From 8ae06d522310b09282c3bcb71f9793d7a41b5ecb Mon Sep 17 00:00:00 2001
From: Nick Wellnhofer <wellnhofer@aevum.de>
Date: Thu, 29 Aug 2024 00:07:27 +0200
Subject: [PATCH] SAX2: Don't merge CDATA sections

The Document Object Model (DOM) Level 3 Core Specification says:

> Adjacent CDATASection nodes are not merged by use of the normalize
> method of the Node interface.

Fixes #412.
---
 SAX2.c                               | 2 +-
 result/adjacent-cdata.xml            | 2 ++
 result/adjacent-cdata.xml.rde        | 5 +++++
 result/adjacent-cdata.xml.rdr        | 5 +++++
 result/adjacent-cdata.xml.sax        | 8 ++++++++
 result/adjacent-cdata.xml.sax2       | 8 ++++++++
 result/noent/adjacent-cdata.xml      | 2 ++
 result/noent/adjacent-cdata.xml.sax2 | 8 ++++++++
 test/adjacent-cdata.xml              | 1 +
 9 files changed, 40 insertions(+), 1 deletion(-)
 create mode 100644 result/adjacent-cdata.xml
 create mode 100644 result/adjacent-cdata.xml.rde
 create mode 100644 result/adjacent-cdata.xml.rdr
 create mode 100644 result/adjacent-cdata.xml.sax
 create mode 100644 result/adjacent-cdata.xml.sax2
 create mode 100644 result/noent/adjacent-cdata.xml
 create mode 100644 result/noent/adjacent-cdata.xml.sax2
 create mode 100644 test/adjacent-cdata.xml

diff --git a/SAX2.c b/SAX2.c
index 4c3c7329..1a9310d5 100644
--- a/SAX2.c
+++ b/SAX2.c
@@ -2471,7 +2471,7 @@ xmlSAX2Text(xmlParserCtxtPtr ctxt, const xmlChar *ch, int len,
     } else {
 	int coalesceText = (lastChild != NULL) &&
 	    (lastChild->type == type) &&
-	    ((type != XML_TEXT_NODE) ||
+	    (((ctxt->html) && (type != XML_TEXT_NODE)) ||
              (lastChild->name == xmlStringText));
 	if ((coalesceText) && (ctxt->nodemem != 0)) {
             int maxLength = (ctxt->options & XML_PARSE_HUGE) ?
diff --git a/result/adjacent-cdata.xml b/result/adjacent-cdata.xml
new file mode 100644
index 00000000..fada9eee
--- /dev/null
+++ b/result/adjacent-cdata.xml
@@ -0,0 +1,2 @@
+<?xml version="1.0"?>
+<doc><![CDATA[abc]]><![CDATA[def]]><![CDATA[ghi]]></doc>
diff --git a/result/adjacent-cdata.xml.rde b/result/adjacent-cdata.xml.rde
new file mode 100644
index 00000000..026949e0
--- /dev/null
+++ b/result/adjacent-cdata.xml.rde
@@ -0,0 +1,5 @@
+0 1 doc 0 0
+1 4 #cdata-section 0 1 abc
+1 4 #cdata-section 0 1 def
+1 4 #cdata-section 0 1 ghi
+0 15 doc 0 0
diff --git a/result/adjacent-cdata.xml.rdr b/result/adjacent-cdata.xml.rdr
new file mode 100644
index 00000000..026949e0
--- /dev/null
+++ b/result/adjacent-cdata.xml.rdr
@@ -0,0 +1,5 @@
+0 1 doc 0 0
+1 4 #cdata-section 0 1 abc
+1 4 #cdata-section 0 1 def
+1 4 #cdata-section 0 1 ghi
+0 15 doc 0 0
diff --git a/result/adjacent-cdata.xml.sax b/result/adjacent-cdata.xml.sax
new file mode 100644
index 00000000..c7806da7
--- /dev/null
+++ b/result/adjacent-cdata.xml.sax
@@ -0,0 +1,8 @@
+SAX.setDocumentLocator()
+SAX.startDocument()
+SAX.startElement(doc)
+SAX.pcdata(abc, 3)
+SAX.pcdata(def, 3)
+SAX.pcdata(ghi, 3)
+SAX.endElement(doc)
+SAX.endDocument()
diff --git a/result/adjacent-cdata.xml.sax2 b/result/adjacent-cdata.xml.sax2
new file mode 100644
index 00000000..56b4b3b7
--- /dev/null
+++ b/result/adjacent-cdata.xml.sax2
@@ -0,0 +1,8 @@
+SAX.setDocumentLocator()
+SAX.startDocument()
+SAX.startElementNs(doc, NULL, NULL, 0, 0, 0)
+SAX.pcdata(abc, 3)
+SAX.pcdata(def, 3)
+SAX.pcdata(ghi, 3)
+SAX.endElementNs(doc, NULL, NULL)
+SAX.endDocument()
diff --git a/result/noent/adjacent-cdata.xml b/result/noent/adjacent-cdata.xml
new file mode 100644
index 00000000..fada9eee
--- /dev/null
+++ b/result/noent/adjacent-cdata.xml
@@ -0,0 +1,2 @@
+<?xml version="1.0"?>
+<doc><![CDATA[abc]]><![CDATA[def]]><![CDATA[ghi]]></doc>
diff --git a/result/noent/adjacent-cdata.xml.sax2 b/result/noent/adjacent-cdata.xml.sax2
new file mode 100644
index 00000000..56b4b3b7
--- /dev/null
+++ b/result/noent/adjacent-cdata.xml.sax2
@@ -0,0 +1,8 @@
+SAX.setDocumentLocator()
+SAX.startDocument()
+SAX.startElementNs(doc, NULL, NULL, 0, 0, 0)
+SAX.pcdata(abc, 3)
+SAX.pcdata(def, 3)
+SAX.pcdata(ghi, 3)
+SAX.endElementNs(doc, NULL, NULL)
+SAX.endDocument()
diff --git a/test/adjacent-cdata.xml b/test/adjacent-cdata.xml
new file mode 100644
index 00000000..5859226a
--- /dev/null
+++ b/test/adjacent-cdata.xml
@@ -0,0 +1 @@
+<doc><![CDATA[abc]]><![CDATA[def]]><![CDATA[ghi]]></doc>