[uk] uppercase rule adjustment

epfl-lara · Feb 16, 2023 · 9ee85b5 · 9ee85b5
1 parent e98cfd3
commit 9ee85b5
Show file tree

Hide file tree

Showing 4 changed files with 45 additions and 9 deletions.
diff --git a/languagetool-core/src/main/java/org/languagetool/rules/UppercaseSentenceStartRule.java b/languagetool-core/src/main/java/org/languagetool/rules/UppercaseSentenceStartRule.java
@@ -30,7 +30,6 @@
 import org.languagetool.Language;
 import org.languagetool.tokenizers.WordTokenizer;
 import org.languagetool.tools.StringTools;
-import org.languagetool.tools.Tools;
 
 /**
  * Checks that a sentence starts with an uppercase letter.
@@ -91,6 +90,10 @@ public final String getDescription() {
     return messages.getString("desc_uppercase_sentence");
   }
 
+  protected boolean isException(AnalyzedTokenReadings[] tokens, int tokenIdx) {
+    return false;
+  }
+
   @Override
   public RuleMatch[] match(List<AnalyzedSentence> sentences) throws IOException {
     String lastParagraphString = "";
@@ -123,6 +126,9 @@ public RuleMatch[] match(List<AnalyzedSentence> sentences) throws IOException {
         matchTokenPos = 3;
       }
 
+      if( isException(tokens, matchTokenPos) )
+        return toRuleMatchArray(ruleMatches);
+
       String checkToken = firstToken;
       if (thirdToken != null) {
         checkToken = thirdToken;

diff --git a/languagetool-language-modules/uk/src/main/java/org/languagetool/language/Ukrainian.java b/languagetool-language-modules/uk/src/main/java/org/languagetool/language/Ukrainian.java
@@ -141,7 +141,7 @@ public List<Rule> getRelevantRules(ResourceBundle messages, UserConfig userConfi
             Example.fixed("Ми обідали борщем<marker>,</marker> пловом і салатом,— все смачне")),
 
         // TODO: does not handle dot in abbreviations in the middle of the sentence, and also !.., ?..
-        new UppercaseSentenceStartRule(messages, this,
+        new UkrainianUppercaseSentenceStartRule(messages, this,
             Example.wrong("<marker>речення</marker> має починатися з великої."),
             Example.fixed("<marker>Речення</marker> має починатися з великої")),
 

diff --git a/...dules/uk/src/main/java/org/languagetool/rules/uk/UkrainianUppercaseSentenceStartRule.java b/...dules/uk/src/main/java/org/languagetool/rules/uk/UkrainianUppercaseSentenceStartRule.java
@@ -0,0 +1,28 @@
+package org.languagetool.rules.uk;
+
+import java.util.ResourceBundle;
+
+import org.languagetool.AnalyzedTokenReadings;
+import org.languagetool.Language;
+import org.languagetool.rules.CorrectExample;
+import org.languagetool.rules.IncorrectExample;
+import org.languagetool.rules.UppercaseSentenceStartRule;
+
+public class UkrainianUppercaseSentenceStartRule extends UppercaseSentenceStartRule {
+
+  public UkrainianUppercaseSentenceStartRule(ResourceBundle messages, Language language, IncorrectExample incorrectExample, CorrectExample correctExample) {
+    super(messages, language, incorrectExample, correctExample);
+  }
+
+  @Override
+  protected boolean isException(AnalyzedTokenReadings[] tokens, int tokenIdx) {
+    // list, e.g. а) б) в)
+    if( tokenIdx == 1 && tokenIdx < tokens.length-1
+        && tokens[tokenIdx].getCleanToken().matches("[а-яіїєґ]") 
+        && tokens[tokenIdx+1].getToken().equals(")") )
+      return true;
+
+    return false;
+  }
+
+}
diff --git a/...ge-modules/uk/src/test/java/org/languagetool/rules/uk/UppercaseSentenceStartRuleTest.java b/...ge-modules/uk/src/test/java/org/languagetool/rules/uk/UppercaseSentenceStartRuleTest.java
@@ -18,25 +18,24 @@
  */
 package org.languagetool.rules.uk;
 
+import static org.junit.Assert.assertEquals;
+
+import java.io.IOException;
+import java.util.ArrayList;
+
 import org.junit.Test;
 import org.languagetool.JLanguageTool;
 import org.languagetool.TestTools;
 import org.languagetool.language.Ukrainian;
 import org.languagetool.rules.Example;
 import org.languagetool.rules.RuleMatch;
-import org.languagetool.rules.UppercaseSentenceStartRule;
-
-import java.io.IOException;
-import java.util.ArrayList;
-
-import static org.junit.Assert.assertEquals;
 
 public class UppercaseSentenceStartRuleTest {
 
   @Test
   public void testUkrainian() throws IOException {
     Ukrainian ukrainian = new Ukrainian();
-    UppercaseSentenceStartRule rule = new UppercaseSentenceStartRule(TestTools.getEnglishMessages(), ukrainian,
+    UkrainianUppercaseSentenceStartRule rule = new UkrainianUppercaseSentenceStartRule(TestTools.getEnglishMessages(), ukrainian,
         Example.wrong("<marker>речення</marker> має починатися з великої."),
         Example.fixed("<marker>Речення</marker> має починатися з великої"));
 
@@ -56,6 +55,9 @@ public void testUkrainian() throws IOException {
     assertEquals(new ArrayList<RuleMatch>(), lt.check("Цей список з декількох рядків:\n\nрядок 1,\n\nрядок 2,\n\nрядок 3."));
     assertEquals(0, lt.check("Цей список з декількох рядків:\n\nрядок 1;\n\nрядок 2;\n\nрядок 3.").size());
     assertEquals(0, lt.check("Цей список з декількох рядків:\n\n 1) рядок 1;\n\n2) рядок 2;\n\n3)рядок 3.").size());
+
+
+    assertEquals(new ArrayList<RuleMatch>(), lt.check("а) водопі́й"));
   }
 
 }