[de] adresses problem described in #779

languagetool-org · Sep 9, 2017 · 7950b29 · 7950b29
1 parent 7e38a72
commit 7950b29
Show file tree

Hide file tree

Showing 4 changed files with 24 additions and 6 deletions.
diff --git a/languagetool-core/src/main/java/org/languagetool/rules/AbstractCompoundRule.java b/languagetool-core/src/main/java/org/languagetool/rules/AbstractCompoundRule.java
@@ -18,6 +18,7 @@
  */
 package org.languagetool.rules;
 
+import org.apache.commons.lang3.StringUtils;
 import org.languagetool.AnalyzedSentence;
 import org.languagetool.AnalyzedToken;
 import org.languagetool.AnalyzedTokenReadings;
@@ -125,7 +126,7 @@ public RuleMatch[] match(AnalyzedSentence sentence) {
             msg = withHyphenMessage;
           }
           if (isNotAllUppercase(origStringToCheck) && !getCompoundRuleData().getOnlyDashSuggestion().contains(stringToCheck)) {
-            replacement.add(mergeCompound(origStringToCheck));
+            replacement.add(mergeCompound(origStringToCheck, getCompoundRuleData().getNoDashIgnoreCaseSuggestion().stream().anyMatch(s -> origStringsToCheck.contains(s))));
             msg = withoutHyphenMessage;
           }
           String[] parts = stringToCheck.split(" ");
@@ -199,15 +200,15 @@ private boolean isNotAllUppercase(String str) {
     return true;
   }
 
-  private String mergeCompound(String str) {
+  private String mergeCompound(String str, boolean uncapitalizeMidWords) {
     String[] stringParts = str.split(" ");
     StringBuilder sb = new StringBuilder();
     for (int k = 0; k < stringParts.length; k++) {
       if (isHyphenIgnored() || !"-".equals(stringParts[k])) {
         if (k == 0) {
           sb.append(stringParts[0]);
         } else {
-          sb.append(stringParts[k]);
+          sb.append(uncapitalizeMidWords ? StringUtils.uncapitalize(stringParts[k]) : stringParts[k]);
         }
       }
     }

diff --git a/languagetool-core/src/main/java/org/languagetool/rules/CompoundRuleData.java b/languagetool-core/src/main/java/org/languagetool/rules/CompoundRuleData.java
@@ -18,8 +18,6 @@
  */
 package org.languagetool.rules;
 
-import org.languagetool.JLanguageTool;
-
 import java.io.BufferedReader;
 import java.io.IOException;
 import java.io.InputStream;
@@ -28,6 +26,8 @@
 import java.util.HashSet;
 import java.util.Set;
 
+import org.languagetool.JLanguageTool;
+
 /**
  * Data about words that are compounds and should thus not be written
  * as separate words.
@@ -37,6 +37,7 @@ public class CompoundRuleData {
 
   private final Set<String> incorrectCompounds = new HashSet<>();
   private final Set<String> noDashSuggestion = new HashSet<>();
+  private final Set<String> noDashIgnoreCaseSuggestion = new HashSet<>();
   private final Set<String> onlyDashSuggestion = new HashSet<>();
 
   public CompoundRuleData(String path) {
@@ -65,6 +66,10 @@ Set<String> getOnlyDashSuggestion() {
     return Collections.unmodifiableSet(onlyDashSuggestion);
   }
 
+  Set<String> getNoDashIgnoreCaseSuggestion() {
+	return Collections.unmodifiableSet(noDashIgnoreCaseSuggestion);
+  }
+
   private void loadCompoundFile(String path) throws IOException {
     try (
       InputStream stream = JLanguageTool.getDataBroker().getFromResourceDirAsStream(path);
@@ -84,6 +89,10 @@ private void loadCompoundFile(String path) throws IOException {
         } else if (line.endsWith("*")) {
           line = removeLastCharacter(line);
           onlyDashSuggestion.add(line);
+        } else if (line.endsWith("?")) { // github issue #779
+          line = removeLastCharacter(line);
+          noDashSuggestion.add(line);
+          noDashIgnoreCaseSuggestion.add(line);
         }
         incorrectCompounds.add(line);
       }

diff --git a/...agetool-language-modules/de/src/main/resources/org/languagetool/resource/de/compounds.txt b/...agetool-language-modules/de/src/main/resources/org/languagetool/resource/de/compounds.txt
@@ -6,6 +6,8 @@
 # even if the word is usually not spelled with a hyphen or blank.
 # "+" at the end of the line will turn off the suggestion that
 # uses a hyphen.
+# "?" at the end of the line will turn off the suggestion that
+# uses a hyphen and offers lower-cased joined suggestions.
 # "*" at the end of the line will only suggest hyphenated words.
 # =========================================================================
 # =========================================================================
@@ -302,7 +304,6 @@ Distanz-säule+
 Domino-effekt+
 Doppel-effekt+
 Doppel-säule+
-Doppler-effekt+
 Dreh-effekt+
 Dreh-säule+
 Drogen-effekt+
@@ -1295,6 +1296,11 @@ Zusammen-spiels+
 Zusatz-effekt+
 zuteil-werden+
 # =========================================================================
+# "?" at end of line = no hyphen suggestion and join force lower case when
+# joining words
+# =========================================================================
+Doppler-Effekt?
+# =========================================================================
 # The following part is based on the lists from http://tkltrans.sf.net
 # This part is LGPL licensed, license text in http://www.fsf.org
 # Owner: Lantol,  email: lantol46_AT_nospam_DOT_yahoo_DOT_de

diff --git a/...agetool-language-modules/de/src/test/java/org/languagetool/rules/de/CompoundRuleTest.java b/...agetool-language-modules/de/src/test/java/org/languagetool/rules/de/CompoundRuleTest.java
@@ -64,6 +64,8 @@ public void testRule() throws IOException {
     check(1, "Ein Start Ziel Sieg");
     check(1, "Start Ziel Sieg");
     check(1, "Start Ziel Sieg!");
+    check(1, "Doppler Effekt");
+    check(1, "Doppler effekt");
     check(2, "Der dumme System Administrator legt die CD ROM");
     check(2, "Der dumme System Administrator legt die CD ROM.");
     check(2, "Der dumme System Administrator legt die CD ROM ein blah");