Remove CFX_XMLDoc and call the parser directly

This CL removes the CFX_XMLDoc and calls the CFX_XMLParser directly from
CXFA_DocumentParser.

Change-Id: I4d715cca90cd15b5e1d79827e0bb2781e873e371
Reviewed-on: https://pdfium-review.googlesource.com/30251
Reviewed-by: Henrique Nakashima <hnakashima@chromium.org>
Commit-Queue: dsinclair <dsinclair@chromium.org>

3 files changed, 17 insertions, 9 deletions

diff --git a/xfa/fxfa/parser/cxfa_dataexporter.cpp b/xfa/fxfa/parser/cxfa_dataexporter.cpp
index 7773aeb190..1418d030b5 100644
--- a/xfa/fxfa/parser/cxfa_dataexporter.cpp
+++ b/xfa/fxfa/parser/cxfa_dataexporter.cpp
@@ -7,7 +7,6 @@
 #include "xfa/fxfa/parser/cxfa_dataexporter.h"
 
 #include "core/fxcrt/fx_codepage.h"
-#include "core/fxcrt/xml/cfx_xmldoc.h"
 #include "core/fxcrt/xml/cfx_xmlelement.h"
 #include "core/fxcrt/xml/cfx_xmlnode.h"
 #include "third_party/base/stl_util.h"
diff --git a/xfa/fxfa/parser/cxfa_document_parser.cpp b/xfa/fxfa/parser/cxfa_document_parser.cpp
index 707dfc11d5..c2cc1deb88 100644
--- a/xfa/fxfa/parser/cxfa_document_parser.cpp
+++ b/xfa/fxfa/parser/cxfa_document_parser.cpp
@@ -14,7 +14,6 @@
 #include "core/fxcrt/fx_codepage.h"
 #include "core/fxcrt/fx_extension.h"
 #include "core/fxcrt/xml/cfx_xmlchardata.h"
-#include "core/fxcrt/xml/cfx_xmldoc.h"
 #include "core/fxcrt/xml/cfx_xmlelement.h"
 #include "core/fxcrt/xml/cfx_xmlinstruction.h"
 #include "core/fxcrt/xml/cfx_xmlnode.h"
@@ -345,11 +344,10 @@ bool CXFA_DocumentParser::Parse(const RetainPtr<IFX_SeekableStream>& pStream,
     pStreamProxy->SetCodePage(FX_CODEPAGE_UTF8);
   }
 
-  CFX_XMLDoc doc;
-  if (!doc.Load(pStreamProxy))
+  m_pNodeTree = LoadXML(pStreamProxy);
+  if (!m_pNodeTree)
     return false;
 
-  m_pNodeTree = doc.GetTree();
   m_pRootNode = ParseAsXDPPacket(GetDocumentNode(m_pNodeTree.get()), ePacketID);
   return !!m_pRootNode;
 }
@@ -357,13 +355,21 @@ bool CXFA_DocumentParser::Parse(const RetainPtr<IFX_SeekableStream>& pStream,
 CFX_XMLNode* CXFA_DocumentParser::ParseXMLData(const ByteString& wsXML) {
   auto pStream = pdfium::MakeRetain<CFX_SeekableStreamProxy>(
       const_cast<uint8_t*>(wsXML.raw_str()), wsXML.GetLength());
-  CFX_XMLDoc doc;
-  if (doc.Load(pStream))
-    m_pNodeTree = doc.GetTree();
-
+  m_pNodeTree = LoadXML(pStream);
   return m_pNodeTree ? GetDocumentNode(m_pNodeTree.get()) : nullptr;
 }
 
+std::unique_ptr<CFX_XMLNode> CXFA_DocumentParser::LoadXML(
+    const RetainPtr<CFX_SeekableStreamProxy>& pStream) {
+  ASSERT(pStream);
+
+  auto root = pdfium::MakeUnique<CFX_XMLNode>();
+  root->AppendChild(new CFX_XMLInstruction(L"xml"));
+
+  CFX_XMLParser parser(root.get(), pStream);
+  return parser.Parse() ? std::move(root) : nullptr;
+}
+
 void CXFA_DocumentParser::ConstructXFANode(CXFA_Node* pXFANode,
                                            CFX_XMLNode* pXMLNode) {
   XFA_PacketType ePacketID = pXFANode->GetPacketType();
diff --git a/xfa/fxfa/parser/cxfa_document_parser.h b/xfa/fxfa/parser/cxfa_document_parser.h
index e41bf7bf07..8899494c79 100644
--- a/xfa/fxfa/parser/cxfa_document_parser.h
+++ b/xfa/fxfa/parser/cxfa_document_parser.h
@@ -39,6 +39,9 @@ class CXFA_DocumentParser {
   void SetFactory(CXFA_Document* pFactory);
 
  private:
+  std::unique_ptr<CFX_XMLNode> LoadXML(
+      const RetainPtr<CFX_SeekableStreamProxy>& pStream);
+
   CXFA_Node* ParseAsXDPPacket(CFX_XMLNode* pXMLDocumentNode,
                               XFA_PacketType ePacketID);
   CXFA_Node* ParseAsXDPPacket_XDP(CFX_XMLNode* pXMLDocumentNode);