Call onNodeClosed when closing header tags

Ensures that the end source range for headers is tracked correctly. Fixes #1987
jhy · Sep 8, 2023 · 0eb8232 · 0eb8232
1 parent 04b3a82
commit 0eb8232
Show file tree

Hide file tree

Showing 3 changed files with 23 additions and 12 deletions.
diff --git a/CHANGES b/CHANGES
@@ -37,6 +37,9 @@ Release 1.16.2 [PENDING]
     thread after interrupting it.
     <https://github.com/jhy/jsoup/issues/1991>
 
+  * Bugfix: when tracking HTML source positions, the closing tags for H1...H6 elements were not tracked correctly.
+    <https://github.com/jhy/jsoup/issues/1987>
+
   * Change: removed previously deprecated methods Document#normalise, Element#forEach(org.jsoup.helper.Consumer<>),
     Node#forEach(org.jsoup.helper.Consumer<>), and the org.jsoup.helper.Consumer interface; the latter being a
     previously required compatibility shim prior to Android's de-sugaring support.

diff --git a/src/main/java/org/jsoup/parser/HtmlTreeBuilder.java b/src/main/java/org/jsoup/parser/HtmlTreeBuilder.java
@@ -424,20 +424,12 @@ Element popStackToClose(String elName) {
     // elnames is sorted, comes from Constants
     void popStackToClose(String... elNames) {
         for (int pos = stack.size() -1; pos >= 0; pos--) {
-            Element next = stack.get(pos);
+            Element el = stack.get(pos);
             stack.remove(pos);
-            if (inSorted(next.normalName(), elNames))
-                break;
-        }
-    }
-
-    void popStackToBefore(String elName) {
-        for (int pos = stack.size() -1; pos >= 0; pos--) {
-            Element next = stack.get(pos);
-            if (next.normalName().equals(elName)) {
+            if (inSorted(el.normalName(), elNames)) {
+                if (currentToken instanceof Token.EndTag)
+                    onNodeClosed(el, currentToken);
                 break;
-            } else {
-                stack.remove(pos);
             }
         }
     }

diff --git a/src/test/java/org/jsoup/nodes/PositionTest.java b/src/test/java/org/jsoup/nodes/PositionTest.java
@@ -203,4 +203,20 @@ class PositionTest {
         }
     }
 
+    @Test void tracksClosingHeadingTags() {
+        // https://github.com/jhy/jsoup/issues/1987
+        String html = "<h1>One</h1><h2>Two</h2><h10>Ten</h10>";
+        Document doc = Jsoup.parse(html, TrackingParser);
+
+        Elements els = doc.body().children();
+        for (Element el : els) {
+            assertTrue(el.sourceRange().isTracked());
+            assertTrue(el.endSourceRange().isTracked());
+        }
+
+        Element h2 = doc.expectFirst("h2");
+        assertEquals("1,13:12-1,17:16", h2.sourceRange().toString());
+        assertEquals("1,20:19-1,25:24", h2.endSourceRange().toString());
+    }
+
 }