Switch to iterative version of WKT format parser

Signed-off-by: Heemin Kim <heemin@amazon.com>
opensearch-project · Jun 8, 2024 · 1ef63ad · 1ef63ad
1 parent fbe048f
commit 1ef63ad
Show file tree

Hide file tree

Showing 2 changed files with 86 additions and 5 deletions.
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -65,5 +65,6 @@ The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
 - Fix double invocation of postCollection when MultiBucketCollector is present ([#14015](https://github.com/opensearch-project/OpenSearch/pull/14015))
 
 ### Security
+- Switch to iterative version of WKT format parser ([#14086](https://github.com/opensearch-project/OpenSearch/pull/14086))
 
 [Unreleased 2.x]: https://github.com/opensearch-project/OpenSearch/compare/2.13...2.x
diff --git a/libs/geo/src/main/java/org/opensearch/geometry/utils/WellKnownText.java b/libs/geo/src/main/java/org/opensearch/geometry/utils/WellKnownText.java
@@ -53,6 +53,7 @@
 import java.util.Collections;
 import java.util.List;
 import java.util.Locale;
+import java.util.Stack;
 
 /**
  * Utility class for converting to and from WKT
@@ -301,16 +302,95 @@ private Geometry parseGeometry(StreamTokenizer stream) throws IOException, Parse
         throw new IllegalArgumentException("Unknown geometry type: " + type);
     }
 
+    /**
+     * Iterative version of
+     * <!--
+     * ```java
+     *     private GeometryCollection<Geometry> parseGeometryCollectionA(StreamTokenizer stream) throws IOException, ParseException {
+     *         if (nextEmptyOrOpen(stream).equals(EMPTY)) {
+     *             return GeometryCollection.EMPTY;
+     *         }
+     *         List<Geometry> shapes = new ArrayList<>();
+     *         shapes.add(parseGeometry(stream));
+     *         while (nextCloserOrComma(stream).equals(COMMA)) {
+     *             shapes.add(parseGeometry(stream));
+     *         }
+     *         return new GeometryCollection<>(shapes);
+     *     }
+     * -->
+     * to avoid StackOverflowError when there is a deeply nested structure of GeometryCollection.
+     */
     private GeometryCollection<Geometry> parseGeometryCollection(StreamTokenizer stream) throws IOException, ParseException {
         if (nextEmptyOrOpen(stream).equals(EMPTY)) {
             return GeometryCollection.EMPTY;
         }
-        List<Geometry> shapes = new ArrayList<>();
-        shapes.add(parseGeometry(stream));
-        while (nextCloserOrComma(stream).equals(COMMA)) {
-            shapes.add(parseGeometry(stream));
+
+        List<Geometry> topLevelShapes = new ArrayList<>();
+        Stack<List<Geometry>> stack = new Stack<>();
+        stack.push(topLevelShapes);
+        boolean isFirstIteration = true;
+        List<Geometry> currentLevelShapes = null;
+        while (!stack.isEmpty()) {
+            List<Geometry> previousShapes = stack.pop();
+            if (currentLevelShapes != null) {
+                previousShapes.add(new GeometryCollection<>(currentLevelShapes));
+            }
+            currentLevelShapes = previousShapes;
+
+            if (isFirstIteration == true) {
+                isFirstIteration = false;
+            } else {
+                if (!nextCloserOrComma(stream).equals(COMMA)) {
+                    continue;
+                }
+            }
+            while (true) {
+                final String type = nextWord(stream).toLowerCase(Locale.ROOT);
+                switch (type) {
+                    case "point":
+                        currentLevelShapes.add(parsePoint(stream));
+                        break;
+                    case "multipoint":
+                        currentLevelShapes.add(parseMultiPoint(stream));
+                        break;
+                    case "linestring":
+                        currentLevelShapes.add(parseLine(stream));
+                        break;
+                    case "multilinestring":
+                        currentLevelShapes.add(parseMultiLine(stream));
+                        break;
+                    case "polygon":
+                        currentLevelShapes.add(parsePolygon(stream));
+                        break;
+                    case "multipolygon":
+                        currentLevelShapes.add(parseMultiPolygon(stream));
+                        break;
+                    case "bbox":
+                        currentLevelShapes.add(parseBBox(stream));
+                        break;
+                    case "geometrycollection":
+                        if (nextEmptyOrOpen(stream).equals(EMPTY)) {
+                            currentLevelShapes.add(GeometryCollection.EMPTY);
+                            break;
+                        } else {
+                            stack.push(currentLevelShapes);
+                            currentLevelShapes = new ArrayList<>();
+                            continue;
+                        }
+                    case "circle": // Not part of the standard, but we need it for internal serialization
+                        currentLevelShapes.add(parseCircle(stream));
+                        break;
+                    default:
+                        throw new IllegalArgumentException("Unknown geometry type: " + type);
+                }
+
+                if (!nextCloserOrComma(stream).equals(COMMA)) {
+                    break;
+                }
+            }
         }
-        return new GeometryCollection<>(shapes);
+
+        return new GeometryCollection<>(topLevelShapes);
     }
 
     private Point parsePoint(StreamTokenizer stream) throws IOException, ParseException {