confidential_info_redactor_lite 1.0.2 → 1.0.3

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 5944b7a68fd13e6993315805c0cfdccb59e4fd9d
-  data.tar.gz: bb7a44c64076277287649e006002b7cf396e2043
+  metadata.gz: 39a1611275052f427c8e1c916cb368b0a7337347
+  data.tar.gz: 528d58ac47b37eb1be8b76f5de0b5ae8518ddb72
 SHA512:
-  metadata.gz: 0d060ed3456c2c8eaeebf8b486d70a9a15d4f35e9c7b8102ebf72a14ed4e3486cdfc0c40a151733fb6054e052145717806d8e391747e1a95dce425690440b8df
-  data.tar.gz: 7d0729c5404c6425f720ef0ba3bd7fe602b35e2e8d94df70be6c27aaf2a93f84f9afc29450ecdfea9d721a9f7b7b152063d0f09fa9e41b818031b6d0185c5cf2
+  metadata.gz: 39537bfadcce276750a5e3e1c53a275c125ddd457672063e880832e9e13138054723315bbc6a7d4ae041268009acf27ca1e54168f9b8d384ca31ddba4f34725a
+  data.tar.gz: b18afda080a5f8cc75235d202897576f2e76232b3d0c8903f08063e6aaff59c4149ecb7c62ade7944ef4eeff805c271d5c3bed8fe2f57c907edbae053e409a71

data/lib/confidential_info_redactor_lite/extractor.rb

@@ -18,7 +18,7 @@ module ConfidentialInfoRedactorLite
         next if initial_extracted_terms.length.eql?(segment.split(' ').length) && !in_corpus?(initial_extracted_terms)
         search_ngrams(initial_extracted_terms, extracted_terms)
       end
-      extracted_terms.map { |t| t.gsub(/\{\}/, '') }.uniq.reject(&:empty?)
+      extracted_terms.map { |t| t.gsub(/\{\}/, '') }.delete_if { |t| t.length == 1 }.uniq.reject(&:empty?)
     end
 
     private

data/lib/confidential_info_redactor_lite/version.rb

@@ -1,3 +1,3 @@
 module ConfidentialInfoRedactorLite
-  VERSION = "1.0.2"
+  VERSION = "1.0.3"
 end

data/spec/confidential_info_redactor_lite/extractor_spec.rb

@@ -154,6 +154,11 @@ RSpec.describe ConfidentialInfoRedactorLite::Extractor do
         text = 'Westin{}'
         expect(described_class.new(corpus: corpus, language: 'en').extract(text)).to eq(["Westin"])
       end
+
+      it 'extracts the proper nouns from a text #020' do
+        text = 'Document No:-'
+        expect(described_class.new(corpus: corpus, language: 'en').extract(text)).to eq(["Document No"])
+      end
     end
 
     context 'German (de)' do

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: confidential_info_redactor_lite
 version: !ruby/object:Gem::Version
-  version: 1.0.2
+  version: 1.0.3
 platform: ruby
 authors:
 - Kevin S. Dias