LUCENE-9753: Hunspell: disallow compounds with parts present in dictionary, space-separated (#2335)

lucene/analysis/common/src/java/org/apache/lucene/analysis/hunspell/SpellChecker.java

@@ -261,12 +261,16 @@ public class SpellChecker {
         return false;
       }
 
-      //noinspection RedundantIfStatement
       if (dictionary.checkCompoundRep
           && isMisspelledSimpleWord(length + nextPartLength, originalCase)) {
         return false;
       }
-      return true;
+
+      String spaceSeparated =
+          new String(tail.chars, tail.offset, length)
+              + " "
+              + new String(tail.chars, tail.offset + length, nextPartLength);
+      return !checkWord(spaceSeparated);
     }
 
     private boolean isMisspelledSimpleWord(int length, WordCase originalCase) {

lucene/analysis/common/src/test/org/apache/lucene/analysis/hunspell/SpellCheckerTest.java

@@ -132,6 +132,10 @@ public class SpellCheckerTest extends StemmerTestBase {
     doTest("checkcompoundrep");
   }
 
+  public void testDisallowCompoundsWhenDictionaryContainsSeparatedWordPair() throws Exception {
+    doTest("wordpair");
+  }
+
   public void testCompoundrule() throws Exception {
     doTest("compoundrule");
   }

lucene/analysis/common/src/test/org/apache/lucene/analysis/hunspell/wordpair.aff

@@ -0,0 +1,4 @@
+# a dictionary word pair separated by space
+# will avoid its recognition without space
+# at compound word analysis
+COMPOUNDFLAG Y

lucene/analysis/common/src/test/org/apache/lucene/analysis/hunspell/wordpair.dic

@@ -0,0 +1,4 @@
+3
+word/Y
+compound/Y
+compound word

lucene/analysis/common/src/test/org/apache/lucene/analysis/hunspell/wordpair.good

@@ -0,0 +1,3 @@
+word
+compound
+wordcompound

lucene/analysis/common/src/test/org/apache/lucene/analysis/hunspell/wordpair.wrong

@@ -0,0 +1 @@
+compoundword