[pt] Add tests for freaky modifier diacritics

languagetool-language-modules/pt/src/main/resources/org/languagetool/resource/pt/disambiguation.xml

@@ -3886,9 +3886,9 @@
       <disambig action="ignore_spelling"/>
     </rule>
 
-    <rule id="CURRENCT_SPELLING_IGNORE">
+    <rule id="CURRENCY_SPELLING_IGNORE">
         <pattern>
-            <token regexp="yes">\p{Lu}*&currency_symbols;\d+(\.\d+)?(,\d{2})?</token>
+            <token regexp="yes">\p{Lu}*&currency_symbols;(\d+(\.\d+)?(,\d{2})?)?</token>
         </pattern>
         <disambig action="ignore_spelling"/>
     </rule>

languagetool-language-modules/pt/src/test/java/org/languagetool/rules/pt/MorfologikPortugueseSpellerRuleTest.java

@@ -70,13 +70,9 @@ public void testBrazilPortugueseSpelling() throws Exception {
 
     assertNoErrors("Covid-19, COVID-19, covid-19.", lt, rule);
 
-    // 'ja' not corrected to 'já'!
-    assertSingleError("eu ja fiz isso.", lt, rule, new String[]{"já"});
     assertSingleError("eu so", lt, rule, new String[]{"só"});
     assertSingleError("é so", lt, rule, new String[]{"só"});
 
-    // corrected to bizarre 'autoconheci emen'
-    assertSingleErrorAndPos("- Encontre no autoconheciemen", lt, rule, new String[]{"autoconhecimento"}, 14, 29);
     assertSingleErrorAndPos("Sr. Kato nos ensina inglês", lt, rule, new String[]{"Fato"}, 4, 8);
   }
 
@@ -95,6 +91,7 @@ public void testBrazilPortugueseSpellingDoesNotCheckCurrencyValues() throws Exce
     assertNoErrors("R$45,00", br_lt, br_rule);
     assertNoErrors("US$1.000,00", br_lt, br_rule);
     assertNoErrors("€99,99", br_lt, br_rule);
+    assertNoErrors("US$", br_lt, br_rule);
   }
 
   @Test
@@ -117,6 +114,22 @@ public void testBrazilPortugueseSpellingDoesNotCheckXForVezes() throws Exception
     assertNoErrors("5x", br_lt, br_rule);
   }
 
+  @Test
+  public void testBrazilPortugueseSpellingFailsWithModifierDiacritic() throws Exception {
+    assertNoErrors("Não", br_lt, br_rule);  // proper 'ã' char
+    // this is acceptable because LT converts these compound chars to the proper ones
+    assertSingleError("Não", br_lt, br_rule, new String[]{"Não"});  // modifier tilde
+  }
+
+  @Test
+  public void testBrazilPortugueseSpellingMorfologikWeirdness() throws Exception {
+    // 'ja' not corrected to 'já'!
+    assertSingleError("eu ja fiz isso.", br_lt, br_rule, new String[]{"já"});
+    // corrected to bizarre 'autoconheci emen'
+    assertSingleErrorAndPos("- Encontre no autoconheciemen", br_lt, br_rule,
+      new String[]{"autoconhecimento"}, 14, 29);
+  }
+
   @Test
   public void testEuropeanPortugueseSpelling() throws Exception {
     MorfologikPortugueseSpellerRule rule = new MorfologikPortugueseSpellerRule(TestTools.getMessages("pt"),

languagetool-language-modules/pt/src/test/java/org/languagetool/tokenizers/pt/PortugueseWordTokenizerTest.java

@@ -126,4 +126,10 @@ public void testTokeniseNumberAbbreviation() {
   public void testDoNotTokeniseEmoji() {
     testTokenise("☺☺☺Só", new String[]{"☺☺☺Só"});
   }
+
+  @Test
+  public void testDoNotTokeniseModifierDiacritics() {
+    // the tilde here is a unicode modifier char; normally, the unicode a-tilde (ã) is used
+    testTokenise("Não", new String[]{"Não"});
+  }
 }