diff options
author | Dan Sinclair <dsinclair@chromium.org> | 2017-04-19 09:19:57 -0400 |
---|---|---|
committer | Chromium commit bot <commit-bot@chromium.org> | 2017-04-19 13:33:07 +0000 |
commit | 0d86ecb08e1b2c204333b1f1f6b0b014e5b2971c (patch) | |
tree | f816429f8581c16a60773eb23385dc8e55729bac /core/fxcrt/xml/cfx_xmlsyntaxparser_unittest.cpp | |
parent | 3b71d26f092ebc86ca9177fbbe89d83caa67ae1b (diff) | |
download | pdfium-0d86ecb08e1b2c204333b1f1f6b0b014e5b2971c.tar.xz |
Move fde XML parser to core
This CL moves the XML parser from FDE into FXCRT and renames to CFX_
from CFDE_.
Change-Id: I21a9590bf74daf5517df630d7e7a5de89da99ea4
Reviewed-on: https://pdfium-review.googlesource.com/4312
Commit-Queue: dsinclair <dsinclair@chromium.org>
Reviewed-by: Tom Sepez <tsepez@chromium.org>
Reviewed-by: Nicolás Peña <npm@chromium.org>
Diffstat (limited to 'core/fxcrt/xml/cfx_xmlsyntaxparser_unittest.cpp')
-rw-r--r-- | core/fxcrt/xml/cfx_xmlsyntaxparser_unittest.cpp | 527 |
1 files changed, 527 insertions, 0 deletions
diff --git a/core/fxcrt/xml/cfx_xmlsyntaxparser_unittest.cpp b/core/fxcrt/xml/cfx_xmlsyntaxparser_unittest.cpp new file mode 100644 index 0000000000..60c0d34655 --- /dev/null +++ b/core/fxcrt/xml/cfx_xmlsyntaxparser_unittest.cpp @@ -0,0 +1,527 @@ +// 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. + +#include "core/fxcrt/xml/cfx_xmlsyntaxparser.h" + +#include <memory> + +#include "core/fxcrt/cfx_seekablestreamproxy.h" +#include "core/fxcrt/fx_codepage.h" +#include "testing/gtest/include/gtest/gtest.h" +#include "testing/test_support.h" + +class CFX_XMLSyntaxParserTest : public pdfium::FPDF_Test {}; + +TEST_F(CFX_XMLSyntaxParserTest, CData) { + const char* input = + "<script contentType=\"application/x-javascript\">\n" + " <![CDATA[\n" + " if (a[1] < 3)\n" + " app.alert(\"Tclams\");\n" + " ]]>\n" + "</script>"; + + const wchar_t* cdata = + L"\n" + L" if (a[1] < 3)\n" + L" app.alert(\"Tclams\");\n" + L" "; + + CFX_RetainPtr<CFX_SeekableStreamProxy> stream = + pdfium::MakeRetain<CFX_SeekableStreamProxy>( + reinterpret_cast<uint8_t*>(const_cast<char*>(input)), strlen(input)); + stream->SetCodePage(FX_CODEPAGE_UTF8); + + CFX_XMLSyntaxParser parser(stream); + ASSERT_EQ(FX_XmlSyntaxResult::ElementOpen, parser.DoSyntaxParse()); + ASSERT_EQ(FX_XmlSyntaxResult::TagName, parser.DoSyntaxParse()); + ASSERT_EQ(L"script", parser.GetTagName()); + + ASSERT_EQ(FX_XmlSyntaxResult::AttriName, parser.DoSyntaxParse()); + ASSERT_EQ(L"contentType", parser.GetAttributeName()); + ASSERT_EQ(FX_XmlSyntaxResult::AttriValue, parser.DoSyntaxParse()); + ASSERT_EQ(L"application/x-javascript", parser.GetAttributeValue()); + + ASSERT_EQ(FX_XmlSyntaxResult::ElementBreak, parser.DoSyntaxParse()); + ASSERT_EQ(FX_XmlSyntaxResult::Text, parser.DoSyntaxParse()); + ASSERT_EQ(L"\n ", parser.GetTextData()); + + ASSERT_EQ(FX_XmlSyntaxResult::CData, parser.DoSyntaxParse()); + ASSERT_EQ(cdata, parser.GetTextData()); + + ASSERT_EQ(FX_XmlSyntaxResult::Text, parser.DoSyntaxParse()); + ASSERT_EQ(L"\n", parser.GetTextData()); + + ASSERT_EQ(FX_XmlSyntaxResult::ElementClose, parser.DoSyntaxParse()); + ASSERT_EQ(L"script", parser.GetTagName()); + + ASSERT_EQ(FX_XmlSyntaxResult::EndOfString, parser.DoSyntaxParse()); +} + +TEST_F(CFX_XMLSyntaxParserTest, CDataWithInnerScript) { + const char* input = + "<script contentType=\"application/x-javascript\">\n" + " <![CDATA[\n" + " if (a[1] < 3)\n" + " app.alert(\"Tclams\");\n" + " </script>\n" + " ]]>\n" + "</script>"; + + const wchar_t* cdata = + L"\n" + L" if (a[1] < 3)\n" + L" app.alert(\"Tclams\");\n" + L" </script>\n" + L" "; + + CFX_RetainPtr<CFX_SeekableStreamProxy> stream = + pdfium::MakeRetain<CFX_SeekableStreamProxy>( + reinterpret_cast<uint8_t*>(const_cast<char*>(input)), strlen(input)); + stream->SetCodePage(FX_CODEPAGE_UTF8); + + CFX_XMLSyntaxParser parser(stream); + ASSERT_EQ(FX_XmlSyntaxResult::ElementOpen, parser.DoSyntaxParse()); + ASSERT_EQ(FX_XmlSyntaxResult::TagName, parser.DoSyntaxParse()); + ASSERT_EQ(L"script", parser.GetTagName()); + + ASSERT_EQ(FX_XmlSyntaxResult::AttriName, parser.DoSyntaxParse()); + ASSERT_EQ(L"contentType", parser.GetAttributeName()); + ASSERT_EQ(FX_XmlSyntaxResult::AttriValue, parser.DoSyntaxParse()); + ASSERT_EQ(L"application/x-javascript", parser.GetAttributeValue()); + + ASSERT_EQ(FX_XmlSyntaxResult::ElementBreak, parser.DoSyntaxParse()); + ASSERT_EQ(FX_XmlSyntaxResult::Text, parser.DoSyntaxParse()); + ASSERT_EQ(L"\n ", parser.GetTextData()); + + ASSERT_EQ(FX_XmlSyntaxResult::CData, parser.DoSyntaxParse()); + ASSERT_EQ(cdata, parser.GetTextData()); + + ASSERT_EQ(FX_XmlSyntaxResult::Text, parser.DoSyntaxParse()); + ASSERT_EQ(L"\n", parser.GetTextData()); + + ASSERT_EQ(FX_XmlSyntaxResult::ElementClose, parser.DoSyntaxParse()); + ASSERT_EQ(L"script", parser.GetTagName()); + + ASSERT_EQ(FX_XmlSyntaxResult::EndOfString, parser.DoSyntaxParse()); +} + +TEST_F(CFX_XMLSyntaxParserTest, ArrowBangArrow) { + const char* input = + "<script contentType=\"application/x-javascript\">\n" + " <!>\n" + "</script>"; + + CFX_RetainPtr<CFX_SeekableStreamProxy> stream = + pdfium::MakeRetain<CFX_SeekableStreamProxy>( + reinterpret_cast<uint8_t*>(const_cast<char*>(input)), strlen(input)); + stream->SetCodePage(FX_CODEPAGE_UTF8); + + CFX_XMLSyntaxParser parser(stream); + ASSERT_EQ(FX_XmlSyntaxResult::ElementOpen, parser.DoSyntaxParse()); + ASSERT_EQ(FX_XmlSyntaxResult::TagName, parser.DoSyntaxParse()); + + ASSERT_EQ(L"script", parser.GetTagName()); + + ASSERT_EQ(FX_XmlSyntaxResult::AttriName, parser.DoSyntaxParse()); + ASSERT_EQ(L"contentType", parser.GetAttributeName()); + ASSERT_EQ(FX_XmlSyntaxResult::AttriValue, parser.DoSyntaxParse()); + ASSERT_EQ(L"application/x-javascript", parser.GetAttributeValue()); + + ASSERT_EQ(FX_XmlSyntaxResult::ElementBreak, parser.DoSyntaxParse()); + ASSERT_EQ(FX_XmlSyntaxResult::Text, parser.DoSyntaxParse()); + ASSERT_EQ(L"\n ", parser.GetTextData()); + + ASSERT_EQ(FX_XmlSyntaxResult::Text, parser.DoSyntaxParse()); + ASSERT_EQ(L"\n", parser.GetTextData()); + + ASSERT_EQ(FX_XmlSyntaxResult::ElementClose, parser.DoSyntaxParse()); + ASSERT_EQ(L"script", parser.GetTagName()); + + ASSERT_EQ(FX_XmlSyntaxResult::EndOfString, parser.DoSyntaxParse()); +} + +TEST_F(CFX_XMLSyntaxParserTest, ArrowBangBracketArrow) { + const char* input = + "<script contentType=\"application/x-javascript\">\n" + " <![>\n" + "</script>"; + + CFX_RetainPtr<CFX_SeekableStreamProxy> stream = + pdfium::MakeRetain<CFX_SeekableStreamProxy>( + reinterpret_cast<uint8_t*>(const_cast<char*>(input)), strlen(input)); + stream->SetCodePage(FX_CODEPAGE_UTF8); + + CFX_XMLSyntaxParser parser(stream); + ASSERT_EQ(FX_XmlSyntaxResult::ElementOpen, parser.DoSyntaxParse()); + ASSERT_EQ(FX_XmlSyntaxResult::TagName, parser.DoSyntaxParse()); + ASSERT_EQ(L"script", parser.GetTagName()); + + ASSERT_EQ(FX_XmlSyntaxResult::AttriName, parser.DoSyntaxParse()); + ASSERT_EQ(L"contentType", parser.GetAttributeName()); + ASSERT_EQ(FX_XmlSyntaxResult::AttriValue, parser.DoSyntaxParse()); + ASSERT_EQ(L"application/x-javascript", parser.GetAttributeValue()); + + ASSERT_EQ(FX_XmlSyntaxResult::ElementBreak, parser.DoSyntaxParse()); + ASSERT_EQ(FX_XmlSyntaxResult::Text, parser.DoSyntaxParse()); + ASSERT_EQ(L"\n ", parser.GetTextData()); + + // Parser walks to end of input. + + ASSERT_EQ(FX_XmlSyntaxResult::EndOfString, parser.DoSyntaxParse()); +} + +TEST_F(CFX_XMLSyntaxParserTest, IncompleteCData) { + const char* input = + "<script contentType=\"application/x-javascript\">\n" + " <![CDATA>\n" + "</script>"; + + CFX_RetainPtr<CFX_SeekableStreamProxy> stream = + pdfium::MakeRetain<CFX_SeekableStreamProxy>( + reinterpret_cast<uint8_t*>(const_cast<char*>(input)), strlen(input)); + stream->SetCodePage(FX_CODEPAGE_UTF8); + + CFX_XMLSyntaxParser parser(stream); + ASSERT_EQ(FX_XmlSyntaxResult::ElementOpen, parser.DoSyntaxParse()); + ASSERT_EQ(FX_XmlSyntaxResult::TagName, parser.DoSyntaxParse()); + ASSERT_EQ(L"script", parser.GetTagName()); + + ASSERT_EQ(FX_XmlSyntaxResult::AttriName, parser.DoSyntaxParse()); + ASSERT_EQ(L"contentType", parser.GetAttributeName()); + ASSERT_EQ(FX_XmlSyntaxResult::AttriValue, parser.DoSyntaxParse()); + ASSERT_EQ(L"application/x-javascript", parser.GetAttributeValue()); + + ASSERT_EQ(FX_XmlSyntaxResult::ElementBreak, parser.DoSyntaxParse()); + ASSERT_EQ(FX_XmlSyntaxResult::Text, parser.DoSyntaxParse()); + ASSERT_EQ(L"\n ", parser.GetTextData()); + + // Parser walks to end of input. + + ASSERT_EQ(FX_XmlSyntaxResult::EndOfString, parser.DoSyntaxParse()); +} + +TEST_F(CFX_XMLSyntaxParserTest, UnClosedCData) { + const char* input = + "<script contentType=\"application/x-javascript\">\n" + " <![CDATA[\n" + "</script>"; + + CFX_RetainPtr<CFX_SeekableStreamProxy> stream = + pdfium::MakeRetain<CFX_SeekableStreamProxy>( + reinterpret_cast<uint8_t*>(const_cast<char*>(input)), strlen(input)); + stream->SetCodePage(FX_CODEPAGE_UTF8); + + CFX_XMLSyntaxParser parser(stream); + ASSERT_EQ(FX_XmlSyntaxResult::ElementOpen, parser.DoSyntaxParse()); + ASSERT_EQ(FX_XmlSyntaxResult::TagName, parser.DoSyntaxParse()); + ASSERT_EQ(L"script", parser.GetTagName()); + + ASSERT_EQ(FX_XmlSyntaxResult::AttriName, parser.DoSyntaxParse()); + ASSERT_EQ(L"contentType", parser.GetAttributeName()); + ASSERT_EQ(FX_XmlSyntaxResult::AttriValue, parser.DoSyntaxParse()); + ASSERT_EQ(L"application/x-javascript", parser.GetAttributeValue()); + + ASSERT_EQ(FX_XmlSyntaxResult::ElementBreak, parser.DoSyntaxParse()); + ASSERT_EQ(FX_XmlSyntaxResult::Text, parser.DoSyntaxParse()); + ASSERT_EQ(L"\n ", parser.GetTextData()); + + // Parser walks to end of input. + + ASSERT_EQ(FX_XmlSyntaxResult::EndOfString, parser.DoSyntaxParse()); +} + +TEST_F(CFX_XMLSyntaxParserTest, EmptyCData) { + const char* input = + "<script contentType=\"application/x-javascript\">\n" + " <![CDATA[]]>\n" + "</script>"; + + CFX_RetainPtr<CFX_SeekableStreamProxy> stream = + pdfium::MakeRetain<CFX_SeekableStreamProxy>( + reinterpret_cast<uint8_t*>(const_cast<char*>(input)), strlen(input)); + stream->SetCodePage(FX_CODEPAGE_UTF8); + + CFX_XMLSyntaxParser parser(stream); + ASSERT_EQ(FX_XmlSyntaxResult::ElementOpen, parser.DoSyntaxParse()); + ASSERT_EQ(FX_XmlSyntaxResult::TagName, parser.DoSyntaxParse()); + ASSERT_EQ(L"script", parser.GetTagName()); + + ASSERT_EQ(FX_XmlSyntaxResult::AttriName, parser.DoSyntaxParse()); + ASSERT_EQ(L"contentType", parser.GetAttributeName()); + ASSERT_EQ(FX_XmlSyntaxResult::AttriValue, parser.DoSyntaxParse()); + ASSERT_EQ(L"application/x-javascript", parser.GetAttributeValue()); + + ASSERT_EQ(FX_XmlSyntaxResult::ElementBreak, parser.DoSyntaxParse()); + ASSERT_EQ(FX_XmlSyntaxResult::Text, parser.DoSyntaxParse()); + ASSERT_EQ(L"\n ", parser.GetTextData()); + + ASSERT_EQ(FX_XmlSyntaxResult::CData, parser.DoSyntaxParse()); + ASSERT_EQ(L"", parser.GetTextData()); + + ASSERT_EQ(FX_XmlSyntaxResult::Text, parser.DoSyntaxParse()); + ASSERT_EQ(L"\n", parser.GetTextData()); + + ASSERT_EQ(FX_XmlSyntaxResult::ElementClose, parser.DoSyntaxParse()); + ASSERT_EQ(L"script", parser.GetTagName()); + + ASSERT_EQ(FX_XmlSyntaxResult::EndOfString, parser.DoSyntaxParse()); +} + +TEST_F(CFX_XMLSyntaxParserTest, Comment) { + const char* input = + "<script contentType=\"application/x-javascript\">\n" + " <!-- A Comment -->\n" + "</script>"; + + CFX_RetainPtr<CFX_SeekableStreamProxy> stream = + pdfium::MakeRetain<CFX_SeekableStreamProxy>( + reinterpret_cast<uint8_t*>(const_cast<char*>(input)), strlen(input)); + stream->SetCodePage(FX_CODEPAGE_UTF8); + + CFX_XMLSyntaxParser parser(stream); + ASSERT_EQ(FX_XmlSyntaxResult::ElementOpen, parser.DoSyntaxParse()); + ASSERT_EQ(FX_XmlSyntaxResult::TagName, parser.DoSyntaxParse()); + ASSERT_EQ(L"script", parser.GetTagName()); + + ASSERT_EQ(FX_XmlSyntaxResult::AttriName, parser.DoSyntaxParse()); + ASSERT_EQ(L"contentType", parser.GetAttributeName()); + ASSERT_EQ(FX_XmlSyntaxResult::AttriValue, parser.DoSyntaxParse()); + ASSERT_EQ(L"application/x-javascript", parser.GetAttributeValue()); + + ASSERT_EQ(FX_XmlSyntaxResult::ElementBreak, parser.DoSyntaxParse()); + ASSERT_EQ(FX_XmlSyntaxResult::Text, parser.DoSyntaxParse()); + ASSERT_EQ(L"\n ", parser.GetTextData()); + + ASSERT_EQ(FX_XmlSyntaxResult::Text, parser.DoSyntaxParse()); + ASSERT_EQ(L"\n", parser.GetTextData()); + + ASSERT_EQ(FX_XmlSyntaxResult::ElementClose, parser.DoSyntaxParse()); + ASSERT_EQ(L"script", parser.GetTagName()); + + ASSERT_EQ(FX_XmlSyntaxResult::EndOfString, parser.DoSyntaxParse()); +} + +TEST_F(CFX_XMLSyntaxParserTest, IncorrectCommentStart) { + const char* input = + "<script contentType=\"application/x-javascript\">\n" + " <!- A Comment -->\n" + "</script>"; + + CFX_RetainPtr<CFX_SeekableStreamProxy> stream = + pdfium::MakeRetain<CFX_SeekableStreamProxy>( + reinterpret_cast<uint8_t*>(const_cast<char*>(input)), strlen(input)); + stream->SetCodePage(FX_CODEPAGE_UTF8); + + CFX_XMLSyntaxParser parser(stream); + ASSERT_EQ(FX_XmlSyntaxResult::ElementOpen, parser.DoSyntaxParse()); + ASSERT_EQ(FX_XmlSyntaxResult::TagName, parser.DoSyntaxParse()); + ASSERT_EQ(L"script", parser.GetTagName()); + + ASSERT_EQ(FX_XmlSyntaxResult::AttriName, parser.DoSyntaxParse()); + ASSERT_EQ(L"contentType", parser.GetAttributeName()); + ASSERT_EQ(FX_XmlSyntaxResult::AttriValue, parser.DoSyntaxParse()); + ASSERT_EQ(L"application/x-javascript", parser.GetAttributeValue()); + + ASSERT_EQ(FX_XmlSyntaxResult::ElementBreak, parser.DoSyntaxParse()); + ASSERT_EQ(FX_XmlSyntaxResult::Text, parser.DoSyntaxParse()); + ASSERT_EQ(L"\n ", parser.GetTextData()); + + ASSERT_EQ(FX_XmlSyntaxResult::Text, parser.DoSyntaxParse()); + ASSERT_EQ(L"\n", parser.GetTextData()); + + ASSERT_EQ(FX_XmlSyntaxResult::ElementClose, parser.DoSyntaxParse()); + ASSERT_EQ(L"script", parser.GetTagName()); + + ASSERT_EQ(FX_XmlSyntaxResult::EndOfString, parser.DoSyntaxParse()); +} + +TEST_F(CFX_XMLSyntaxParserTest, CommentEmpty) { + const char* input = + "<script contentType=\"application/x-javascript\">\n" + " <!---->\n" + "</script>"; + + CFX_RetainPtr<CFX_SeekableStreamProxy> stream = + pdfium::MakeRetain<CFX_SeekableStreamProxy>( + reinterpret_cast<uint8_t*>(const_cast<char*>(input)), strlen(input)); + stream->SetCodePage(FX_CODEPAGE_UTF8); + + CFX_XMLSyntaxParser parser(stream); + ASSERT_EQ(FX_XmlSyntaxResult::ElementOpen, parser.DoSyntaxParse()); + ASSERT_EQ(FX_XmlSyntaxResult::TagName, parser.DoSyntaxParse()); + ASSERT_EQ(L"script", parser.GetTagName()); + + ASSERT_EQ(FX_XmlSyntaxResult::AttriName, parser.DoSyntaxParse()); + ASSERT_EQ(L"contentType", parser.GetAttributeName()); + ASSERT_EQ(FX_XmlSyntaxResult::AttriValue, parser.DoSyntaxParse()); + ASSERT_EQ(L"application/x-javascript", parser.GetAttributeValue()); + + ASSERT_EQ(FX_XmlSyntaxResult::ElementBreak, parser.DoSyntaxParse()); + ASSERT_EQ(FX_XmlSyntaxResult::Text, parser.DoSyntaxParse()); + ASSERT_EQ(L"\n ", parser.GetTextData()); + + ASSERT_EQ(FX_XmlSyntaxResult::Text, parser.DoSyntaxParse()); + ASSERT_EQ(L"\n", parser.GetTextData()); + + ASSERT_EQ(FX_XmlSyntaxResult::ElementClose, parser.DoSyntaxParse()); + ASSERT_EQ(L"script", parser.GetTagName()); + + ASSERT_EQ(FX_XmlSyntaxResult::EndOfString, parser.DoSyntaxParse()); +} + +TEST_F(CFX_XMLSyntaxParserTest, CommentThreeDash) { + const char* input = + "<script contentType=\"application/x-javascript\">\n" + " <!--->\n" + "</script>"; + + CFX_RetainPtr<CFX_SeekableStreamProxy> stream = + pdfium::MakeRetain<CFX_SeekableStreamProxy>( + reinterpret_cast<uint8_t*>(const_cast<char*>(input)), strlen(input)); + stream->SetCodePage(FX_CODEPAGE_UTF8); + + CFX_XMLSyntaxParser parser(stream); + ASSERT_EQ(FX_XmlSyntaxResult::ElementOpen, parser.DoSyntaxParse()); + ASSERT_EQ(FX_XmlSyntaxResult::TagName, parser.DoSyntaxParse()); + ASSERT_EQ(L"script", parser.GetTagName()); + + ASSERT_EQ(FX_XmlSyntaxResult::AttriName, parser.DoSyntaxParse()); + ASSERT_EQ(L"contentType", parser.GetAttributeName()); + ASSERT_EQ(FX_XmlSyntaxResult::AttriValue, parser.DoSyntaxParse()); + ASSERT_EQ(L"application/x-javascript", parser.GetAttributeValue()); + + ASSERT_EQ(FX_XmlSyntaxResult::ElementBreak, parser.DoSyntaxParse()); + ASSERT_EQ(FX_XmlSyntaxResult::Text, parser.DoSyntaxParse()); + ASSERT_EQ(L"\n ", parser.GetTextData()); + + ASSERT_EQ(FX_XmlSyntaxResult::EndOfString, parser.DoSyntaxParse()); +} + +TEST_F(CFX_XMLSyntaxParserTest, CommentTwoDash) { + const char* input = + "<script contentType=\"application/x-javascript\">\n" + " <!-->\n" + "</script>"; + + CFX_RetainPtr<CFX_SeekableStreamProxy> stream = + pdfium::MakeRetain<CFX_SeekableStreamProxy>( + reinterpret_cast<uint8_t*>(const_cast<char*>(input)), strlen(input)); + stream->SetCodePage(FX_CODEPAGE_UTF8); + + CFX_XMLSyntaxParser parser(stream); + ASSERT_EQ(FX_XmlSyntaxResult::ElementOpen, parser.DoSyntaxParse()); + ASSERT_EQ(FX_XmlSyntaxResult::TagName, parser.DoSyntaxParse()); + ASSERT_EQ(L"script", parser.GetTagName()); + + ASSERT_EQ(FX_XmlSyntaxResult::AttriName, parser.DoSyntaxParse()); + ASSERT_EQ(L"contentType", parser.GetAttributeName()); + ASSERT_EQ(FX_XmlSyntaxResult::AttriValue, parser.DoSyntaxParse()); + ASSERT_EQ(L"application/x-javascript", parser.GetAttributeValue()); + + ASSERT_EQ(FX_XmlSyntaxResult::ElementBreak, parser.DoSyntaxParse()); + ASSERT_EQ(FX_XmlSyntaxResult::Text, parser.DoSyntaxParse()); + ASSERT_EQ(L"\n ", parser.GetTextData()); + + ASSERT_EQ(FX_XmlSyntaxResult::EndOfString, parser.DoSyntaxParse()); +} + +TEST_F(CFX_XMLSyntaxParserTest, Entities) { + const char* input = + "<script contentType=\"application/x-javascript\">" + "B" + "T" + "H" + "ꭈ" + "�" + "</script>"; + + CFX_RetainPtr<CFX_SeekableStreamProxy> stream = + pdfium::MakeRetain<CFX_SeekableStreamProxy>( + reinterpret_cast<uint8_t*>(const_cast<char*>(input)), strlen(input)); + stream->SetCodePage(FX_CODEPAGE_UTF8); + + CFX_XMLSyntaxParser parser(stream); + ASSERT_EQ(FX_XmlSyntaxResult::ElementOpen, parser.DoSyntaxParse()); + ASSERT_EQ(FX_XmlSyntaxResult::TagName, parser.DoSyntaxParse()); + ASSERT_EQ(L"script", parser.GetTagName()); + + ASSERT_EQ(FX_XmlSyntaxResult::AttriName, parser.DoSyntaxParse()); + ASSERT_EQ(L"contentType", parser.GetAttributeName()); + ASSERT_EQ(FX_XmlSyntaxResult::AttriValue, parser.DoSyntaxParse()); + ASSERT_EQ(L"application/x-javascript", parser.GetAttributeValue()); + + ASSERT_EQ(FX_XmlSyntaxResult::ElementBreak, parser.DoSyntaxParse()); + ASSERT_EQ(FX_XmlSyntaxResult::Text, parser.DoSyntaxParse()); + ASSERT_EQ(L"BTH\xab48", parser.GetTextData()); + + ASSERT_EQ(FX_XmlSyntaxResult::ElementClose, parser.DoSyntaxParse()); + ASSERT_EQ(L"script", parser.GetTagName()); + + ASSERT_EQ(FX_XmlSyntaxResult::EndOfString, parser.DoSyntaxParse()); +} + +TEST_F(CFX_XMLSyntaxParserTest, EntityOverflowHex) { + const char* input = + "<script contentType=\"application/x-javascript\">" + "�" + "�" + "</script>"; + + CFX_RetainPtr<CFX_SeekableStreamProxy> stream = + pdfium::MakeRetain<CFX_SeekableStreamProxy>( + reinterpret_cast<uint8_t*>(const_cast<char*>(input)), strlen(input)); + stream->SetCodePage(FX_CODEPAGE_UTF8); + + CFX_XMLSyntaxParser parser(stream); + ASSERT_EQ(FX_XmlSyntaxResult::ElementOpen, parser.DoSyntaxParse()); + ASSERT_EQ(FX_XmlSyntaxResult::TagName, parser.DoSyntaxParse()); + ASSERT_EQ(L"script", parser.GetTagName()); + + ASSERT_EQ(FX_XmlSyntaxResult::AttriName, parser.DoSyntaxParse()); + ASSERT_EQ(L"contentType", parser.GetAttributeName()); + ASSERT_EQ(FX_XmlSyntaxResult::AttriValue, parser.DoSyntaxParse()); + ASSERT_EQ(L"application/x-javascript", parser.GetAttributeValue()); + + ASSERT_EQ(FX_XmlSyntaxResult::ElementBreak, parser.DoSyntaxParse()); + ASSERT_EQ(FX_XmlSyntaxResult::Text, parser.DoSyntaxParse()); + ASSERT_EQ(L" ", parser.GetTextData()); + + ASSERT_EQ(FX_XmlSyntaxResult::ElementClose, parser.DoSyntaxParse()); + ASSERT_EQ(L"script", parser.GetTagName()); + + ASSERT_EQ(FX_XmlSyntaxResult::EndOfString, parser.DoSyntaxParse()); +} + +TEST_F(CFX_XMLSyntaxParserTest, EntityOverflowDecimal) { + const char* input = + "<script contentType=\"application/x-javascript\">" + "�" + "�" + "</script>"; + + CFX_RetainPtr<CFX_SeekableStreamProxy> stream = + pdfium::MakeRetain<CFX_SeekableStreamProxy>( + reinterpret_cast<uint8_t*>(const_cast<char*>(input)), strlen(input)); + stream->SetCodePage(FX_CODEPAGE_UTF8); + + CFX_XMLSyntaxParser parser(stream); + ASSERT_EQ(FX_XmlSyntaxResult::ElementOpen, parser.DoSyntaxParse()); + ASSERT_EQ(FX_XmlSyntaxResult::TagName, parser.DoSyntaxParse()); + ASSERT_EQ(L"script", parser.GetTagName()); + + ASSERT_EQ(FX_XmlSyntaxResult::AttriName, parser.DoSyntaxParse()); + ASSERT_EQ(L"contentType", parser.GetAttributeName()); + ASSERT_EQ(FX_XmlSyntaxResult::AttriValue, parser.DoSyntaxParse()); + ASSERT_EQ(L"application/x-javascript", parser.GetAttributeValue()); + + ASSERT_EQ(FX_XmlSyntaxResult::ElementBreak, parser.DoSyntaxParse()); + ASSERT_EQ(FX_XmlSyntaxResult::Text, parser.DoSyntaxParse()); + ASSERT_EQ(L" ", parser.GetTextData()); + + ASSERT_EQ(FX_XmlSyntaxResult::ElementClose, parser.DoSyntaxParse()); + ASSERT_EQ(L"script", parser.GetTagName()); + + ASSERT_EQ(FX_XmlSyntaxResult::EndOfString, parser.DoSyntaxParse()); +} |