adds case insensitivity to the language filtering (#1432)

app/services/language_filter.rb

@@ -10,6 +10,7 @@ class LanguageFilter
   def initialize(filter_file_path = Rails.root.join('config', 'language_filter.yml'))
     @filter_data = YAML.load_file(filter_file_path)
     @terms = @filter_data.keys.sort { |a, b| b.length <=> a.length }
+    @term_regexes = @terms.map { |term| [Regexp.new(Regexp.escape(term), Regexp::IGNORECASE), @filter_data[term]['replacement']] }.to_h
   end
 
   # Checks if the input is valid, i.e., doesn't need replacement.
@@ -19,11 +20,7 @@ def initialize(filter_file_path = Rails.root.join('config', 'language_filter.yml
   def valid?(input)
     return true if input.blank?
 
-    @terms.each do |term|
-      return false if input.include?(term)
-    end
-
-    true
+    @term_regexes.keys.none? { |regex| input.match?(regex) }
   end
 
   # Gets the filtered version of the input text.
@@ -35,8 +32,8 @@ def filter(input)
 
     output = input.dup
 
-    @terms.each do |term|
-      output.gsub!(term, @filter_data[term]['replacement']) if output.include?(term)
+    @term_regexes.each do |regex, replacement|
+      output.gsub!(regex, replacement)
     end
 
     output

config/language_filter.yml

@@ -1,11 +1,11 @@
-"Gender identity disorder":
+"gender identity disorder":
   replacement: "Gender dysphoria"
   rationale: "https://docs.google.com/spreadsheets/d/1uqiP5PPKXJt35uFLrG5ytsLbz2kgWXBjG__hyD12KlY/edit?gid=0#gid=0"
-"Gender identity disorders":
+"gender identity disorders":
   replacement: "Gender dysphoria"
-"Gender identity disorders in children":
+"gender identity disorders in children":
   replacement: "Gender dysphoria in children"
-"Gender identity disorders in adolescence":
+"gender identity disorders in adolescence":
   replacement: "Gender dysphoria in adolescence"
 "african american gays in literature":
   replacement: "African American gay people in literature"

spec/services/language_filter_spec.rb

@@ -22,19 +22,21 @@
 
     it 'returns true when input is valid' do
       input = "Georgia"
-
       expect(filter.valid?(input)).to eq(true)
     end
 
     it 'returns false when input is invalid' do
       input = "Gender identity disorder"
+      expect(filter.valid?(input)).to eq(false)
+    end
 
+    it 'returns false when input is invalid with different case' do
+      input = "gEnDeR iDeNtItY dIsOrDeR"
       expect(filter.valid?(input)).to eq(false)
     end
 
     it 'returns true when input is nil' do
       input = nil
-
       expect(filter.valid?(input)).to eq(true)
     end
   end
@@ -59,7 +61,12 @@
     it 'prioritizes terms with higher length during replacement' do
       input = 'Gender identity disorders'
       expected = 'Gender dysphoria'
+      expect(filter.filter(input)).to eq(expected)
+    end
 
+    it 'replaces harmful text with corresponding replacements regardless of case' do
+      input = 'gEnDeR iDeNtItY dIsOrDeR'
+      expected = 'Gender dysphoria'
       expect(filter.filter(input)).to eq(expected)
     end
 
@@ -70,6 +77,12 @@
       expect(filter.filter(input)).to eq(expected)
     end
 
+    it 'replaces harmful terms when multiple terms are combined regardless of case' do
+      input = 'GeNdEr IdEnTiTy DiSoRdErS--United States'
+      expected = 'Gender dysphoria--United States'
+      expect(filter.filter(input)).to eq(expected)
+    end
+
     it 'returns nil when input is nil' do
       input = nil
       expected = nil