asXml(): emit formatting line breaks only where the document contains whitespace (#1158)

shapiroronny · rbri · web-flow · commit 05ab2b1f4323 · 2026-07-14T17:24:18.000+02:00
Adjacent elements without whitespace in between no longer gain a fabricated
line break, which rendered as a visible space between inline elements after a
parse round-trip. Whitespace-only text nodes now mark where the formatting
break belongs instead of being silently dropped, so the output follows the
whitespace actually present in the document.

Co-authored-by: RBRi &lt;rbri@rbri.de&gt;
diff --git a/src/changes/changes.xml b/src/changes/changes.xml
@@ -8,6 +8,10 @@
 
     <body>
         <release version="5.3.0" date="July xx, 2026" description="Chrome/Edge 150, Bugfixes">
+            <action type="fix" dev="Ronny Shapiro">
+                asXml(): emit formatting line breaks only where the document contains whitespace;
+                adjacent elements no longer gain a fabricated line break that renders as a space.
+            </action>
             <action type="update" dev="rbri">
                 Many, many javadoc improvements.
             </action>
diff --git a/src/main/java/org/htmlunit/html/DomElement.java b/src/main/java/org/htmlunit/html/DomElement.java
@@ -362,7 +362,7 @@ protected boolean printXml(final String indent, final boolean tagBefore, final P
 
         if (hasChildren) {
             printWriter.print(">");
-            final boolean tag = printChildrenAsXml(indent, true, printWriter);
+            final boolean tag = printChildrenAsXml(indent, false, printWriter);
             if (tag) {
                 printWriter.print("\r\n");
                 printWriter.print(indent);
@@ -380,7 +380,7 @@ else if (isEmptyXmlTagExpanded()) {
             printWriter.print("/>");
         }
 
-        return true;
+        return false;
     }
 
     /**
diff --git a/src/main/java/org/htmlunit/html/DomText.java b/src/main/java/org/htmlunit/html/DomText.java
@@ -141,6 +141,9 @@ protected boolean printXml(final String indent, final boolean tagBefore, final P
             printWriter.print(data);
             tag = false;
         }
+        else if (data != null && !data.isEmpty()) {
+            tag = true;
+        }
         return printChildrenAsXml(indent, tag, printWriter);
     }
 
diff --git a/src/test/java/org/htmlunit/WebClient8Test.java b/src/test/java/org/htmlunit/WebClient8Test.java
@@ -59,15 +59,45 @@ public void asXml() throws Exception {
 
         try (WebClient webClient = new WebClient(getBrowserVersion(), false, null, -1)) {
             final HtmlPage page = loadPage(webClient, html, null, URL_FIRST);
-            assertEquals("<?xml version=\"1.0\" encoding=\"ISO-8859-1\"?>\r\n<html>\r\n"
-                    + "  <head>\r\n"
-                    + "    <title>foo</title>\r\n"
-                    + "  </head>\r\n"
-                    + "  <body>\r\n"
-                    + "    <div>Hello <b>HtmlUnit</b>\r\n"
-                    + "    </div>\r\n"
-                    + "  </body>\r\n"
-                    + "</html>",
+            assertEquals("<?xml version=\"1.0\" encoding=\"ISO-8859-1\"?>\r\n"
+                    + "<html><head><title>foo</title></head>"
+                    + "<body><div>Hello <b>HtmlUnit</b></div></body></html>",
+                    page.asXml());
+        }
+    }
+
+    /**
+     * @throws Exception if something goes wrong
+     */
+    @Test
+    public void asXmlNoWhitespaceBetweenAdjacentElements() throws Exception {
+        final String html = DOCTYPE_HTML
+                + "<html><head><title>foo</title></head>"
+                + "<body><div><span>a</span><span>b</span></div></body></html>";
+
+        try (WebClient webClient = new WebClient(getBrowserVersion(), false, null, -1)) {
+            final HtmlPage page = loadPage(webClient, html, null, URL_FIRST);
+            assertEquals("<?xml version=\"1.0\" encoding=\"ISO-8859-1\"?>\r\n"
+                    + "<html><head><title>foo</title></head>"
+                    + "<body><div><span>a</span><span>b</span></div></body></html>",
+                    page.asXml());
+        }
+    }
+
+    /**
+     * @throws Exception if something goes wrong
+     */
+    @Test
+    public void asXmlWhitespaceOnlyTextBetweenElements() throws Exception {
+        final String html = DOCTYPE_HTML
+                + "<html><head><title>foo</title></head>"
+                + "<body><div><span>a</span> <span>b</span></div></body></html>";
+
+        try (WebClient webClient = new WebClient(getBrowserVersion(), false, null, -1)) {
+            final HtmlPage page = loadPage(webClient, html, null, URL_FIRST);
+            assertEquals("<?xml version=\"1.0\" encoding=\"ISO-8859-1\"?>\r\n"
+                    + "<html><head><title>foo</title></head>"
+                    + "<body><div><span>a</span>\r\n      <span>b</span></div></body></html>",
                     page.asXml());
         }
     }
diff --git a/src/test/java/org/htmlunit/html/HtmlFormTest.java b/src/test/java/org/htmlunit/html/HtmlFormTest.java
@@ -1199,13 +1199,8 @@ public void asXml_emptyTag() throws Exception {
 
         final String xml =
                 "<?xml version=\"1.0\" encoding=\"ISO-8859-1\"?>\r\n"
-                        + "<html>\r\n"
-                        + "  <head/>\r\n"
-                        + "  <body>\r\n"
-                        + "    <form></form>\r\n"
-                        + "    <div>test</div>\r\n"
-                        + "  </body>\r\n"
-                        + "</html>";
+                        + "<html><head/><body>\r\n    <form></form>\r\n    <div>test</div>\r\n"
+                        + "  </body></html>";
 
         final HtmlPage page = loadPage(html);
         assertEquals(xml, page.asXml());
diff --git a/src/test/java/org/htmlunit/html/HtmlNoScript2Test.java b/src/test/java/org/htmlunit/html/HtmlNoScript2Test.java
@@ -34,8 +34,7 @@ public class HtmlNoScript2Test extends SimpleWebTestCase {
      */
     @Test
     @Alerts("<body>\r\n"
-                + "  <noscript>&lt;div&gt;hello</noscript>\r\n"
-                + "</body>")
+                + "  <noscript>&lt;div&gt;hello</noscript></body>")
     public void asXml_jsEnabled() throws Exception {
         final String html = DOCTYPE_HTML
             + "<html><body>\n"
@@ -57,10 +56,7 @@ public void asXml_jsDisabled() throws Exception {
             + "</body></html>";
 
         final String expected = "<body>\r\n"
-            + "  <noscript>\r\n"
-            + "    <div>hello</div>\r\n"
-            + "  </noscript>\r\n"
-            + "</body>";
+            + "  <noscript><div>hello</div></noscript></body>";
 
         final WebClient client = getWebClient();
         client.getOptions().setJavaScriptEnabled(false);
diff --git a/src/test/java/org/htmlunit/html/HtmlOrderedListTest.java b/src/test/java/org/htmlunit/html/HtmlOrderedListTest.java
@@ -70,12 +70,8 @@ public void asXml() throws Exception {
 
         // assertEquals("<ol id=\"myNode\"></ol>", element.asXml());
         assertEquals("<?xml version=\"1.0\" encoding=\"ISO-8859-1\"?>\r\n"
-                + "<html>\r\n"
-                + "  <head/>\r\n"
-                + "  <body>\r\n"
-                + "    <ol id=\"myNode\"></ol>\n"
+                + "<html><head/><body>\r\n    <ol id=\"myNode\"></ol>\n"
                 + "foo\n"
-                + "</body>\r\n"
-                + "</html>", page.asXml());
+                + "</body></html>", page.asXml());
     }
 }
diff --git a/src/test/java/org/htmlunit/html/HtmlPageTest.java b/src/test/java/org/htmlunit/html/HtmlPageTest.java
@@ -1042,14 +1042,9 @@ public void asXml_noscript() throws Exception {
             + "</body></html>";
 
         final String expected = "<?xml version=\"1.0\" encoding=\"ISO-8859-1\"?>\r\n"
-                + "<html>\r\n"
-                + "  <head/>\r\n"
-                + "  <body>\r\n"
-                + "    <noscript>"
-                            + "&lt;p&gt;&lt;strong&gt;your browser does not support JavaScript&lt;/strong&gt;&lt;/p&gt;"
-                            + "</noscript>\r\n"
-                + "  </body>\r\n"
-                + "</html>";
+                + "<html><head/><body><noscript>"
+                + "&lt;p&gt;&lt;strong&gt;your browser does not support JavaScript&lt;/strong&gt;&lt;/p&gt;"
+                + "</noscript></body></html>";
 
         final HtmlPage page = loadPage(html);
         assertEquals(expected, page.asXml());
diff --git a/src/test/java/org/htmlunit/html/HtmlScriptTest.java b/src/test/java/org/htmlunit/html/HtmlScriptTest.java
@@ -110,20 +110,13 @@ public void asXml() throws Exception {
         // asXml() should be reusable
         final String xml = page.asXml();
         assertEquals("<?xml version=\"1.0\" encoding=\"ISO-8859-1\"?>\r\n"
-                + "<html>\r\n"
-                + "  <head>\r\n"
-                + "    <title>foo</title>\r\n"
-                + "  </head>\r\n"
-                + "  <body>\r\n"
-                + "    <script id=\"script1\">\r\n"
+                + "<html><head><title>foo</title></head><body>\r\n    <script id=\"script1\">\r\n"
                 + "//<![CDATA[\r\n"
                 + "\n"
                 + "  alert('hello');\n"
                 + "\r\n"
                 + "//]]>\r\n"
-                + "    </script>\r\n"
-                + "  </body>\r\n"
-                + "</html>",
+                + "    </script></body></html>",
                 xml);
 
         loadPageWithAlerts(xml);
diff --git a/src/test/java/org/htmlunit/html/HtmlUnorderedList2Test.java b/src/test/java/org/htmlunit/html/HtmlUnorderedList2Test.java
@@ -68,12 +68,8 @@ public void asXml() throws Exception {
 
         assertEquals("<ul id=\"myNode\"></ul>", element.asXml());
         assertEquals("<?xml version=\"1.0\" encoding=\"ISO-8859-1\"?>\r\n"
-                + "<html>\r\n"
-                + "  <head/>\r\n"
-                + "  <body>\r\n"
-                + "    <ul id=\"myNode\"></ul>\n"
+                + "<html><head/><body>\r\n    <ul id=\"myNode\"></ul>\n"
                 + "foo\n"
-                + "</body>\r\n"
-                + "</html>", page.asXml());
+                + "</body></html>", page.asXml());
     }
 }

Original file line number	Diff line number	Diff line change
`@@ -362,7 +362,7 @@ protected boolean printXml(final String indent, final boolean tagBefore, final P`
`362`	`362`
`363`	`363`	`if (hasChildren) {`
`364`	`364`	`printWriter.print(">");`
`365`		`- final boolean tag = printChildrenAsXml(indent, true, printWriter);`
	`365`	`+ final boolean tag = printChildrenAsXml(indent, false, printWriter);`
`366`	`366`	`if (tag) {`
`367`	`367`	`printWriter.print("\r\n");`
`368`	`368`	`printWriter.print(indent);`
`@@ -380,7 +380,7 @@ else if (isEmptyXmlTagExpanded()) {`
`380`	`380`	`printWriter.print("/>");`
`381`	`381`	`}`
`382`	`382`
`383`		`- return true;`
	`383`	`+ return false;`
`384`	`384`	`}`
`385`	`385`
`386`	`386`	`/**`
Original file line number	Diff line number	Diff line change
`@@ -141,6 +141,9 @@ protected boolean printXml(final String indent, final boolean tagBefore, final P`
`141`	`141`	`printWriter.print(data);`
`142`	`142`	`tag = false;`
`143`	`143`	`}`
	`144`	`+ else if (data != null && !data.isEmpty()) {`
	`145`	`+ tag = true;`
	`146`	`+ }`
`144`	`147`	`return printChildrenAsXml(indent, tag, printWriter);`
`145`	`148`	`}`
`146`	`149`
Original file line number	Diff line number	Diff line change
`@@ -70,12 +70,8 @@ public void asXml() throws Exception {`
`70`	`70`
`71`	`71`	`// assertEquals("<ol id=\"myNode\"></ol>", element.asXml());`
`72`	`72`	`assertEquals("<?xml version=\"1.0\" encoding=\"ISO-8859-1\"?>\r\n"`
`73`		`- + "<html>\r\n"`
`74`		`- + " <head/>\r\n"`
`75`		`- + " <body>\r\n"`
`76`		`- + " <ol id=\"myNode\"></ol>\n"`
	`73`	`+ + "<html><head/><body>\r\n <ol id=\"myNode\"></ol>\n"`
`77`	`74`	`+ "foo\n"`
`78`		`- + "</body>\r\n"`
`79`		`- + "</html>", page.asXml());`
	`75`	`+ + "</body></html>", page.asXml());`
`80`	`76`	`}`
`81`	`77`	`}`
Original file line number	Diff line number	Diff line change
`@@ -68,12 +68,8 @@ public void asXml() throws Exception {`
`68`	`68`
`69`	`69`	`assertEquals("<ul id=\"myNode\"></ul>", element.asXml());`
`70`	`70`	`assertEquals("<?xml version=\"1.0\" encoding=\"ISO-8859-1\"?>\r\n"`
`71`		`- + "<html>\r\n"`
`72`		`- + " <head/>\r\n"`
`73`		`- + " <body>\r\n"`
`74`		`- + " <ul id=\"myNode\"></ul>\n"`
	`71`	`+ + "<html><head/><body>\r\n <ul id=\"myNode\"></ul>\n"`
`75`	`72`	`+ "foo\n"`
`76`		`- + "</body>\r\n"`
`77`		`- + "</html>", page.asXml());`
	`73`	`+ + "</body></html>", page.asXml());`
`78`	`74`	`}`
`79`	`75`	`}`