
-          it 'is empty initially' do
-            used_context_subjects.should be_empty
-          end
-
-          it 'adding a context_fact alone does not create an entry' do
-            subject << context_fact_visibility
-            used_context_subjects.should be_empty
-          end
-
-          it 'adding a context_fact and a depending fact create an entry' do
-            subject << context_fact_visibility
-            subject << fact_2_with_subject
-            used_context_subjects[context_subject_1].should == true
-          end
-        end
-      end
-
       describe 'validate that facts do not have errors when loading in the Fact::Collection' do
         it 'succeeds with a fact from factory' do
            subject << fact_2_with_subject # should_not raise_error

data/spec/lib/dbd/fact/new_spec.rb

@@ -74,13 +74,5 @@ module Dbd
         end
       end
     end
-
-    describe 'update_used_context_subjects' do
-      it 'sets the value for context_subject to true for a fact' do
-        h = {}
-        fact_1.update_used_context_subjects(h)
-        h[fact_1.context_subject].should == true
-      end
-    end
   end
 end

data/spec/lib/dbd/graph/from_csv_spec.rb

@@ -89,5 +89,14 @@ module Dbd
         graph_from_CSV.first.should be_equivalent(graph.first)
       end
     end
+
+    describe 'error handling' do
+      it 'gives better description on incorrect fact size' do
+        #TODO present a nicer error (check the amount of entries in a CSV input line)
+        csv = 'foo, bar'
+        lambda { described_class.new.from_CSV(csv) }.should raise_error IndexError
+      end
+
+    end
   end
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: dbd
 version: !ruby/object:Gem::Version
-  version: 0.0.14
+  version: 0.0.15
 platform: ruby
 authors:
 - Peter Vandenabeele
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2013-08-09 00:00:00.000000000 Z
+date: 2013-08-25 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: bundler