Guard against undefined shift.

xfa/fde/xml/fde_xml_imp_unittest.cpp

 // 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 "xfa/fde/xml/fde_xml_imp.h"
 
 #include "xfa/fgas/crt/fgas_stream.h"
 #include "testing/gtest/include/gtest/gtest.h"
 
 TEST(CFDE_XMLSyntaxParser, CData) {
@@ skipping 502 matching lines @@
   parser.GetAttributeValue(data);
   EXPECT_EQ(L"application/x-javascript", data);
 
   EXPECT_EQ(FDE_XmlSyntaxResult::ElementBreak, parser.DoSyntaxParse());
   EXPECT_EQ(FDE_XmlSyntaxResult::Text, parser.DoSyntaxParse());
   parser.GetTextData(data);
   EXPECT_EQ(L"\n  ", data);
 
   EXPECT_EQ(FDE_XmlSyntaxResult::EndOfString, parser.DoSyntaxParse());
 }
+
+TEST(CFDE_XMLSyntaxParser, Entities) {
+  const FX_WCHAR* input =
+      L"<script contentType=\"application/x-javascript\">"
+      L"&#66;"
+      L"&#x54;"
+      L"</script>";
+
+  // We * sizeof(FX_WCHAR) because we pass in the uint8_t, not the FX_WCHAR.
+  size_t len = FXSYS_wcslen(input) * sizeof(FX_WCHAR);
+  std::unique_ptr<IFX_Stream> stream(IFX_Stream::CreateStream(
+      reinterpret_cast<uint8_t*>(const_cast<FX_WCHAR*>(input)), len, 0));
+  CFDE_XMLSyntaxParser parser;
+  parser.Init(stream.get(), 256);
+
+  CFX_WideString data;
+
+  EXPECT_EQ(FDE_XmlSyntaxResult::ElementOpen, parser.DoSyntaxParse());
+  EXPECT_EQ(FDE_XmlSyntaxResult::TagName, parser.DoSyntaxParse());
+  parser.GetTagName(data);
+  EXPECT_EQ(L"script", data);
+
+  EXPECT_EQ(FDE_XmlSyntaxResult::AttriName, parser.DoSyntaxParse());
+  parser.GetAttributeName(data);
+  EXPECT_EQ(L"contentType", data);
+  EXPECT_EQ(FDE_XmlSyntaxResult::AttriValue, parser.DoSyntaxParse());
+  parser.GetAttributeValue(data);
+  EXPECT_EQ(L"application/x-javascript", data);
+
+  EXPECT_EQ(FDE_XmlSyntaxResult::ElementBreak, parser.DoSyntaxParse());
+  EXPECT_EQ(FDE_XmlSyntaxResult::Text, parser.DoSyntaxParse());
+  parser.GetTextData(data);
+  EXPECT_EQ(L"BT", data);
+
+  EXPECT_EQ(FDE_XmlSyntaxResult::ElementClose, parser.DoSyntaxParse());
+  parser.GetTagName(data);
+  EXPECT_EQ(L"script", data);
+
+  EXPECT_EQ(FDE_XmlSyntaxResult::EndOfString, parser.DoSyntaxParse());
+}
+
+TEST(CFDE_XMLSyntaxParser, EntityOverflowHex) {
+  const FX_WCHAR* input =
+      L"<script contentType=\"application/x-javascript\">"
+      L"&#xaDBDFFFFF;"
+      L"</script>";
+
+  // We * sizeof(FX_WCHAR) because we pass in the uint8_t, not the FX_WCHAR.
+  size_t len = FXSYS_wcslen(input) * sizeof(FX_WCHAR);
+  std::unique_ptr<IFX_Stream> stream(IFX_Stream::CreateStream(
+      reinterpret_cast<uint8_t*>(const_cast<FX_WCHAR*>(input)), len, 0));
+  CFDE_XMLSyntaxParser parser;
+  parser.Init(stream.get(), 256);
+
+  CFX_WideString data;
+
+  EXPECT_EQ(FDE_XmlSyntaxResult::ElementOpen, parser.DoSyntaxParse());
+  EXPECT_EQ(FDE_XmlSyntaxResult::TagName, parser.DoSyntaxParse());
+  parser.GetTagName(data);
+  EXPECT_EQ(L"script", data);
+
+  EXPECT_EQ(FDE_XmlSyntaxResult::AttriName, parser.DoSyntaxParse());
+  parser.GetAttributeName(data);
+  EXPECT_EQ(L"contentType", data);
+  EXPECT_EQ(FDE_XmlSyntaxResult::AttriValue, parser.DoSyntaxParse());
+  parser.GetAttributeValue(data);
+  EXPECT_EQ(L"application/x-javascript", data);
+
+  EXPECT_EQ(FDE_XmlSyntaxResult::ElementBreak, parser.DoSyntaxParse());
+  EXPECT_EQ(FDE_XmlSyntaxResult::Text, parser.DoSyntaxParse());
+  parser.GetTextData(data);
+
+// The Windows wchar_t is unsigned so this doesn't end up overflowing to
+// zero and has a different output value.
+#if _FXM_PLATFORM_ == _FXM_PLATFORM_WINDOWS_
+  EXPECT_EQ(L"\xffff", data);
+#else
+  EXPECT_EQ(L" ", data);
+#endif  // _FXM_PLATFORM_ == _FXM_PLATFORM_WINDOWS_
+
+  EXPECT_EQ(FDE_XmlSyntaxResult::ElementClose, parser.DoSyntaxParse());
+  parser.GetTagName(data);
+  EXPECT_EQ(L"script", data);
+
+  EXPECT_EQ(FDE_XmlSyntaxResult::EndOfString, parser.DoSyntaxParse());
+}
+
+TEST(CFDE_XMLSyntaxParser, EntityOverflowDecimal) {
+  const FX_WCHAR* input =
+      L"<script contentType=\"application/x-javascript\">"
+      L"&#2914910205;"
+      L"</script>";
+
+  // We * sizeof(FX_WCHAR) because we pass in the uint8_t, not the FX_WCHAR.
+  size_t len = FXSYS_wcslen(input) * sizeof(FX_WCHAR);
+  std::unique_ptr<IFX_Stream> stream(IFX_Stream::CreateStream(
+      reinterpret_cast<uint8_t*>(const_cast<FX_WCHAR*>(input)), len, 0));
+  CFDE_XMLSyntaxParser parser;
+  parser.Init(stream.get(), 256);
+
+  CFX_WideString data;
+
+  EXPECT_EQ(FDE_XmlSyntaxResult::ElementOpen, parser.DoSyntaxParse());
+  EXPECT_EQ(FDE_XmlSyntaxResult::TagName, parser.DoSyntaxParse());
+  parser.GetTagName(data);
+  EXPECT_EQ(L"script", data);
+
+  EXPECT_EQ(FDE_XmlSyntaxResult::AttriName, parser.DoSyntaxParse());
+  parser.GetAttributeName(data);
+  EXPECT_EQ(L"contentType", data);
+  EXPECT_EQ(FDE_XmlSyntaxResult::AttriValue, parser.DoSyntaxParse());
+  parser.GetAttributeValue(data);
+  EXPECT_EQ(L"application/x-javascript", data);
+
+  EXPECT_EQ(FDE_XmlSyntaxResult::ElementBreak, parser.DoSyntaxParse());
+  EXPECT_EQ(FDE_XmlSyntaxResult::Text, parser.DoSyntaxParse());
+  parser.GetTextData(data);
+
+// The Windows wchar_t is unsigned so this doesn't end up overflowing to
+// zero and has a different output value.
+#if _FXM_PLATFORM_ == _FXM_PLATFORM_WINDOWS_
+  EXPECT_EQ(L"\xfffd", data);
+#else
+  EXPECT_EQ(L" ", data);
+#endif  // _FXM_PLATFORM_ == _FXM_PLATFORM_WINDOWS_
+
+  EXPECT_EQ(FDE_XmlSyntaxResult::ElementClose, parser.DoSyntaxParse());
+  parser.GetTagName(data);
+  EXPECT_EQ(L"script", data);
+
+  EXPECT_EQ(FDE_XmlSyntaxResult::EndOfString, parser.DoSyntaxParse());
+}