// Copyright 2016 PDFium Authors. All rights reserved. // Use of this source code is governed by a BSD-style license that can be // found in the LICENSE file. // Original code copyright 2014 Foxit Software Inc. http://www.foxitsoftware.com #ifndef CORE_FXCRT_XML_CFX_XMLPARSER_H_ #define CORE_FXCRT_XML_CFX_XMLPARSER_H_ #include #include #include #include "core/fxcrt/fx_string.h" #include "core/fxcrt/retain_ptr.h" #include "core/fxcrt/xml/cfx_xmlnode.h" class CFX_XMLElement; class CFX_XMLNode; class IFX_SeekableReadStream; enum class FX_XmlSyntaxResult { None, InstructionOpen, InstructionClose, ElementOpen, ElementBreak, ElementClose, TargetName, TagName, AttriName, AttriValue, Text, CData, TargetData, Error, EndOfString }; class CFX_XMLParser { public: static bool IsXMLNameChar(wchar_t ch, bool bFirstChar); CFX_XMLParser(CFX_XMLNode* pParent, const RetainPtr& pStream); virtual ~CFX_XMLParser(); bool Parse(); protected: FX_XmlSyntaxResult DoSyntaxParse(); WideString GetTextData(); private: enum class FDE_XmlSyntaxState { Text, Node, Target, Tag, AttriName, AttriEqualSign, AttriQuotation, AttriValue, Entity, EntityDecimal, EntityHex, CloseInstruction, BreakElement, CloseElement, SkipDeclNode, DeclCharData, SkipComment, SkipCommentOrDecl, SkipCData, TargetData }; void ParseTextChar(wchar_t ch); bool GetStatus() const; CFX_XMLNode* m_pParent; CFX_XMLNode* m_pChild = nullptr; WideString current_attribute_name_; RetainPtr m_pStream; FX_FILESIZE m_Start = 0; // Start position in m_Buffer FX_FILESIZE m_End = 0; // End position in m_Buffer FX_XmlSyntaxResult m_syntaxParserResult = FX_XmlSyntaxResult::None; FDE_XmlSyntaxState m_syntaxParserState = FDE_XmlSyntaxState::Text; std::stack m_NodeStack; std::stack m_XMLNodeTypeStack; std::stack m_SkipStack; std::vector m_Buffer; std::vector current_text_; size_t m_iXMLPlaneSize = 1024; int32_t m_iEntityStart = -1; wchar_t m_wQuotationMark = 0; wchar_t m_SkipChar = 0; }; #endif // CORE_FXCRT_XML_CFX_XMLPARSER_H_