alf 0.10.0 → 0.10.1

data/CHANGELOG.md

@@ -1,3 +1,25 @@
+# 0.10.1 / 2011-08-31
+
+## Miscellaneous enhancements
+
+* Added Variance and Stddev aggregation operators; they are available under 
+  Aggregator.variance{} and Aggregator.stddev{}, respectively
+* Added a --pretty option to 'alf', whose semantics is delegated to the output 
+  renderer. Same option is available on 'alf show'.
+* 'alf show' now accepts an optional ordering argument. This avoids explicitely
+  including non-relational sort invocations in your pipe (sorting should be 
+  seen as a displaying issue)
+* Added an options hash to text renderer, :float_format among others
+* Added a --ff option to 'alf show', for float format in text rendering
+
+## Bugfixes
+
+* Alf::Environment::(Folder/Explicit)#dataset now correctly raises a 
+  NoSuchDatasetError when the dataset cannot be found, as required by the
+  specification.
+* Alf::Reader.reader now correctly returns a Rash reader when invoked on a 
+  StringIO
+
 # 0.10.0 / 2011-08-15
 
 ## New recognized data sources

data/Gemfile

@@ -1,8 +1,9 @@
 source 'http://rubygems.org'
 
 group :runtime do
-  gem "quickl", "~> 0.4.0"
-  gem "myrrha", "~> 1.2.0"
+  gem "quickl", "~> 0.4.1"
+  gem "myrrha", "~> 1.2.1"
+  gem "highline", "~> 1.6.2"
 end
 
 group :test do
@@ -17,7 +18,6 @@ group :release do
 end
 
 group :doc do
-  gem "rcov", "~> 0.9.9"
   gem "yard", "~> 0.7.2"
   gem "bluecloth", "~> 2.1.0"
   gem "redcarpet", "~> 1.17.2"

data/Gemfile.lock

@@ -2,15 +2,14 @@ GEM
   remote: http://rubygems.org/
   specs:
     bluecloth (2.1.0)
-    diff-lcs (1.1.2)
+    diff-lcs (1.1.3)
     fastercsv (1.5.4)
-    myrrha (1.2.0)
-    quickl (0.4.0)
+    highline (1.6.2)
+    myrrha (1.2.1)
+    quickl (0.4.1)
     rake (0.9.2)
-    rcov (0.9.9)
-    rcov (0.9.9-java)
     redcarpet (1.17.2)
-    request-log-analyzer (1.11.0)
+    request-log-analyzer (1.11.1)
     rspec (2.6.0)
       rspec-core (~> 2.6.0)
       rspec-expectations (~> 2.6.0)
@@ -20,7 +19,7 @@ GEM
       diff-lcs (~> 1.1.2)
     rspec-mocks (2.6.0)
     sequel (3.25.0)
-    sqlite3 (1.3.3)
+    sqlite3 (1.3.4)
     wlang (0.10.2)
     yard (0.7.2)
 
@@ -31,10 +30,10 @@ PLATFORMS
 DEPENDENCIES
   bluecloth (~> 2.1.0)
   fastercsv (~> 1.5.4)
-  myrrha (~> 1.2.0)
-  quickl (~> 0.4.0)
+  highline (~> 1.6.2)
+  myrrha (~> 1.2.1)
+  quickl (~> 0.4.1)
   rake (~> 0.9.2)
-  rcov (~> 0.9.9)
   redcarpet (~> 1.17.2)
   request-log-analyzer (~> 1.11.0)
   rspec (~> 2.6.0)

data/alf.gemspec

@@ -126,7 +126,6 @@ Gem::Specification.new do |s|
   s.add_development_dependency("rake", "~> 0.9.2")
   s.add_development_dependency("rspec", "~> 2.6.0")
   s.add_development_dependency("wlang", "~> 0.10.2")
-  s.add_development_dependency("rcov", "~> 0.9.9")
   s.add_development_dependency("yard", "~> 0.7.2")
   s.add_development_dependency("bluecloth", "~> 2.1.0")
   s.add_development_dependency("redcarpet", "~> 1.17.2")
@@ -134,8 +133,9 @@ Gem::Specification.new do |s|
   s.add_development_dependency("request-log-analyzer", "~> 1.11.0")
   s.add_development_dependency("sequel", "~> 3.25.0")
   s.add_development_dependency("sqlite3", "~> 1.3.0")
-  s.add_dependency("quickl", "~> 0.4.0")
-  s.add_dependency("myrrha", "~> 1.2.0")
+  s.add_dependency("quickl", "~> 0.4.1")
+  s.add_dependency("myrrha", "~> 1.2.1")
+  s.add_dependency("highline", "~> 1.6.2")
 
   # The version of ruby required by this gem
   #

data/alf.noespec

@@ -16,7 +16,7 @@ variables:
   upper:
     Alf
   version:
-    0.10.0
+    0.10.1
   summary: |-
     Relational Algebra at your fingertips
   description: |-
@@ -33,12 +33,13 @@ variables:
     - http://rubygems.org/gems/alf
     - http://rubydoc.info/gems/alf
   dependencies: 
-    - {name: quickl,       version: "~> 0.4.0",  groups: [runtime      ]}
-    - {name: myrrha,       version: "~> 1.2.0",  groups: [runtime      ]}
+    - {name: quickl,       version: "~> 0.4.1",  groups: [runtime      ]}
+    - {name: myrrha,       version: "~> 1.2.1",  groups: [runtime      ]}
+    - {name: highline,     version: "~> 1.6.2",  groups: [runtime      ]}
     - {name: rake,         version: "~> 0.9.2",  groups: [test, release]}
     - {name: rspec,        version: "~> 2.6.0",  groups: [test, release]}
     - {name: wlang,        version: "~> 0.10.2", groups: [release      ]}
-    - {name: rcov,         version: "~> 0.9.9",  groups: [doc          ]}
+    #- {name: rcov,         version: "~> 0.9.9",  groups: [doc          ]}
     - {name: yard,         version: "~> 0.7.2",  groups: [doc          ]}
     - {name: bluecloth,    version: "~> 2.1.0",  groups: [doc          ]}
     - {name: redcarpet,    version: "~> 1.17.2", groups: [doc          ]}

data/doc/commands/show.md

@@ -3,7 +3,7 @@ Output input tuples through a specific renderer (text, yaml, ...)
 
 SYNOPSIS
 
-    alf #(command_name) DATASET
+    alf #(command_name) DATASET -- [ORDERING]
 
 OPTIONS
 

data/doc/operators/relational/heading.md

@@ -0,0 +1,20 @@
+
+Relational heading (print the relation type)
+
+SYNOPSIS
+
+    #(signature)
+
+DESCRIPTION
+
+This operator simply outputs a relation containing one tuple: the heading of 
+its operand. 
+
+EXAMPLE
+
+    # What is the heading of the `suppliers` relation?
+    !(alf heading suppliers)
+
+    # What is the heading of the following join?
+    !(alf join suppliers supplies | alf heading)
+

data/doc/operators/relational/join.md

@@ -9,9 +9,10 @@ DESCRIPTION
 
 This operator computes the (natural) join of its operands. 
 
-Natural join means that, unlike what is commonly used in SQL, the join is 
-performed on common attribute names. You can use the `rename` operator if this 
-behavior does not fit your needs.
+Natural join means that, unlike what is commonly used in SQL, joining is based 
+on equality of all attributes that are shared between LEFT and RIGHT performed 
+on common attribute names. You can use the `rename` operator if this  behavior 
+does not fit your needs.
 
 When operands have no attribute in common, this operator naturally "degenerates" 
 to a cartesian product.

data/doc/operators/relational/matching.md

@@ -7,12 +7,16 @@ SYNOPSIS
 
 DESCRIPTION
 
-This operator restricts its LEFT operand to tuples for which there exists 
-at least one tuple in RIGHT that (naturally) joins. This is a shortcut 
-operator for the following longer expression:
+This operator restricts its LEFT operand to tuples for which there exists at 
+least one tuple in RIGHT that (naturally) joins. This is a shortcut operator 
+for the following longer expression:
 
     (project (join xxx, yyy), [xxx's attributes])
 
+Or equivalently,
+
+    (join xxx, (project xxx, [common attributes]))
+
 EXAMPLE
 
     # Which suppliers supply at least one part?

data/doc/operators/relational/restrict.md

@@ -7,8 +7,8 @@ SYNOPSIS
 
 DESCRIPTION
 
-This command restricts tuples to those for which PREDICATE evaluates to 
-true.
+This command restricts restricts relations to those tuples for which PREDICATE 
+evaluates to true.
 
 PREDICATE must be a valid tuple expression that returns a truth-value. It 
 may be specified as a ruby code literal, or a mapping between (name, value) 

data/examples/hitradio/compact.rb

@@ -0,0 +1,48 @@
+require 'rubygems'
+require 'alf'
+require 'benchmark'
+
+#rel = Alf.lispy('.').compile{ environment.dataset("hitradio.log") } 
+#keys = nil
+#rel.each do |tuple|
+#  unless keys
+#    keys = tuple.keys
+#    puts Alf::Tools.to_ruby_literal(keys)
+#  end
+#  values = keys.collect{|k| tuple[k]}
+#  puts Alf::Tools.to_ruby_literal(values)
+#end
+
+m = Benchmark.measure do 
+  keys = nil
+  File.open("hitradio.mar").each_line do |line|
+    eval(line)
+  end
+end
+puts "hitradio.mar: #{m}"
+
+m = Benchmark.measure do 
+  rel = Alf.lispy('.').compile{ environment.dataset("hitradio.log") }
+  rel.each do |tuple|
+  end 
+end
+puts "hitradio.log: #{m}"
+
+m = Benchmark.measure do 
+  rel = Alf.lispy('.').compile{ environment.dataset("hitradio.rash") }
+  rel.each do |tuple|
+  end 
+end
+puts "hitradio.rash: #{m}"
+
+require 'zlib'
+m = Benchmark.measure do 
+  File.open("hitradio.mar.gz", "r") do |io|
+    reader = Zlib::GzipReader.new(io)
+    reader.each_line do |line|
+      eval(line)
+    end
+  end
+end
+puts "hitradio.mar.gz: #{m}"
+

data/examples/hitradio/compare1.rb

@@ -0,0 +1,28 @@
+require 'rubygems'
+require 'alf'
+require File.expand_path('../mar_reader', __FILE__)
+require File.expand_path('../gz_reader', __FILE__)
+require 'benchmark'
+
+def which_ruby
+  if Object.const_defined?(:RUBY_DESCRIPTION)
+    RUBY_DESCRIPTION =~ /^([^\s]+\s*[^\s]+)/
+    $1
+  else
+    "ruby #{RUBY_VERSION} (#{RUBY_PLATFORM})"
+  end
+end
+
+
+log  = Alf::Environment.folder('.').dataset("hitradio.log")
+rash = Alf::Environment.folder('.').dataset("hitradio.rash")
+mar  = Alf::Environment.folder('.').dataset("hitradio.mar")
+gz   = Alf::Environment.folder('.').dataset("hitradio.mar.gz")
+
+puts which_ruby
+Benchmark.bm do |bench|
+  bench.report('.log'){ log.each{} }
+  bench.report('.rash'){ rash.each{} }
+  bench.report('.mar'){ mar.each{} }
+  bench.report('.gz'){ gz.each{} }
+end

data/examples/hitradio/gz_reader.rb

@@ -0,0 +1,23 @@
+require 'alf'
+require 'zlib'
+class GzReader < Alf::Reader
+
+  def each
+    keys = nil
+    with_input_io do |io|
+      reader = Zlib::GzipReader.new(io)
+      reader.each_line do |line|
+        decoded = eval(line)
+        if keys.nil?
+          keys = decoded
+        else
+          yield Hash[keys.zip(decoded)]
+        end
+      end
+    end
+  end
+
+  Alf::Reader.register(:gz, [".gz"], self)  
+end
+
+

data/examples/hitradio/hitradio-1.alf

@@ -0,0 +1 @@
+(restrict (rank (summarize "hitradio.log", [:path], :weight => sum{ bytes_sent }), [[:weight, :desc]], :podium), lambda{ podium < 10 })