From 96d7b7abe732c430ea37afb0e9d3cdf8db10eb23 Mon Sep 17 00:00:00 2001
From: Nicola Isotta <nicola.isotta@gmail.com>
Date: Tue, 4 Aug 2026 08:54:52 +0200
Subject: [PATCH] XMLSyntaxParser: tolerate a UTF-8 BOM at the start of XML
 documents

XMLSyntaxParser rejected any leading text before the document
element with "Invalid token ... found in document", so a BOM
character (U+FEFF) caused every BOM-prefixed XML file to fail
parsing. Treat a trimmed BOM the same as blank/whitespace when
validating text under the Document node.

This also adds a test with a BOM xml file and removes some dead code
(tokenSequence.token() is always null before the first moveNext())
---
 .../xml/xdm/nodes/XMLSyntaxParser.java        | 22 +++++--------------
 .../xml/xdm/nodes/XMLSyntaxParserTest.java    | 18 ++++++++++++++-
 .../modules/xml/xdm/nodes/testBOM.xml         |  9 ++++++++
 3 files changed, 31 insertions(+), 18 deletions(-)
 create mode 100644 ide/xml.xdm/test/unit/src/org/netbeans/modules/xml/xdm/nodes/testBOM.xml

diff --git a/ide/xml.xdm/src/org/netbeans/modules/xml/xdm/nodes/XMLSyntaxParser.java b/ide/xml.xdm/src/org/netbeans/modules/xml/xdm/nodes/XMLSyntaxParser.java
index 7627ccc8debb..cba5d81d0037 100644
--- a/ide/xml.xdm/src/org/netbeans/modules/xml/xdm/nodes/XMLSyntaxParser.java
+++ b/ide/xml.xdm/src/org/netbeans/modules/xml/xdm/nodes/XMLSyntaxParser.java
@@ -31,7 +31,9 @@
 import org.netbeans.api.xml.lexer.XMLTokenId;
 
 public class XMLSyntaxParser {
-            
+
+    private static final String BOM = "\uFEFF";
+
     public Document parse(BaseDocument basedoc)
     throws IOException, BadLocationException {
         try {
@@ -44,22 +46,8 @@ public Document parse(BaseDocument basedoc)
             List<Token> currentTokens = new ArrayList<Token>();
             TokenHierarchy th = TokenHierarchy.get(basedoc);
             TokenSequence<XMLTokenId> tokenSequence = th.tokenSequence();
-            org.netbeans.api.lexer.Token<XMLTokenId> token = tokenSequence.token();
-            // Add the text token, if any, before xml decalration to document node
-            if(token != null && token.id() == XMLTokenId.TEXT) {
-                currentTokens.add(Token.create(token.text().toString(),TokenType.TOKEN_CHARACTER_DATA));
-                if(tokenSequence.moveNext()) {
-                    token = tokenSequence.token();
-                }
-                // if the xml decalration is not there assign this token to document
-                if(token.id() != XMLTokenId.PI_START) {
-                    currentNode.setTokens(new ArrayList<Token>(currentTokens));
-                    currentTokens.clear();
-                }
-            }
-
             while (tokenSequence.moveNext()) {
-                token = tokenSequence.token();
+                org.netbeans.api.lexer.Token<XMLTokenId> token = tokenSequence.token();
                 XMLTokenId tokenId = token.id();
                 String image = token.text().toString();
                 TokenType tokenType = TokenType.TOKEN_WHITESPACE;
@@ -245,7 +233,7 @@ public Document parse(BaseDocument basedoc)
                             ((Element)parent).appendChild(currentNode, false);
                         } else {//parent is Document
                             if(token.id() != XMLTokenId.BLOCK_COMMENT &&
-                                    token.text().toString().trim().length() > 0) {
+                                    !token.text().toString().isBlank() && !BOM.equals(token.text().toString().trim())) {
                                 throw new IOException("Invalid token '" + token.text() +
                                         "' found in document: " +
                                         "Please use the text editor to resolve the issues...");
diff --git a/ide/xml.xdm/test/unit/src/org/netbeans/modules/xml/xdm/nodes/XMLSyntaxParserTest.java b/ide/xml.xdm/test/unit/src/org/netbeans/modules/xml/xdm/nodes/XMLSyntaxParserTest.java
index fff1190ec0b3..c6401fb139c3 100644
--- a/ide/xml.xdm/test/unit/src/org/netbeans/modules/xml/xdm/nodes/XMLSyntaxParserTest.java
+++ b/ide/xml.xdm/test/unit/src/org/netbeans/modules/xml/xdm/nodes/XMLSyntaxParserTest.java
@@ -64,6 +64,7 @@ public static Test suite() {
         suite.addTest(new XMLSyntaxParserTest("testParseWSDL"));
 //        Disabled as referenced files were partly not donated by oracle to apache
 //        suite.addTest(new XMLSyntaxParserTest("testParsePerformace"));
+        suite.addTest(new XMLSyntaxParserTest("testParseBOM"));
         return suite;
     }
     
@@ -254,5 +255,20 @@ public void testParsePerformace() throws Exception {
         //FlushVisitor fv = new FlushVisitor();
         //String docBuf = fv.flushModel(doc);
         //assertEquals("The document should be unaltered",basedoc.getText(0,basedoc.getLength()),docBuf);
-    }    
+    }
+
+    /**
+     * Test of parse method, of class org.netbeans.modules.xmltools.xmlmodel.nodes.XMLSyntaxParser.
+     * XMLSyntaxParser should handle xml files with BOM
+     */
+    public void testParseBOM() throws Exception {
+        BaseDocument basedoc = getDocument("nodes/testBOM.xml");
+        XMLSyntaxParser parser = new XMLSyntaxParser();
+        Document doc = parser.parse(basedoc);
+        assertNotNull("Document can not be null", doc);
+        FlushVisitor fv = new FlushVisitor();
+        String docBuf = fv.flushModel(doc);
+        assertEquals("The document should be unaltered",basedoc.getText(0,basedoc.getLength()),docBuf);
+    }
+
 }
diff --git a/ide/xml.xdm/test/unit/src/org/netbeans/modules/xml/xdm/nodes/testBOM.xml b/ide/xml.xdm/test/unit/src/org/netbeans/modules/xml/xdm/nodes/testBOM.xml
new file mode 100644
index 000000000000..83943ceda3ff
--- /dev/null
+++ b/ide/xml.xdm/test/unit/src/org/netbeans/modules/xml/xdm/nodes/testBOM.xml
@@ -0,0 +1,9 @@
+<?xml version="1.0"?>
+<company>
+    <employee ssn="xx-xx-xxxx" id="123"
+        address="16 Network Circle"
+        phone="123-456-7890" >Vidhya Narayanan
+    </employee>
+    <!-- -->    
+    <employee/>
+</company>