Reset entity start when clearing textchromium/3410

When we extract the text data we need to make sure we clear any
entity start positions, otherwise our entity extraction will go badly.

Bug: chromium:836661
Change-Id: Icbafdef912b1f5b495eafef426961c5df66cd3fd
Reviewed-on: https://pdfium-review.googlesource.com/31450
Commit-Queue: dsinclair <dsinclair@chromium.org>
Reviewed-by: Ryan Harrison <rharrison@chromium.org>
Reviewed-by: Henrique Nakashima <hnakashima@chromium.org>

2 files changed, 33 insertions, 0 deletions

diff --git a/core/fxcrt/xml/cfx_xmlparser.cpp b/core/fxcrt/xml/cfx_xmlparser.cpp
index 685655e815..eb79637095 100644
--- a/core/fxcrt/xml/cfx_xmlparser.cpp
+++ b/core/fxcrt/xml/cfx_xmlparser.cpp
@@ -586,6 +586,7 @@ void CFX_XMLParser::ParseTextChar(wchar_t character) {
 WideString CFX_XMLParser::GetTextData() {
   WideString ret(current_text_.data(), current_text_.size());
   current_text_.clear();
+  m_iEntityStart = -1;
   current_text_.reserve(kCurrentTextReserve);
   return ret;
 }
diff --git a/core/fxcrt/xml/cfx_xmlparser_unittest.cpp b/core/fxcrt/xml/cfx_xmlparser_unittest.cpp
index b5c9be57cb..73d6685dad 100644
--- a/core/fxcrt/xml/cfx_xmlparser_unittest.cpp
+++ b/core/fxcrt/xml/cfx_xmlparser_unittest.cpp
@@ -569,3 +569,35 @@ TEST(CFX_XMLParserTest, DoubleElementClose) {
   ASSERT_EQ(L"p", parser.GetTextData());
   ASSERT_EQ(FX_XmlSyntaxResult::Error, parser.DoSyntaxParse());
 }
+
+TEST(CFX_XMLParserTest, BadEntity) {
+  const char* input =
+      "<script>"
+      "Test &<p>; thing"
+      "</script>";
+
+  auto stream = MakeProxy(input);
+  auto root = pdfium::MakeUnique<CFX_XMLElement>(L"ROOT");
+
+  CFX_XMLTestParser parser(root.get(), stream);
+  ASSERT_EQ(FX_XmlSyntaxResult::ElementOpen, parser.DoSyntaxParse());
+  ASSERT_EQ(FX_XmlSyntaxResult::TagName, parser.DoSyntaxParse());
+  ASSERT_EQ(L"script", parser.GetTextData());
+
+  ASSERT_EQ(FX_XmlSyntaxResult::ElementBreak, parser.DoSyntaxParse());
+  ASSERT_EQ(FX_XmlSyntaxResult::Text, parser.DoSyntaxParse());
+  ASSERT_EQ(L"Test &", parser.GetTextData());
+
+  ASSERT_EQ(FX_XmlSyntaxResult::ElementOpen, parser.DoSyntaxParse());
+  ASSERT_EQ(FX_XmlSyntaxResult::TagName, parser.DoSyntaxParse());
+  ASSERT_EQ(L"p", parser.GetTextData());
+  ASSERT_EQ(FX_XmlSyntaxResult::ElementBreak, parser.DoSyntaxParse());
+
+  ASSERT_EQ(FX_XmlSyntaxResult::Text, parser.DoSyntaxParse());
+  ASSERT_EQ(L"; thing", parser.GetTextData());
+
+  ASSERT_EQ(FX_XmlSyntaxResult::ElementClose, parser.DoSyntaxParse());
+  ASSERT_EQ(L"script", parser.GetTextData());
+
+  ASSERT_EQ(FX_XmlSyntaxResult::EndOfString, parser.DoSyntaxParse());
+}