From ef755ea5d5e8fbbd9f6a9d2494960744d8607a46 Mon Sep 17 00:00:00 2001 From: Dan Sinclair Date: Mon, 7 May 2018 14:25:15 +0000 Subject: [xml] Move members to method variables in CFX_XMLParser This CL moves several of the member variables to CFX_XMLParser::DoSyntaxParse as they are only used in that method. The names of the members have been updated to make their usage clearer. Change-Id: I460f2bad0181ece278752bd8d93d6bf63dc6d70a Reviewed-on: https://pdfium-review.googlesource.com/32050 Commit-Queue: dsinclair Reviewed-by: Henrique Nakashima Reviewed-by: Ryan Harrison --- core/fxcrt/xml/cfx_xmlparser.cpp | 259 +++++++++++++++++++++------------------ core/fxcrt/xml/cfx_xmlparser.h | 9 -- 2 files changed, 138 insertions(+), 130 deletions(-) (limited to 'core/fxcrt') diff --git a/core/fxcrt/xml/cfx_xmlparser.cpp b/core/fxcrt/xml/cfx_xmlparser.cpp index 7a571e38b0..268774a10f 100644 --- a/core/fxcrt/xml/cfx_xmlparser.cpp +++ b/core/fxcrt/xml/cfx_xmlparser.cpp @@ -83,62 +83,74 @@ std::unique_ptr CFX_XMLParser::Parse() { auto doc = pdfium::MakeUnique(); current_node_ = doc->GetRoot(); + return DoSyntaxParse(doc.get()) ? std::move(doc) : nullptr; +} + +bool CFX_XMLParser::DoSyntaxParse(CFX_XMLDocument* doc) { + FX_FILESIZE current_buffer_idx = 0; + FX_FILESIZE buffer_size = 0; + FX_SAFE_SIZE_T alloc_size_safe = m_iXMLPlaneSize; alloc_size_safe += 1; // For NUL. if (!alloc_size_safe.IsValid() || alloc_size_safe.ValueOrDie() <= 0) - return nullptr; + return false; - m_Buffer.resize(pdfium::base::ValueOrDieForType(alloc_size_safe)); + std::vector buffer; + buffer.resize(pdfium::base::ValueOrDieForType(alloc_size_safe)); - return DoSyntaxParse(doc.get()) ? std::move(doc) : nullptr; -} - -bool CFX_XMLParser::DoSyntaxParse(CFX_XMLDocument* doc) { + std::stack character_to_skip_too_stack; + std::stack node_type_stack; + WideString current_attribute_name; + FDE_XmlSyntaxState current_parser_state = FDE_XmlSyntaxState::Text; int32_t iCount = 0; + wchar_t current_quote_character = 0; + wchar_t current_character_to_skip_to = 0; + while (true) { - if (m_Start >= m_End) { + if (current_buffer_idx >= buffer_size) { if (m_pStream->IsEOF()) return true; size_t buffer_chars = - m_pStream->ReadBlock(m_Buffer.data(), m_iXMLPlaneSize); + m_pStream->ReadBlock(buffer.data(), m_iXMLPlaneSize); if (buffer_chars == 0) return true; - m_Start = 0; - m_End = buffer_chars; + current_buffer_idx = 0; + buffer_size = buffer_chars; } - while (m_Start < m_End) { - wchar_t ch = m_Buffer[m_Start]; - switch (m_syntaxParserState) { + while (current_buffer_idx < buffer_size) { + wchar_t ch = buffer[current_buffer_idx]; + switch (current_parser_state) { case FDE_XmlSyntaxState::Text: if (ch == L'<') { if (!current_text_.empty()) { current_node_->AppendChild( doc->CreateNode(GetTextData())); } else { - m_Start++; - m_syntaxParserState = FDE_XmlSyntaxState::Node; + current_buffer_idx++; + current_parser_state = FDE_XmlSyntaxState::Node; } } else { ProcessTextChar(ch); + current_buffer_idx++; } break; case FDE_XmlSyntaxState::Node: if (ch == L'!') { - m_Start++; - m_syntaxParserState = FDE_XmlSyntaxState::SkipCommentOrDecl; + current_buffer_idx++; + current_parser_state = FDE_XmlSyntaxState::SkipCommentOrDecl; } else if (ch == L'/') { - m_Start++; - m_syntaxParserState = FDE_XmlSyntaxState::CloseElement; + current_buffer_idx++; + current_parser_state = FDE_XmlSyntaxState::CloseElement; } else if (ch == L'?') { - m_XMLNodeTypeStack.push(FX_XMLNODE_Instruction); - m_Start++; - m_syntaxParserState = FDE_XmlSyntaxState::Target; + node_type_stack.push(FX_XMLNODE_Instruction); + current_buffer_idx++; + current_parser_state = FDE_XmlSyntaxState::Target; } else { - m_XMLNodeTypeStack.push(FX_XMLNODE_Element); - m_syntaxParserState = FDE_XmlSyntaxState::Tag; + node_type_stack.push(FX_XMLNODE_Element); + current_parser_state = FDE_XmlSyntaxState::Tag; } break; case FDE_XmlSyntaxState::Target: @@ -146,7 +158,7 @@ bool CFX_XMLParser::DoSyntaxParse(CFX_XMLDocument* doc) { if (current_text_.empty()) return false; - m_syntaxParserState = FDE_XmlSyntaxState::TargetData; + current_parser_state = FDE_XmlSyntaxState::TargetData; WideString target_name = GetTextData(); if (target_name == L"originalXFAVersion" || @@ -157,7 +169,7 @@ bool CFX_XMLParser::DoSyntaxParse(CFX_XMLDocument* doc) { } } else { current_text_.push_back(ch); - m_Start++; + current_buffer_idx++; } break; case FDE_XmlSyntaxState::Tag: @@ -165,114 +177,115 @@ bool CFX_XMLParser::DoSyntaxParse(CFX_XMLDocument* doc) { if (current_text_.empty()) return false; - m_syntaxParserState = FDE_XmlSyntaxState::AttriName; + current_parser_state = FDE_XmlSyntaxState::AttriName; auto* child = doc->CreateNode(GetTextData()); current_node_->AppendChild(child); current_node_ = child; } else { current_text_.push_back(ch); - m_Start++; + current_buffer_idx++; } break; case FDE_XmlSyntaxState::AttriName: if (current_text_.empty() && IsXMLWhiteSpace(ch)) { - m_Start++; + current_buffer_idx++; break; } if (!IsXMLNameChar(ch, current_text_.empty())) { if (current_text_.empty()) { - if (m_XMLNodeTypeStack.top() == FX_XMLNODE_Element) { + if (node_type_stack.top() == FX_XMLNODE_Element) { if (ch == L'>' || ch == L'/') { - m_syntaxParserState = FDE_XmlSyntaxState::BreakElement; + current_parser_state = FDE_XmlSyntaxState::BreakElement; break; } - } else if (m_XMLNodeTypeStack.top() == FX_XMLNODE_Instruction) { + } else if (node_type_stack.top() == FX_XMLNODE_Instruction) { if (ch == L'?') { - m_syntaxParserState = FDE_XmlSyntaxState::CloseInstruction; - m_Start++; + current_parser_state = FDE_XmlSyntaxState::CloseInstruction; + current_buffer_idx++; } else { - m_syntaxParserState = FDE_XmlSyntaxState::TargetData; + current_parser_state = FDE_XmlSyntaxState::TargetData; } break; } return false; } else { - if (m_XMLNodeTypeStack.top() == FX_XMLNODE_Instruction) { + if (node_type_stack.top() == FX_XMLNODE_Instruction) { if (ch != '=' && !IsXMLWhiteSpace(ch)) { - m_syntaxParserState = FDE_XmlSyntaxState::TargetData; + current_parser_state = FDE_XmlSyntaxState::TargetData; break; } } - m_syntaxParserState = FDE_XmlSyntaxState::AttriEqualSign; - current_attribute_name_ = GetTextData(); + current_parser_state = FDE_XmlSyntaxState::AttriEqualSign; + current_attribute_name = GetTextData(); } } else { current_text_.push_back(ch); - m_Start++; + current_buffer_idx++; } break; case FDE_XmlSyntaxState::AttriEqualSign: if (IsXMLWhiteSpace(ch)) { - m_Start++; + current_buffer_idx++; break; } if (ch != L'=') { - if (m_XMLNodeTypeStack.top() == FX_XMLNODE_Instruction) { - m_syntaxParserState = FDE_XmlSyntaxState::TargetData; + if (node_type_stack.top() == FX_XMLNODE_Instruction) { + current_parser_state = FDE_XmlSyntaxState::TargetData; break; } return false; } else { - m_syntaxParserState = FDE_XmlSyntaxState::AttriQuotation; - m_Start++; + current_parser_state = FDE_XmlSyntaxState::AttriQuotation; + current_buffer_idx++; } break; case FDE_XmlSyntaxState::AttriQuotation: if (IsXMLWhiteSpace(ch)) { - m_Start++; + current_buffer_idx++; break; } if (ch != L'\"' && ch != L'\'') { return false; } - m_wQuotationMark = ch; - m_syntaxParserState = FDE_XmlSyntaxState::AttriValue; - m_Start++; + current_quote_character = ch; + current_parser_state = FDE_XmlSyntaxState::AttriValue; + current_buffer_idx++; break; case FDE_XmlSyntaxState::AttriValue: - if (ch == m_wQuotationMark) { + if (ch == current_quote_character) { if (m_iEntityStart > -1) return false; - m_wQuotationMark = 0; - m_Start++; - m_syntaxParserState = FDE_XmlSyntaxState::AttriName; + current_quote_character = 0; + current_buffer_idx++; + current_parser_state = FDE_XmlSyntaxState::AttriName; if (current_node_ && current_node_->GetType() == FX_XMLNODE_Element) { static_cast(current_node_) - ->SetAttribute(current_attribute_name_, GetTextData()); + ->SetAttribute(current_attribute_name, GetTextData()); } - current_attribute_name_.clear(); + current_attribute_name.clear(); } else { ProcessTextChar(ch); + current_buffer_idx++; } break; case FDE_XmlSyntaxState::CloseInstruction: if (ch != L'>') { current_text_.push_back(ch); - m_syntaxParserState = FDE_XmlSyntaxState::TargetData; + current_parser_state = FDE_XmlSyntaxState::TargetData; } else if (!current_text_.empty()) { ProcessTargetData(); } else { - m_Start++; - if (m_XMLNodeTypeStack.empty()) + current_buffer_idx++; + if (node_type_stack.empty()) return false; - m_XMLNodeTypeStack.pop(); - m_syntaxParserState = FDE_XmlSyntaxState::Text; + node_type_stack.pop(); + current_parser_state = FDE_XmlSyntaxState::Text; if (current_node_ && current_node_->GetType() == FX_XMLNODE_Instruction) @@ -281,22 +294,22 @@ bool CFX_XMLParser::DoSyntaxParse(CFX_XMLDocument* doc) { break; case FDE_XmlSyntaxState::BreakElement: if (ch == L'>') { - m_syntaxParserState = FDE_XmlSyntaxState::Text; + current_parser_state = FDE_XmlSyntaxState::Text; } else if (ch == L'/') { - m_syntaxParserState = FDE_XmlSyntaxState::CloseElement; + current_parser_state = FDE_XmlSyntaxState::CloseElement; } else { return false; } - m_Start++; + current_buffer_idx++; break; case FDE_XmlSyntaxState::CloseElement: if (!IsXMLNameChar(ch, current_text_.empty())) { if (ch == L'>') { - if (m_XMLNodeTypeStack.empty()) + if (node_type_stack.empty()) return false; - m_XMLNodeTypeStack.pop(); - m_syntaxParserState = FDE_XmlSyntaxState::Text; + node_type_stack.pop(); + current_parser_state = FDE_XmlSyntaxState::Text; if (current_node_->GetType() != FX_XMLNODE_Element) return false; @@ -316,119 +329,124 @@ bool CFX_XMLParser::DoSyntaxParse(CFX_XMLDocument* doc) { } else { current_text_.push_back(ch); } - m_Start++; + current_buffer_idx++; break; case FDE_XmlSyntaxState::SkipCommentOrDecl: - if (FXSYS_wcsnicmp(m_Buffer.data() + m_Start, L"--", 2) == 0) { - m_Start += 2; - m_syntaxParserState = FDE_XmlSyntaxState::SkipComment; - } else if (FXSYS_wcsnicmp(m_Buffer.data() + m_Start, L"[CDATA[", 7) == - 0) { - m_Start += 7; - m_syntaxParserState = FDE_XmlSyntaxState::SkipCData; + if (FXSYS_wcsnicmp(buffer.data() + current_buffer_idx, L"--", 2) == + 0) { + current_buffer_idx += 2; + current_parser_state = FDE_XmlSyntaxState::SkipComment; + } else if (FXSYS_wcsnicmp(buffer.data() + current_buffer_idx, + L"[CDATA[", 7) == 0) { + current_buffer_idx += 7; + current_parser_state = FDE_XmlSyntaxState::SkipCData; } else { - m_syntaxParserState = FDE_XmlSyntaxState::SkipDeclNode; - m_SkipChar = L'>'; - m_SkipStack.push(L'>'); + current_parser_state = FDE_XmlSyntaxState::SkipDeclNode; + current_character_to_skip_to = L'>'; + character_to_skip_too_stack.push(L'>'); } break; case FDE_XmlSyntaxState::SkipCData: { - if (FXSYS_wcsnicmp(m_Buffer.data() + m_Start, L"]]>", 3) == 0) { - m_Start += 3; - m_syntaxParserState = FDE_XmlSyntaxState::Text; + if (FXSYS_wcsnicmp(buffer.data() + current_buffer_idx, L"]]>", 3) == + 0) { + current_buffer_idx += 3; + current_parser_state = FDE_XmlSyntaxState::Text; current_node_->AppendChild( doc->CreateNode(GetTextData())); } else { current_text_.push_back(ch); - m_Start++; + current_buffer_idx++; } break; } case FDE_XmlSyntaxState::SkipDeclNode: - if (m_SkipChar == L'\'' || m_SkipChar == L'\"') { - m_Start++; - if (ch != m_SkipChar) + if (current_character_to_skip_to == L'\'' || + current_character_to_skip_to == L'\"') { + current_buffer_idx++; + if (ch != current_character_to_skip_to) break; - m_SkipStack.pop(); - if (m_SkipStack.empty()) - m_syntaxParserState = FDE_XmlSyntaxState::Text; + character_to_skip_too_stack.pop(); + if (character_to_skip_too_stack.empty()) + current_parser_state = FDE_XmlSyntaxState::Text; else - m_SkipChar = m_SkipStack.top(); + current_character_to_skip_to = character_to_skip_too_stack.top(); } else { switch (ch) { case L'<': - m_SkipChar = L'>'; - m_SkipStack.push(L'>'); + current_character_to_skip_to = L'>'; + character_to_skip_too_stack.push(L'>'); break; case L'[': - m_SkipChar = L']'; - m_SkipStack.push(L']'); + current_character_to_skip_to = L']'; + character_to_skip_too_stack.push(L']'); break; case L'(': - m_SkipChar = L')'; - m_SkipStack.push(L')'); + current_character_to_skip_to = L')'; + character_to_skip_too_stack.push(L')'); break; case L'\'': - m_SkipChar = L'\''; - m_SkipStack.push(L'\''); + current_character_to_skip_to = L'\''; + character_to_skip_too_stack.push(L'\''); break; case L'\"': - m_SkipChar = L'\"'; - m_SkipStack.push(L'\"'); + current_character_to_skip_to = L'\"'; + character_to_skip_too_stack.push(L'\"'); break; default: - if (ch == m_SkipChar) { - m_SkipStack.pop(); - if (m_SkipStack.empty()) { - m_syntaxParserState = FDE_XmlSyntaxState::Text; + if (ch == current_character_to_skip_to) { + character_to_skip_too_stack.pop(); + if (character_to_skip_too_stack.empty()) { + current_parser_state = FDE_XmlSyntaxState::Text; } else { - m_SkipChar = m_SkipStack.top(); + current_character_to_skip_to = + character_to_skip_too_stack.top(); } } break; } - m_Start++; + current_buffer_idx++; } break; case FDE_XmlSyntaxState::SkipComment: - if (FXSYS_wcsnicmp(m_Buffer.data() + m_Start, L"-->", 3) == 0) { - m_Start += 2; - m_syntaxParserState = FDE_XmlSyntaxState::Text; + if (FXSYS_wcsnicmp(buffer.data() + current_buffer_idx, L"-->", 3) == + 0) { + current_buffer_idx += 2; + current_parser_state = FDE_XmlSyntaxState::Text; } - m_Start++; + current_buffer_idx++; break; case FDE_XmlSyntaxState::TargetData: if (IsXMLWhiteSpace(ch)) { if (current_text_.empty()) { - m_Start++; + current_buffer_idx++; break; } - if (m_wQuotationMark == 0) { - m_Start++; + if (current_quote_character == 0) { + current_buffer_idx++; ProcessTargetData(); break; } } if (ch == '?') { - m_syntaxParserState = FDE_XmlSyntaxState::CloseInstruction; - m_Start++; + current_parser_state = FDE_XmlSyntaxState::CloseInstruction; + current_buffer_idx++; } else if (ch == '\"') { - if (m_wQuotationMark == 0) { - m_wQuotationMark = ch; - m_Start++; - } else if (ch == m_wQuotationMark) { - m_wQuotationMark = 0; - m_Start++; + if (current_quote_character == 0) { + current_quote_character = ch; + current_buffer_idx++; + } else if (ch == current_quote_character) { + current_quote_character = 0; + current_buffer_idx++; ProcessTargetData(); } else { return false; } } else { current_text_.push_back(ch); - m_Start++; + current_buffer_idx++; } break; default: @@ -495,7 +513,6 @@ void CFX_XMLParser::ProcessTextChar(wchar_t character) { } else if (m_iEntityStart < 0 && character == L'&') { m_iEntityStart = current_text_.size() - 1; } - m_Start++; } void CFX_XMLParser::ProcessTargetData() { diff --git a/core/fxcrt/xml/cfx_xmlparser.h b/core/fxcrt/xml/cfx_xmlparser.h index aa5ad5761c..4d9184de3e 100644 --- a/core/fxcrt/xml/cfx_xmlparser.h +++ b/core/fxcrt/xml/cfx_xmlparser.h @@ -55,19 +55,10 @@ class CFX_XMLParser { void ProcessTargetData(); CFX_XMLNode* current_node_ = 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 - FDE_XmlSyntaxState m_syntaxParserState = FDE_XmlSyntaxState::Text; - 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_ -- cgit v1.2.3