stimmung 0.0.2 → 0.0.3

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: cabcb933856e7bbdf3aa844be363c1ce1eb3d341
-  data.tar.gz: 7d2dd009a5b9ddb6076f9ca3a4326fb9ccb5094c
+  metadata.gz: 6f9169fc6c1557a510676990396f70de9a8b4cf6
+  data.tar.gz: 07d8be74953098bdd5a104475e03518df04baf15
 SHA512:
-  metadata.gz: acd7f86db207dc8748c61e33927f169e2ca60030ff5cc20958db98d11243790041c7c46c5d648b6ba4f9256da40ce32d581c07de9fc3c2c44a3c60c982afc876
-  data.tar.gz: 4f5beb6caddecd13c09f11d5ebd0a6101ad958b428a55ade161139739c7191a2deaf69bc63c9db64cc7f88e083c4222c7a2697cdeb1690886a0f385a4058c3fc
+  metadata.gz: 497d9424d744ab4b54775857f12b1dfb0293fe289d3f3acb5887bda1ae01e70c84bedcbdefd2b8f22e47900020038c0e284fdb6f75041106e28d2deba9d384cc
+  data.tar.gz: c46e355f12a984014826a005a18e979c3022f9a8d7c27c30a874b928aacc95ad01690d7e99297dfce19ee963698b6cec96c1d137f362427cf4879403304fbd62

data/lib/emoticons.txt

@@ -40,7 +40,6 @@
 1.0	:-P
 1.0	:D
 1.0	:P
-1.0	:P
 1.0	:]
 1.0	:o)
 1.0	:p

data/lib/stimmung.rb

@@ -1,25 +1,25 @@
+#<Encoding:UTF-8>
+
 class Stimmung
   attr_accessor :words
 
   def initialize(*dicts)
     @words = Hash.new(0.0)
-    if dicts.empty?
-      load_dictionary(File.join(File.dirname(__FILE__), 'emoticons.txt'), File.join(File.dirname(__FILE__), 'german.txt'))
-    else
-      load_dictionary(*dicts)
-    end
+    directory = File.dirname(__FILE__)
+    dicts = %w(emoticons.txt german.txt).map! {|file| File.join(directory, file)} if dicts.empty?
+    load_dictionary(*dicts)
   end
 
   def score(str)
-    str.to_s.downcase.split(/[\s\!\?\.]+/).reduce(0.0){|s,e| s += @words[e] }
+    str.downcase.gsub(/(\w{3,})(\W)(?=\s|$)/, '\1').split(/\s|\b(?=\W)/).reduce(0.0) {|s,e| s += @words[e]}
   end
 
   def load_dictionary(*files)
     files.each do |file|
-      File.readlines(file).each do |line|
-        score, text = line.chomp.split(/\s/)
-        @words[text] = score.to_f
+      File.foreach(file, :encoding => 'UTF-8') do |line|
+        score, text = line.chomp!.split
+        @words[text.downcase.freeze] = score.to_f
       end
     end
   end
-end
+end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: stimmung
 version: !ruby/object:Gem::Version
-  version: 0.0.2
+  version: 0.0.3
 platform: ruby
 authors:
 - Pachacamac
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2015-07-10 00:00:00.000000000 Z
+date: 2015-07-11 00:00:00.000000000 Z
 dependencies: []
 description: Sentiment analysis for the German language
 email: pachacamac@inboxalias.com