Explicitly sort output token in RST visualizer

CHANGELOG.md

@@ -15,6 +15,8 @@ ANNIS is a frontend to the graphANNIS webservice, which has its [own changelog](
 
 - Document names containing an umlaut or other special characters were shown
   percent encoded in the match list ([#837](https://github.com/korpling/ANNIS/issues/837)).
+- RST segments did not sort their token theirself, but relied on external mechanisms to 
+  already get sorted tokens. Changed to explicitly sort the output. (#913)
 
 ## [4.15.3] - 2026-06-25
 

src/main/java/org/corpus_tools/annis/gui/visualizers/component/rst/RSTImpl.java

@@ -22,7 +22,6 @@
 import java.util.HashMap;
 import java.util.HashSet;
 import java.util.Iterator;
-import java.util.LinkedHashSet;
 import java.util.List;
 import java.util.Map;
 import java.util.Set;
@@ -262,6 +261,7 @@ private boolean isSecondarySignalNode(SNode sNode) {
 
   private String transformSaltToJSON(VisualizerInput visInput) {
     graph = visInput.getSResult().getDocumentGraph();
+
     List<SNode> rootSNodes = graph.getRoots();
     List<SNode> rstRoots = new ArrayList<SNode>();
 
@@ -376,7 +376,7 @@ private JSONObject createJsonEntry(SNode currNode) {
     JSONObject data = new JSONObject();
     StringBuilder sb = new StringBuilder();
     // use a hash set so we don't get any duplicate entries
-    LinkedHashSet<SToken> token = new LinkedHashSet<>();
+    HashSet<SToken> tokenSet = new HashSet<>();
     List<SRelation<SNode, SNode>> edges;
 
     if (currNode instanceof SStructure) {
@@ -386,11 +386,12 @@ private JSONObject createJsonEntry(SNode currNode) {
       // get all tokens directly dominated tokens and build a string
       for (SRelation<SNode, SNode> sedge : edges) {
         if (sedge.getTarget() instanceof SToken) {
-          token.add((SToken) sedge.getTarget());
+          tokenSet.add((SToken) sedge.getTarget());
         }
       }
 
-      // build strings
+      // build strings from sorted token
+      List<SToken> token = graph.getSortedTokenByText(new ArrayList<>(tokenSet));
       Iterator<SToken> tokIterator = token.iterator();
       while (tokIterator.hasNext()) {
         SToken tok = tokIterator.next();
@@ -441,7 +442,7 @@ private JSONObject createJsonEntry(SNode currNode) {
 
 
     // since we have found some tokens, it must be a sentence in RST.
-    if (token.size() > 0) {
+    if (!tokenSet.isEmpty()) {
       data.put("sentence", sb.toString());
     }