ruby-nuggets 0.9.3 → 0.9.4

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 94ea66bd359c2d1ab6160014687f6a133965b0b6
-  data.tar.gz: f243ceb32806ffec4f5d0ea04e5d673c18c0d42c
+  metadata.gz: 1310c734144767d70dee0cd8479b0077f54f5a45
+  data.tar.gz: 6e76159514c5a08df644c83baf042175705a5497
 SHA512:
-  metadata.gz: d09992bb8f671087d09cba6b9f9aa0720e99e4ed8be878c8153b23510e26876cf3677b62af51d95290a0e535a2daddd22dd7d6a1fc78c55b2e105972a9a0e4fc
-  data.tar.gz: b16039377e13cacb573c2708c6d43684ca6f8fa6efdd4e87f8f74e809cedae505b5d9886f845d8bd3570a7de18270d0893492f6d0297c47891e4fb4942a2aa93
+  metadata.gz: 4cb2f5ba60d908871b3802a0d4795a3b390a2cb5a83f004e484677a8940288d2bf9d1728d191629040629b531a90073b57ee872de80a99612f02a97101764329
+  data.tar.gz: 981e031f0731040b3b08857ca9a772fe48c5af564b47a0367ce9c80aa777b6ae7283b0394ceaee04696bdf214781c010784430576efbd3f05f844852713ebad5

data/ChangeLog

@@ -1,3 +1,5 @@
+# markup: rd
+
 = Revision history for ruby-nuggets
 
 == x.y.z [yyyy-mm-dd]

data/README

@@ -2,7 +2,7 @@
 
 == VERSION
 
-This documentation refers to ruby-nuggets version 0.9.3
+This documentation refers to ruby-nuggets version 0.9.4
 
 
 == DESCRIPTION

data/lib/nuggets/argv/option.rb

@@ -0,0 +1,3 @@
+require 'nuggets/argv/option_mixin'
+
+ARGV.extend(Nuggets::Argv::OptionMixin)

data/lib/nuggets/argv/option_mixin.rb

@@ -0,0 +1,85 @@
+#--
+###############################################################################
+#                                                                             #
+# A component of ruby-nuggets, some extensions to the Ruby programming        #
+# language.                                                                   #
+#                                                                             #
+# Copyright (C) 2007-2013 Jens Wille                                          #
+#                                                                             #
+# Authors:                                                                    #
+#     Jens Wille <jens.wille@gmail.com>                                       #
+#                                                                             #
+# ruby-nuggets is free software; you can redistribute it and/or modify it     #
+# under the terms of the GNU Affero General Public License as published by    #
+# the Free Software Foundation; either version 3 of the License, or (at your  #
+# option) any later version.                                                  #
+#                                                                             #
+# ruby-nuggets is distributed in the hope that it will be useful, but WITHOUT #
+# ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or       #
+# FITNESS FOR A PARTICULAR PURPOSE. See the GNU Affero General Public License #
+# for more details.                                                           #
+#                                                                             #
+# You should have received a copy of the GNU Affero General Public License    #
+# along with ruby-nuggets. If not, see <http://www.gnu.org/licenses/>.        #
+#                                                                             #
+###############################################################################
+#++
+
+module Nuggets
+  module Argv
+    module OptionMixin
+
+  # call-seq:
+  #   ARGV.switch(short[, long]) -> true | false
+  #
+  # Whether ARGV includes the switch +short+ (or +long+).
+  def switch(*args)
+    !!(__key(*args) { |key| include?(key) })
+  end
+
+  # call-seq:
+  #   ARGV.option(short[, long]) -> aString
+  #   ARGV.option(short[, long]) { |value| ... } -> anObject
+  #
+  # Returns the value associated with the option +short+ (or +long+) if present
+  # in ARGV. Yields that value to the block if given and returns its result.
+  def option(*args, &block)
+    __opt(block, *args) { |index| at(index + 1) }
+  end
+
+  # call-seq:
+  #   ARGV.switch!(short[, long]) -> true | false
+  #
+  # Whether ARGV includes the switch +short+ (or +long+). Removes the matching
+  # switch from ARGV.
+  def switch!(*args)
+    !!(__key(*args) { |key| delete(key) })
+  end
+
+  # call-seq:
+  #   ARGV.option!(short[, long]) -> aString
+  #   ARGV.option!(short[, long]) { |value| ... } -> anObject
+  #
+  # Returns the value associated with the option +short+ (or +long+) if present
+  # in ARGV and removes both from ARGV. Yields that value to the block if given
+  # and returns its result.
+  def option!(*args, &block)
+    __opt(block, *args) { |index| delete_at(index); delete_at(index) }
+  end
+
+  private
+
+  def __key(short, long = nil)  # :yield: key
+    long && yield("--#{long}") || yield("-#{short}")
+  end
+
+  def __opt(block, *args)
+    index = __key(*args) { |key| index(key) } or return
+
+    value = yield(index)
+    block ? block[value] : value
+  end
+
+    end
+  end
+end

data/lib/nuggets/lsi.rb

@@ -0,0 +1,198 @@
+#--
+###############################################################################
+#                                                                             #
+# A component of ruby-nuggets, some extensions to the Ruby programming        #
+# language.                                                                   #
+#                                                                             #
+# Copyright (C) 2007-2013 Jens Wille                                          #
+#                                                                             #
+# Authors:                                                                    #
+#     Jens Wille <jens.wille@gmail.com>                                       #
+#                                                                             #
+# ruby-nuggets is free software; you can redistribute it and/or modify it     #
+# under the terms of the GNU Affero General Public License as published by    #
+# the Free Software Foundation; either version 3 of the License, or (at your  #
+# option) any later version.                                                  #
+#                                                                             #
+# ruby-nuggets is distributed in the hope that it will be useful, but WITHOUT #
+# ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or       #
+# FITNESS FOR A PARTICULAR PURPOSE. See the GNU Affero General Public License #
+# for more details.                                                           #
+#                                                                             #
+# You should have received a copy of the GNU Affero General Public License    #
+# along with ruby-nuggets. If not, see <http://www.gnu.org/licenses/>.        #
+#                                                                             #
+###############################################################################
+#++
+
+require 'gsl'
+
+module Nuggets
+
+  class LSI
+
+    include Enumerable
+
+    DEFAULT_EPSILON = Float::EPSILON * 10
+
+    def self.each_norm(items, options = {}, &block)
+      lsi = new(items)
+      lsi.each_norm(nil, options, &block) if lsi.build
+    end
+
+    def initialize(items = {})
+      reset
+      items.each { |k, v| self[k] = v }
+    end
+
+    def [](key)
+      @hash[key]
+    end
+
+    def []=(key, value)
+      @hash[key] = Doc.new(key, value, @list)
+    end
+
+    def add(key, value = key)
+      self[key] = value
+      self
+    end
+
+    def <<(value)
+      add(value.object_id, value)
+    end
+
+    def size
+      @hash.size
+    end
+
+    def keys
+      @hash.keys
+    end
+
+    def docs
+      @hash.values
+    end
+
+    def each(&block)
+      @hash.each(&block)
+    end
+
+    # min:: minimum value to consider
+    # abs:: minimum absolute value to consider
+    # nul:: exclude null values (true or Float)
+    # new:: exclude original terms / only yield new ones
+    def each_norm(key = nil, options = {})
+      min, abs, nul, new = options.values_at(:min, :abs, :nul, :new)
+      nul = DEFAULT_EPSILON if nul == true
+
+      list = @invlist
+
+      (key ? [self[key]] : docs).each { |doc|
+        if doc && norm = doc.norm
+          i = 0
+
+          norm.each { |v|
+            yield doc, list[i], v unless (min && v < min) ||
+                                         (abs && v.abs < abs) ||
+                                         (nul && v.abs < nul) ||
+                                         (new && doc.include?(i))
+            i += 1
+          }
+        end
+      }
+    end
+
+    def related(key, num = 5)
+      if doc = self[key] and norm = doc.norm
+        a = []; norm *= -1
+        each { |k, v| a << [norm * v.norm.col, k] unless k == key }
+        a.sort![0, num].map! { |_, k| k }
+      end
+    end
+
+    def build(cutoff = 0.75)
+      build!(docs, @list, cutoff) if size > 1
+    end
+
+    def reset
+      @hash, @list, @invlist = {}, Hash.new { |h, k| h[k] = h.size }, {}
+    end
+
+    private
+
+    def build!(docs, list, cutoff)
+      u, v, s = GSL::Matrix.alloc(*vectors(docs, list)).trans.SV_decomp
+      reduce(u, v, cutoff(s, cutoff), docs)
+      size
+    end
+
+    def vectors(docs, list)
+      @invlist, size = list.invert, list.size
+      docs.map { |doc| transform(doc.raw_vector(size)) }
+    end
+
+    # FIXME: "first-order association transform" ???
+    def transform(vec, q = 0)
+      return vec #unless (sum = vec.sum) > 1
+
+      vec.each { |v| q -= (w = v / sum) * Math.log(w) if v > 0 }
+      vec.map! { |v| Math.log(v + 1) / q }
+    end
+
+    def cutoff(s, c)
+      w, i = s.sort[-(s.size * c).round], 0
+      s.each { |v| s[i] = 0 if v < w; i += 1 }
+      s
+    end
+
+    def reduce(u, v, s, d, i = -1)
+      (u * GSL::Matrix.diagonal(s) * v.trans).each_col { |c|
+        d[i += 1].vector = c.row
+      }
+    end
+
+    class Doc
+
+      TOKEN_RE = %r{\s+}
+
+      def initialize(key, value, list)
+        @key = key
+        @map = !value.is_a?(Hash) ? build_hash(value, list) :
+          value.inject({}) { |h, (k, v)| h[list[k]] = v; h }
+      end
+
+      attr_reader :key, :vector, :norm
+
+      def raw_vector(size)
+        vec = GSL::Vector.alloc(size)
+        @map.each { |k, v| vec[k] = v }
+        vec
+      end
+
+      def vector=(vec)
+        @vector, @norm = vec, vec.normalize
+      end
+
+      def include?(k)
+        @map.include?(k)
+      end
+
+      private
+
+      def build_hash(value, list, hash = Hash.new(0))
+        build_array(value).each { |i| hash[list[i]] += 1 }
+        hash
+      end
+
+      def build_array(value, re = TOKEN_RE)
+        value = value.read if value.respond_to?(:read)
+        value = value.split(re) if value.respond_to?(:split)
+        value
+      end
+
+    end
+
+  end
+
+end

data/lib/nuggets/midos.rb

@@ -42,9 +42,22 @@ module Nuggets
       # Line break indicator
       DEFAULT_NL = '^'
 
-      def self.parse(input, *args, &block)
-        parser = new(*args).parse(input, &block)
-        block_given? ? parser : parser.records
+      # Default encoding for ::parse_file.
+      DEFAULT_ENCODING = 'iso-8859-1'
+
+      class << self
+
+        def parse(input, *args, &block)
+          parser = new(*args).parse(input, &block)
+          block_given? ? parser : parser.records
+        end
+
+        def parse_file(file, encoding = nil, *args, &block)
+          File.open(file, :encoding => encoding || DEFAULT_ENCODING) { |input|
+            parse(input, *args, &block)
+          }
+        end
+
       end
 
       def initialize(options = {})

data/lib/nuggets/version.rb

@@ -4,7 +4,7 @@ module Nuggets
 
     MAJOR = 0
     MINOR = 9
-    TINY  = 3
+    TINY  = 4
 
     class << self
 

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: ruby-nuggets
 version: !ruby/object:Gem::Version
-  version: 0.9.3
+  version: 0.9.4
 platform: ruby
 authors:
 - Jens Wille
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2013-04-15 00:00:00.000000000 Z
+date: 2013-06-07 00:00:00.000000000 Z
 dependencies: []
 description: Some extensions to the Ruby programming language.
 email: jens.wille@gmail.com
@@ -23,6 +23,8 @@ files:
 - lib/nuggets/all.rb
 - lib/nuggets/all_mixins.rb
 - lib/nuggets/ansicolor2css.rb
+- lib/nuggets/argv/option.rb
+- lib/nuggets/argv/option_mixin.rb
 - lib/nuggets/array/boost.rb
 - lib/nuggets/array/boost_mixin.rb
 - lib/nuggets/array/combination.rb
@@ -105,6 +107,7 @@ files:
 - lib/nuggets/log_parser.rb
 - lib/nuggets/log_parser/apache.rb
 - lib/nuggets/log_parser/rails.rb
+- lib/nuggets/lsi.rb
 - lib/nuggets/midos.rb
 - lib/nuggets/mysql.rb
 - lib/nuggets/net/success.rb
@@ -224,7 +227,7 @@ rdoc_options:
 - --line-numbers
 - --all
 - --title
-- ruby-nuggets Application documentation (v0.9.3)
+- ruby-nuggets Application documentation (v0.9.4)
 - --main
 - README
 require_paths: