| // Copyright 2014 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 |
| |
| #include <algorithm> |
| #include <memory> |
| #include <sstream> |
| #include <string> |
| #include <utility> |
| #include <vector> |
| |
| #include "core/fxcrt/cfx_utf8decoder.h" |
| #include "core/fxcrt/cfx_widetextbuf.h" |
| #include "core/fxcrt/fx_extension.h" |
| #include "core/fxcrt/xml/cxml_content.h" |
| #include "core/fxcrt/xml/cxml_element.h" |
| #include "core/fxcrt/xml/cxml_parser.h" |
| #include "third_party/base/ptr_util.h" |
| #include "third_party/base/stl_util.h" |
| |
| namespace { |
| |
| #define FXCRTM_XML_CHARTYPE_Normal 0x00 |
| #define FXCRTM_XML_CHARTYPE_SpaceChar 0x01 |
| #define FXCRTM_XML_CHARTYPE_Letter 0x02 |
| #define FXCRTM_XML_CHARTYPE_Digital 0x04 |
| #define FXCRTM_XML_CHARTYPE_NameIntro 0x08 |
| #define FXCRTM_XML_CHARTYPE_NameChar 0x10 |
| #define FXCRTM_XML_CHARTYPE_HexDigital 0x20 |
| #define FXCRTM_XML_CHARTYPE_HexLowerLetter 0x40 |
| #define FXCRTM_XML_CHARTYPE_HexUpperLetter 0x60 |
| #define FXCRTM_XML_CHARTYPE_HexChar 0x60 |
| |
| const uint8_t g_FXCRT_XML_ByteTypes[256] = { |
| 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, |
| 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, |
| 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x00, 0x00, 0x00, |
| 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x10, 0x10, 0x00, |
| 0x34, 0x34, 0x34, 0x34, 0x34, 0x34, 0x34, 0x34, 0x34, 0x34, 0x08, 0x00, |
| 0x00, 0x00, 0x00, 0x00, 0x00, 0x7A, 0x7A, 0x7A, 0x7A, 0x7A, 0x7A, 0x1A, |
| 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, |
| 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x00, 0x00, 0x00, 0x00, 0x18, |
| 0x00, 0x5A, 0x5A, 0x5A, 0x5A, 0x5A, 0x5A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, |
| 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, |
| 0x1A, 0x1A, 0x1A, 0x00, 0x00, 0x00, 0x00, 0x00, 0x1A, 0x1A, 0x1A, 0x1A, |
| 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, |
| 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, |
| 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, |
| 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, |
| 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, |
| 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, |
| 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, |
| 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, |
| 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, |
| 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, 0x1A, |
| 0x1A, 0x1A, 0x01, 0x01, |
| }; |
| |
| constexpr int kMaxDepth = 1024; |
| |
| bool g_FXCRT_XML_IsWhiteSpace(uint8_t ch) { |
| return !!(g_FXCRT_XML_ByteTypes[ch] & FXCRTM_XML_CHARTYPE_SpaceChar); |
| } |
| |
| bool g_FXCRT_XML_IsDigital(uint8_t ch) { |
| return !!(g_FXCRT_XML_ByteTypes[ch] & FXCRTM_XML_CHARTYPE_Digital); |
| } |
| |
| bool g_FXCRT_XML_IsNameIntro(uint8_t ch) { |
| return !!(g_FXCRT_XML_ByteTypes[ch] & FXCRTM_XML_CHARTYPE_NameIntro); |
| } |
| |
| bool g_FXCRT_XML_IsNameChar(uint8_t ch) { |
| return !!(g_FXCRT_XML_ByteTypes[ch] & FXCRTM_XML_CHARTYPE_NameChar); |
| } |
| |
| } // namespace |
| |
| void FX_XML_SplitQualifiedName(const ByteStringView& bsFullName, |
| ByteStringView& bsSpace, |
| ByteStringView& bsName) { |
| if (bsFullName.IsEmpty()) |
| return; |
| |
| auto iStart = bsFullName.Find(':'); |
| if (!iStart.has_value()) { |
| bsName = bsFullName; |
| } else { |
| bsSpace = bsFullName.Left(iStart.value()); |
| bsName = bsFullName.Right(bsFullName.GetLength() - (iStart.value() + 1)); |
| } |
| } |
| |
| CXML_Parser::CXML_Parser() |
| : m_nOffset(0), |
| m_pBuffer(nullptr), |
| m_dwBufferSize(0), |
| m_nBufferOffset(0), |
| m_dwIndex(0) {} |
| |
| CXML_Parser::~CXML_Parser() {} |
| |
| bool CXML_Parser::Init(const uint8_t* pBuffer, size_t size) { |
| m_pDataAcc = pdfium::MakeUnique<CXML_DataBufAcc>(pBuffer, size); |
| m_nOffset = 0; |
| return ReadNextBlock(); |
| } |
| |
| bool CXML_Parser::ReadNextBlock() { |
| if (!m_pDataAcc->ReadNextBlock()) |
| return false; |
| |
| m_pBuffer = m_pDataAcc->GetBlockBuffer(); |
| m_dwBufferSize = m_pDataAcc->GetBlockSize(); |
| m_nBufferOffset = 0; |
| m_dwIndex = 0; |
| return m_dwBufferSize > 0; |
| } |
| |
| bool CXML_Parser::IsEOF() { |
| return m_pDataAcc->IsEOF() && m_dwIndex >= m_dwBufferSize; |
| } |
| |
| void CXML_Parser::SkipWhiteSpaces() { |
| m_nOffset = m_nBufferOffset + static_cast<FX_FILESIZE>(m_dwIndex); |
| if (IsEOF()) |
| return; |
| |
| do { |
| while (m_dwIndex < m_dwBufferSize && |
| g_FXCRT_XML_IsWhiteSpace(m_pBuffer[m_dwIndex])) { |
| m_dwIndex++; |
| } |
| m_nOffset = m_nBufferOffset + static_cast<FX_FILESIZE>(m_dwIndex); |
| if (m_dwIndex < m_dwBufferSize || IsEOF()) |
| break; |
| } while (ReadNextBlock()); |
| } |
| |
| void CXML_Parser::GetName(ByteString* space, ByteString* name) { |
| m_nOffset = m_nBufferOffset + static_cast<FX_FILESIZE>(m_dwIndex); |
| if (IsEOF()) |
| return; |
| |
| std::ostringstream buf; |
| do { |
| while (m_dwIndex < m_dwBufferSize) { |
| uint8_t ch = m_pBuffer[m_dwIndex]; |
| if (ch == ':') { |
| *space = ByteString(buf); |
| buf.str(""); |
| } else if (g_FXCRT_XML_IsNameChar(ch)) { |
| buf << static_cast<char>(ch); |
| } else { |
| break; |
| } |
| m_dwIndex++; |
| } |
| m_nOffset = m_nBufferOffset + static_cast<FX_FILESIZE>(m_dwIndex); |
| if (m_dwIndex < m_dwBufferSize || IsEOF()) |
| break; |
| } while (ReadNextBlock()); |
| *name = ByteString(buf); |
| } |
| |
| void CXML_Parser::SkipLiterals(const ByteStringView& str) { |
| m_nOffset = m_nBufferOffset + static_cast<FX_FILESIZE>(m_dwIndex); |
| if (IsEOF()) { |
| return; |
| } |
| int32_t i = 0, iLen = str.GetLength(); |
| do { |
| while (m_dwIndex < m_dwBufferSize) { |
| if (str[i] != m_pBuffer[m_dwIndex++]) { |
| i = 0; |
| continue; |
| } |
| i++; |
| if (i == iLen) |
| break; |
| } |
| m_nOffset = m_nBufferOffset + static_cast<FX_FILESIZE>(m_dwIndex); |
| if (i == iLen) |
| return; |
| |
| if (m_dwIndex < m_dwBufferSize || IsEOF()) |
| break; |
| } while (ReadNextBlock()); |
| while (!m_pDataAcc->IsEOF()) { |
| ReadNextBlock(); |
| m_nOffset = m_nBufferOffset + static_cast<FX_FILESIZE>(m_dwBufferSize); |
| } |
| m_dwIndex = m_dwBufferSize; |
| } |
| |
| uint32_t CXML_Parser::GetCharRef() { |
| m_nOffset = m_nBufferOffset + static_cast<FX_FILESIZE>(m_dwIndex); |
| if (IsEOF()) |
| return 0; |
| |
| uint8_t ch; |
| int32_t iState = 0; |
| std::ostringstream buf; |
| uint32_t code = 0; |
| do { |
| while (m_dwIndex < m_dwBufferSize) { |
| ch = m_pBuffer[m_dwIndex]; |
| switch (iState) { |
| case 0: |
| if (ch == '#') { |
| m_dwIndex++; |
| iState = 2; |
| break; |
| } |
| iState = 1; |
| case 1: |
| m_dwIndex++; |
| if (ch == ';') { |
| std::string ref = buf.str(); |
| if (ref == "gt") |
| code = '>'; |
| else if (ref == "lt") |
| code = '<'; |
| else if (ref == "amp") |
| code = '&'; |
| else if (ref == "apos") |
| code = '\''; |
| else if (ref == "quot") |
| code = '"'; |
| iState = 10; |
| break; |
| } |
| buf << static_cast<char>(ch); |
| break; |
| case 2: |
| if (ch == 'x') { |
| m_dwIndex++; |
| iState = 4; |
| break; |
| } |
| iState = 3; |
| case 3: |
| m_dwIndex++; |
| if (ch == ';') { |
| iState = 10; |
| break; |
| } |
| if (g_FXCRT_XML_IsDigital(ch)) |
| code = code * 10 + FXSYS_DecimalCharToInt(static_cast<wchar_t>(ch)); |
| break; |
| case 4: |
| m_dwIndex++; |
| if (ch == ';') { |
| iState = 10; |
| break; |
| } |
| uint8_t nHex = |
| g_FXCRT_XML_ByteTypes[ch] & FXCRTM_XML_CHARTYPE_HexChar; |
| if (nHex) { |
| if (nHex == FXCRTM_XML_CHARTYPE_HexDigital) { |
| code = (code << 4) + |
| FXSYS_DecimalCharToInt(static_cast<wchar_t>(ch)); |
| } else if (nHex == FXCRTM_XML_CHARTYPE_HexLowerLetter) { |
| code = (code << 4) + ch - 87; |
| } else { |
| code = (code << 4) + ch - 55; |
| } |
| } |
| break; |
| } |
| if (iState == 10) |
| break; |
| } |
| m_nOffset = m_nBufferOffset + static_cast<FX_FILESIZE>(m_dwIndex); |
| if (iState == 10 || m_dwIndex < m_dwBufferSize || IsEOF()) { |
| break; |
| } |
| } while (ReadNextBlock()); |
| return code; |
| } |
| |
| void CXML_Parser::GetAttrValue(WideString& value) { |
| m_nOffset = m_nBufferOffset + static_cast<FX_FILESIZE>(m_dwIndex); |
| if (IsEOF()) |
| return; |
| |
| CFX_UTF8Decoder decoder; |
| uint8_t mark = 0, ch = 0; |
| do { |
| while (m_dwIndex < m_dwBufferSize) { |
| ch = m_pBuffer[m_dwIndex]; |
| if (mark == 0) { |
| if (ch != '\'' && ch != '"') |
| return; |
| |
| mark = ch; |
| m_dwIndex++; |
| ch = 0; |
| continue; |
| } |
| m_dwIndex++; |
| if (ch == mark) |
| break; |
| |
| if (ch == '&') { |
| decoder.AppendCodePoint(GetCharRef()); |
| if (IsEOF()) { |
| value = decoder.GetResult(); |
| return; |
| } |
| } else { |
| decoder.Input(ch); |
| } |
| } |
| m_nOffset = m_nBufferOffset + static_cast<FX_FILESIZE>(m_dwIndex); |
| if (ch == mark || m_dwIndex < m_dwBufferSize || IsEOF()) |
| break; |
| } while (ReadNextBlock()); |
| value = decoder.GetResult(); |
| } |
| |
| void CXML_Parser::GetTagName(bool bStartTag, |
| bool* bEndTag, |
| ByteString* space, |
| ByteString* name) { |
| m_nOffset = m_nBufferOffset + static_cast<FX_FILESIZE>(m_dwIndex); |
| if (IsEOF()) |
| return; |
| |
| *bEndTag = false; |
| uint8_t ch; |
| int32_t iState = bStartTag ? 1 : 0; |
| do { |
| while (m_dwIndex < m_dwBufferSize) { |
| ch = m_pBuffer[m_dwIndex]; |
| switch (iState) { |
| case 0: |
| m_dwIndex++; |
| if (ch != '<') |
| break; |
| |
| iState = 1; |
| break; |
| case 1: |
| if (ch == '?') { |
| m_dwIndex++; |
| SkipLiterals("?>"); |
| iState = 0; |
| break; |
| } |
| if (ch == '!') { |
| m_dwIndex++; |
| SkipLiterals("-->"); |
| iState = 0; |
| break; |
| } |
| if (ch == '/') { |
| m_dwIndex++; |
| GetName(space, name); |
| *bEndTag = true; |
| } else { |
| GetName(space, name); |
| *bEndTag = false; |
| } |
| return; |
| } |
| } |
| m_nOffset = m_nBufferOffset + static_cast<FX_FILESIZE>(m_dwIndex); |
| if (m_dwIndex < m_dwBufferSize || IsEOF()) |
| break; |
| } while (ReadNextBlock()); |
| } |
| |
| std::unique_ptr<CXML_Element> CXML_Parser::ParseElement(CXML_Element* pParent, |
| bool bStartTag) { |
| return ParseElementInternal(pParent, bStartTag, 0); |
| } |
| |
| std::unique_ptr<CXML_Element> CXML_Parser::ParseElementInternal( |
| CXML_Element* pParent, |
| bool bStartTag, |
| int nDepth) { |
| if (nDepth > kMaxDepth) |
| return nullptr; |
| |
| m_nOffset = m_nBufferOffset + static_cast<FX_FILESIZE>(m_dwIndex); |
| if (IsEOF()) |
| return nullptr; |
| |
| ByteString tag_name; |
| ByteString tag_space; |
| bool bEndTag; |
| GetTagName(bStartTag, &bEndTag, &tag_space, &tag_name); |
| if (tag_name.IsEmpty() || bEndTag) |
| return nullptr; |
| |
| auto pElement = pdfium::MakeUnique<CXML_Element>( |
| pParent, tag_space.AsStringView(), tag_name.AsStringView()); |
| do { |
| ByteString attr_space; |
| ByteString attr_name; |
| while (m_dwIndex < m_dwBufferSize) { |
| SkipWhiteSpaces(); |
| if (IsEOF()) |
| break; |
| |
| if (!g_FXCRT_XML_IsNameIntro(m_pBuffer[m_dwIndex])) |
| break; |
| |
| GetName(&attr_space, &attr_name); |
| SkipWhiteSpaces(); |
| if (IsEOF()) |
| break; |
| |
| if (m_pBuffer[m_dwIndex] != '=') |
| break; |
| |
| m_dwIndex++; |
| SkipWhiteSpaces(); |
| if (IsEOF()) |
| break; |
| |
| WideString attr_value; |
| GetAttrValue(attr_value); |
| pElement->SetAttribute(attr_space, attr_name, attr_value); |
| } |
| m_nOffset = m_nBufferOffset + static_cast<FX_FILESIZE>(m_dwIndex); |
| if (m_dwIndex < m_dwBufferSize || IsEOF()) |
| break; |
| } while (ReadNextBlock()); |
| SkipWhiteSpaces(); |
| if (IsEOF()) |
| return pElement; |
| |
| uint8_t ch = m_pBuffer[m_dwIndex++]; |
| if (ch == '/') { |
| m_dwIndex++; |
| m_nOffset = m_nBufferOffset + static_cast<FX_FILESIZE>(m_dwIndex); |
| return pElement; |
| } |
| if (ch != '>') { |
| m_nOffset = m_nBufferOffset + static_cast<FX_FILESIZE>(m_dwIndex); |
| return nullptr; |
| } |
| SkipWhiteSpaces(); |
| if (IsEOF()) |
| return pElement; |
| |
| CFX_UTF8Decoder decoder; |
| CFX_WideTextBuf content; |
| bool bCDATA = false; |
| int32_t iState = 0; |
| do { |
| while (m_dwIndex < m_dwBufferSize) { |
| ch = m_pBuffer[m_dwIndex++]; |
| switch (iState) { |
| case 0: |
| if (ch == '<') { |
| iState = 1; |
| } else if (ch == '&') { |
| decoder.ClearStatus(); |
| decoder.AppendCodePoint(GetCharRef()); |
| } else { |
| decoder.Input(ch); |
| } |
| break; |
| case 1: |
| if (ch == '!') { |
| iState = 2; |
| } else if (ch == '?') { |
| SkipLiterals("?>"); |
| SkipWhiteSpaces(); |
| iState = 0; |
| } else if (ch == '/') { |
| ByteString space; |
| ByteString name; |
| GetName(&space, &name); |
| SkipWhiteSpaces(); |
| m_dwIndex++; |
| iState = 10; |
| } else { |
| content << decoder.GetResult(); |
| WideString dataStr = content.MakeString(); |
| if (!bCDATA) |
| dataStr.TrimRight(L" \t\r\n"); |
| |
| InsertContentSegment(bCDATA, dataStr.AsStringView(), |
| pElement.get()); |
| content.Clear(); |
| decoder.Clear(); |
| bCDATA = false; |
| iState = 0; |
| m_dwIndex--; |
| std::unique_ptr<CXML_Element> pSubElement = |
| ParseElementInternal(pElement.get(), true, nDepth + 1); |
| if (!pSubElement) |
| break; |
| |
| pElement->AppendChild(std::move(pSubElement)); |
| SkipWhiteSpaces(); |
| } |
| break; |
| case 2: |
| if (ch == '[') { |
| SkipLiterals("]]>"); |
| } else if (ch == '-') { |
| m_dwIndex++; |
| SkipLiterals("-->"); |
| } else { |
| SkipLiterals(">"); |
| } |
| decoder.Clear(); |
| SkipWhiteSpaces(); |
| iState = 0; |
| break; |
| } |
| if (iState == 10) { |
| break; |
| } |
| } |
| m_nOffset = m_nBufferOffset + static_cast<FX_FILESIZE>(m_dwIndex); |
| if (iState == 10 || m_dwIndex < m_dwBufferSize || IsEOF()) |
| break; |
| } while (ReadNextBlock()); |
| content << decoder.GetResult(); |
| WideString dataStr = content.MakeString(); |
| dataStr.TrimRight(L" \t\r\n"); |
| |
| InsertContentSegment(bCDATA, dataStr.AsStringView(), pElement.get()); |
| content.Clear(); |
| decoder.Clear(); |
| bCDATA = false; |
| return pElement; |
| } |
| |
| void CXML_Parser::InsertContentSegment(bool bCDATA, |
| const WideStringView& content, |
| CXML_Element* pElement) { |
| if (content.IsEmpty()) |
| return; |
| |
| pElement->AppendChild(pdfium::MakeUnique<CXML_Content>(bCDATA, content)); |
| } |
