pragmatic_segmenter 0.1.0 → 0.1.1

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 96aa232b769e566aad1de6ace0fa6d8591e3859c
-  data.tar.gz: 2ed0518fb7c236821b0986795e7bc1ce5e19e1a8
+  metadata.gz: a6efcebc92c7cc21017d9acacea4283de1003094
+  data.tar.gz: 0c64c57f9c5e506848295199e9c8195cce89a8d3
 SHA512:
-  metadata.gz: fef9536b2c2dc9896aedcb363f89bc10af37c09d66c034d8814ccb784619a9ea82fa3d20e9e69316dfcef06d7013183d9dcbd564430dc7ce2303a7cfbe32ba87
-  data.tar.gz: 813a64e15c83b9dab5a2fe29c0bebba638fba0663c3debb26442cd42f060fa87e927846309355ec15953ca0a45c90d764fae796533f7d054282bdf8b5bcb06d4
+  metadata.gz: edc26a49545a32ae220caac43979861ebe2f28fb892d857d5c920bb857df38a62bfe161ed268168710b155510c780edf1d625e3f61e7e7bdaeaa54c52a1f87aa
+  data.tar.gz: 970301bbcb8e984e1c9d3c8ed222831d35fa7a0731af9b9916ca593d746dbad9db2922d20af5690be810de1100df021ef799c93c2e03901381681fb99980ee8a

data/README.md

@@ -747,6 +747,9 @@ To test the relative performance of different segmentation tools and libraries I
 **Version 0.1.0**  
 * Add Kommanditgesellschaft Rule  
 
+**Version 0.1.1**  
+* Fix handling of German dates  
+
 ## Contributing
 
 If you find a text that is incorrectly segmented using this gem, please submit an issue.

data/lib/pragmatic_segmenter/languages/deutsch.rb

@@ -32,9 +32,20 @@ module PragmaticSegmenter
         # Rubular: http://rubular.com/r/ityNMwdghj
         NegativeNumberPeriodSpaceRule = Rule.new(/(?<=-[0-9]|-([1-9][0-9]))\.(?=\s)/, '∯')
 
+        DE_MONTHS = ['Januar', 'Februar', 'März', 'April', 'Mai', 'Juni', 'Juli', 'August', 'September', 'Oktober', 'November', 'Dezember']
+
         def replace
           super
           @text.apply(NumberPeriodSpaceRule).apply(NegativeNumberPeriodSpaceRule)
+          replace_period_in_deutsch_dates(@text)
+        end
+
+        def replace_period_in_deutsch_dates(txt)
+          DE_MONTHS.each do |month|
+            # Rubular: http://rubular.com/r/zlqgj7G5dA
+            txt.gsub!(/(?<=\d)\.(?=\s*#{Regexp.escape(month)})/, '∯')
+          end
+          txt
         end
       end
 

data/lib/pragmatic_segmenter/version.rb

@@ -1,3 +1,3 @@
 module PragmaticSegmenter
-  VERSION = "0.1.0"
+  VERSION = "0.1.1"
 end

data/spec/pragmatic_segmenter_spec.rb

@@ -1508,6 +1508,11 @@ RSpec.describe PragmaticSegmenter::Segmenter do
         ps = PragmaticSegmenter::Segmenter.new(text: "s. vorherige Anmerkung.", language: 'de')
         expect(ps.segment).to eq(["s. vorherige Anmerkung."])
       end
+
+      it 'correctly segments text #033' do
+        ps = PragmaticSegmenter::Segmenter.new(text: "Mit Inkrafttreten des Mindestlohngesetzes (MiLoG) zum 01. Januar 2015 werden in Bezug auf den Einsatz von Leistungs.", language: 'de')
+        expect(ps.segment).to eq(["Mit Inkrafttreten des Mindestlohngesetzes (MiLoG) zum 01. Januar 2015 werden in Bezug auf den Einsatz von Leistungs."])
+      end
     end
   end
 

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: pragmatic_segmenter
 version: !ruby/object:Gem::Version
-  version: 0.1.0
+  version: 0.1.1
 platform: ruby
 authors:
 - Kevin S. Dias