added Poco::Net::escapeHTML() and Poco::Net::EscapeHTMLStream; PageCompiler now supports optional HTML-escaping of <%= %> expressions.

Net/Makefile

@@ -33,7 +33,8 @@ objects = \
 	WebSocket WebSocketImpl \
 	OAuth10Credentials OAuth20Credentials \
 	PollSet UDPClient UDPServerParams \
-	NTLMCredentials SSPINTLMCredentials HTTPNTLMCredentials
+	NTLMCredentials SSPINTLMCredentials HTTPNTLMCredentials \
+	EscapeHTMLStream
 
 target         = PocoNet
 target_version = $(LIBVERSION)

Net/include/Poco/Net/EscapeHTMLStream.h

@@ -0,0 +1,89 @@
+//
+// EscapeHTMLStream.h
+//
+// Library: Net
+// Package: HTTP
+// Module:  EscapeHTMLStream
+//
+// Definition of the EscapeHTMLStream class.
+//
+// Copyright (c) 1029, Applied Informatics Software Engineering GmbH.
+// and Contributors.
+//
+// SPDX-License-Identifier:	BSL-1.0
+//
+
+
+#ifndef Net_EscapeHTMLStream_INCLUDED
+#define Net_EscapeHTMLStream_INCLUDED
+
+
+#include "Poco/Net/Net.h"
+#include "Poco/UnbufferedStreamBuf.h"
+#include <ostream>
+
+
+namespace Poco {
+namespace Net {
+
+
+class Net_API EscapeHTMLStreamBuf: public Poco::UnbufferedStreamBuf
+	/// This stream buffer replaces all occurrences of special HTML
+	/// characters < > " & with their respective character
+	/// entities &lt; &gt; &quot; &amp;.
+{
+public:
+	EscapeHTMLStreamBuf(std::ostream& ostr);
+		/// Creates the EscapeHTMLStreamBuf and connects it
+		/// to the given output stream.
+
+	~EscapeHTMLStreamBuf();
+		/// Destroys the EscapeHTMLStreamBuf.
+
+protected:
+	int readFromDevice();
+	int writeToDevice(char c);
+
+private:
+	std::ostream* _pOstr;
+};
+
+
+class Net_API EscapeHTMLIOS: public virtual std::ios
+	/// The base class for EscapeHTMLOutputStream.
+{
+public:
+	EscapeHTMLIOS(std::ostream& ostr);
+		/// Creates the MailIOS and connects it
+		/// to the given output stream.
+
+	~EscapeHTMLIOS();
+		/// Destroys the stream.
+
+	EscapeHTMLStreamBuf* rdbuf();
+		/// Returns a pointer to the underlying streambuf.
+
+protected:
+	EscapeHTMLStreamBuf _buf;
+};
+
+
+class Net_API EscapeHTMLOutputStream: public EscapeHTMLIOS, public std::ostream
+	/// This stream replaces all occurrences of special HTML
+	/// characters < > " & with their respective character
+	/// entities &lt; &gt; &quot; &amp;.
+{
+public:
+	EscapeHTMLOutputStream(std::ostream& ostr);
+		/// Creates the MailOutputStream and connects it
+		/// to the given input stream.
+
+	~EscapeHTMLOutputStream();
+		/// Destroys the MailOutputStream.
+};
+
+
+} } // namespace Poco::Net
+
+
+#endif // Net_EscapeHTMLStream_INCLUDED

Net/include/Poco/Net/Net.h

@@ -81,7 +81,12 @@ void Net_API uninitializeNetwork();
 	/// (Windows only, no-op elsewhere)
 
 
-}} // namespace Poco::Net
+std::string htmlize(const std::string& str);
+	/// Returns a copy of html with reserved HTML
+	/// characters (<, >, ", &) propery escaped.
+
+
+} } // namespace Poco::Net
 
 
 //

Net/src/EscapeHTMLStream.cpp

@@ -0,0 +1,92 @@
+//
+// EscapeHTMLStream.cpp
+//
+// Library: Net
+// Package: Mail
+// Module:  EscapeHTMLStream
+//
+// Copyright (c) 2019, Applied Informatics Software Engineering GmbH.
+// and Contributors.
+//
+// SPDX-License-Identifier:	BSL-1.0
+//
+
+
+#include "Poco/Net/EscapeHTMLStream.h"
+
+
+namespace Poco {
+namespace Net {
+
+
+EscapeHTMLStreamBuf::EscapeHTMLStreamBuf(std::ostream& ostr):
+	_pOstr(&ostr)
+{
+}
+
+
+EscapeHTMLStreamBuf::~EscapeHTMLStreamBuf()
+{
+}
+
+
+int EscapeHTMLStreamBuf::readFromDevice()
+{
+	return std::char_traits<char>::eof();
+}
+
+
+int EscapeHTMLStreamBuf::writeToDevice(char c)
+{
+	switch (c)
+	{
+	case '<':
+		*_pOstr << "&lt;";
+		break;
+	case '>':
+		*_pOstr << "&gt;";
+		break;
+	case '"':
+		*_pOstr << "&quot;";
+		break;
+	case '&':
+		*_pOstr << "&amp;";
+		break;
+	default:
+		_pOstr->put(c);
+		break;
+	}
+	return charToInt(c);
+}
+
+
+EscapeHTMLIOS::EscapeHTMLIOS(std::ostream& ostr): _buf(ostr)
+{
+	poco_ios_init(&_buf);
+}
+
+
+EscapeHTMLIOS::~EscapeHTMLIOS()
+{
+}
+
+
+EscapeHTMLStreamBuf* EscapeHTMLIOS::rdbuf()
+{
+	return &_buf;
+}
+
+
+EscapeHTMLOutputStream::EscapeHTMLOutputStream(std::ostream& ostr):
+	EscapeHTMLIOS(ostr),
+	std::ostream(&_buf)
+{
+}
+
+
+EscapeHTMLOutputStream::~EscapeHTMLOutputStream()
+{
+}
+
+
+} } // namespace Poco::Net

Net/src/Net.cpp

@@ -42,6 +42,26 @@ void Net_API uninitializeNetwork()
 }
 
 
+std::string htmlize(const std::string& str)
+{
+	std::string::const_iterator it(str.begin());
+	std::string::const_iterator end(str.end());
+	std::string html;
+	for (; it != end; ++it)
+	{
+		switch (*it)
+		{
+		case '<': html += "&lt;"; break;
+		case '>': html += "&gt;"; break;
+		case '"': html += "&quot;"; break;
+		case '&': html += "&amp;"; break;
+		default:  html += *it; break;
+		}
+	}
+	return html;
+}
+
+
 } } // namespace Poco::Net
 
 

PageCompiler/doc/PageCompilerUserGuide.page

@@ -2,11 +2,16 @@ POCO C++ Server Page Compiler User Guide
 POCO PageCompiler
 
 !!!Introduction
+
 PageCompiler is a command line tool that translates HTML files (and other kinds of files) into
 C++ code, more precisely, subclasses of Poco::Net::HTTPRequestHandler.
 The source files can contain special directives that allow embedding of C++ code.
 The syntax of these directives is based on the syntax used for
-Java Server Pages (JSP) and Active Server Pages (ASP).
+Java Server Pages (JSP), Active Server Pages (ASP) or Embedded JavaScript (EJS) templating.
+
+This makes PageCompiler a C++-based HTML templating system. Since the translation of
+the HTML template into a C++ class happens at compile time, runtime performance is
+excellent.
 
 The following introductory sample shows the code for a simple page that displays the
 current date and time.
@@ -16,8 +21,8 @@ current date and time.
         #include "Poco/DateTime.h"
         #include "Poco/DateTimeFormatter.h"
         #include "Poco/DateTimeFormat.h"
-        
-        
+
+
         using Poco::DateTime;
         using Poco::DateTimeFormatter;
         using Poco::DateTimeFormat;
@@ -88,11 +93,24 @@ is required.
 
 The result of any valid C++ expression can be directly inserted into the page,
 provided the result can be written to an output stream. Note that the expression
-must not end with a semicolon.
+must not end with a semicolon. If HTML escape mode is enabled (new in POCO C++ Libraries
+release 1.10.0), HTML special characters < > " &
+will be replaced with corresponding character entities.
 
     <%= <expression> %>
 ----
 
+An alternative form of this is:
+
+    <%- <expression> %>
+----
+
+The latter form always copies the result of expression to the page verbatim,
+without escaping special HTML characters.
+
+Note that if escape mode is disabled, both forms are equivalent.
+
+
 !!Scriptlet
 
 Arbitrary C++ code fragments can be included using the Scriptlet directive.
@@ -214,6 +232,13 @@ constructors of the base class must also accept a single argument of the specifi
 
 Cannot be used together with <[context]>.
 
+!escape
+
+Enable (set to <[true]>) or disable (<[false]>, default) HTML escape mode.
+If enabled, the result of any expression enclosed in <%= %> tags will be HTML-escaped,
+which means HTML special characters < > " & will be replaced by corresponding
+character entities.
+
 !export
 
 Allows to specify a DLL import/export directive that is being added to the request
@@ -378,6 +403,10 @@ system (e.g., <[/help]> on Windows, <[--help]> or <[-h]> on Unix).
   * config-file (f): load configuration properties from a file
   * osp (O): add factory class definition/implementation for use with OSP
   * apache (A): add factory class definition/implementation and shared library manifest for use with ApacheConnector
+  * escape (e): make HTML-escape mode default for all pages
+
+Run the PageCompiler with the --help option to see all available options.
+
 
 !!Configuration Properties
 

PageCompiler/src/CodeWriter.cpp

@@ -55,6 +55,10 @@ void CodeWriter::writeImpl(std::ostream& ostr, const std::string& headerFileName
 {
 	ostr << "#include \"" << headerFileName << "\"\n";
 	writeImplIncludes(ostr);
+	if (_page.getBool("page.escape", false))
+	{
+		ostr << "#include \"Poco/Net/EscapeHTMLStream.h\"\n";
+	}
 	if (_page.getBool("page.compressed", false))
 	{
 		ostr << "#include \"Poco/DeflatingStream.h\"\n";
@@ -349,6 +353,7 @@ void CodeWriter::writeResponse(std::ostream& ostr)
 
 void CodeWriter::writeContent(std::ostream& ostr)
 {
+	bool escape(_page.getBool("page.escape", false));
 	bool buffered(_page.getBool("page.buffered", false));
 	bool chunked(_page.getBool("page.chunked", !buffered));
 	bool compressed(_page.getBool("page.compressed", false));
@@ -359,6 +364,10 @@ void CodeWriter::writeContent(std::ostream& ostr)
 	if (buffered)
 	{
 		ostr << "\tstd::stringstream responseStream;\n";
+		if (escape)
+		{
+			ostr << "\tPoco::Net::EscapeHTMLOutputStream _escapeStream(responseStream);\n";
+		}
 		ostr << cleanupHandler(_page.handler().str());
 		if (!chunked)
 		{
@@ -371,12 +380,20 @@ void CodeWriter::writeContent(std::ostream& ostr)
 		ostr << "\tstd::ostream& _responseStream = response.send();\n"
 		     << "\tPoco::DeflatingOutputStream _gzipStream(_responseStream, Poco::DeflatingStreamBuf::STREAM_GZIP, " << compressionLevel << ");\n"
 		     << "\tstd::ostream& responseStream = _compressResponse ? _gzipStream : _responseStream;\n";
+		if (escape)
+		{
+			ostr << "\tPoco::Net::EscapeHTMLOutputStream _escapeStream(responseStream);\n";
+		}
 		ostr << cleanupHandler(_page.handler().str());
 		ostr << "\tif (_compressResponse) _gzipStream.close();\n";
 	}
 	else
 	{
 		ostr << "\tstd::ostream& responseStream = response.send();\n";
+		if (escape)
+		{
+			ostr << "\tPoco::Net::EscapeHTMLOutputStream _escapeStream(responseStream);\n";
+		}
 		ostr << cleanupHandler(_page.handler().str());
 	}
 }

PageCompiler/src/PageCompiler.cpp

@@ -58,7 +58,8 @@ public:
 		_helpRequested(false),
 		_generateOSPCode(false),
 		_generateApacheCode(false),
-		_emitLineDirectives(true)
+		_emitLineDirectives(true),
+		_escape(false)
 	{
 	}
 
@@ -141,6 +142,12 @@ protected:
 				.required(false)
 				.repeatable(false)
 				.callback(OptionCallback<CompilerApp>(this, &CompilerApp::handleNoLine)));
+
+		options.addOption(
+			Option("escape", "e", "Escape special HTML characters (<, >, \", &) in <%= %> expressions.")
+				.required(false)
+				.repeatable(false)
+				.callback(OptionCallback<CompilerApp>(this, &CompilerApp::handleEscape)));
 	}
 
 	void handleHelp(const std::string& name, const std::string& value)
@@ -196,6 +203,11 @@ protected:
 		_emitLineDirectives = false;
 	}
 
+	void handleEscape(const std::string& name, const std::string& value)
+	{
+		_escape = true;
+	}
+
 	void displayHelp()
 	{
 		HelpFormatter helpFormatter(options());
@@ -329,6 +341,12 @@ protected:
 	void compile(const std::string& path)
 	{
 		Page page;
+
+		if (_escape)
+		{
+			page.set("page.escape", "true");
+		}
+
 		std::string clazz;
 		parse(path, page, clazz);
 		write(path, page, clazz);
@@ -365,6 +383,7 @@ private:
 	bool _generateOSPCode;
 	bool _generateApacheCode;
 	bool _emitLineDirectives;
+	bool _escape;
 	std::string _outputDir;
 	std::string _headerOutputDir;
 	std::string _headerPrefix;

PageCompiler/src/PageReader.cpp

@@ -21,6 +21,8 @@ const std::string PageReader::MARKUP_BEGIN("\tresponseStream << \"");
 const std::string PageReader::MARKUP_END("\";\n");
 const std::string PageReader::EXPR_BEGIN("\tresponseStream << (");
 const std::string PageReader::EXPR_END(");\n");
+const std::string PageReader::ESC_EXPR_BEGIN("\t_escapeStream << (");
+const std::string PageReader::ESC_EXPR_END(");\n");
 
 
 PageReader::PageReader(Page& page, const std::string& path):
@@ -128,6 +130,25 @@ void PageReader::parse(std::istream& pageStream)
 			else _page.handler() << token;
 		}
 		else if (token == "<%=")
+		{
+			if (state == STATE_MARKUP)
+			{
+				_page.handler() << MARKUP_END;
+				generateLineDirective(_page.handler());
+				if (escape())
+				{
+					_page.handler() << ESC_EXPR_BEGIN;
+					state = STATE_ESC_EXPR;
+				}
+				else
+				{
+					_page.handler() << EXPR_BEGIN;
+					state = STATE_EXPR;
+				}
+			}
+			else _page.handler() << token;
+		}
+		else if (token == "<%-")
 		{
 			if (state == STATE_MARKUP)
 			{
@@ -146,6 +167,12 @@ void PageReader::parse(std::istream& pageStream)
 				_page.handler() << MARKUP_BEGIN;
 				state = STATE_MARKUP;
 			}
+			else if (state == STATE_ESC_EXPR)
+			{
+				_page.handler() << ESC_EXPR_END;
+				_page.handler() << MARKUP_BEGIN;
+				state = STATE_MARKUP;
+			}
 			else if (state == STATE_ATTR)
 			{
 				parseAttributes();
@@ -201,6 +228,7 @@ void PageReader::parse(std::istream& pageStream)
 				_page.handler() << token;
 				break;
 			case STATE_EXPR:
+			case STATE_ESC_EXPR:
 				_page.handler() << token;
 				break;
 			case STATE_COMMENT:
@@ -387,3 +415,9 @@ void PageReader::generateLineDirective(std::ostream& ostr)
 		ostr << "\"\n";
 	}
 }
+
+
+bool PageReader::escape() const
+{
+	return _page.getBool("page.escape", false);
+}

PageCompiler/src/PageReader.h

@@ -35,7 +35,7 @@ public:
 	~PageReader();
 		/// Destroys the PageReader.
 
-	void parse(std::istream& pageStream);	
+	void parse(std::istream& pageStream);
 		/// Parses a HTML file containing server page directives,
 		/// converts the file into C++ code and adds the code
 		/// to the reader's Page object. Also parses page
@@ -53,6 +53,7 @@ protected:
 		STATE_PREHANDLER,
 		STATE_BLOCK,
 		STATE_EXPR,
+		STATE_ESC_EXPR,
 		STATE_COMMENT,
 		STATE_ATTR
 	};
@@ -61,16 +62,19 @@ protected:
 	static const std::string MARKUP_END;
 	static const std::string EXPR_BEGIN;
 	static const std::string EXPR_END;
+	static const std::string ESC_EXPR_BEGIN;
+	static const std::string ESC_EXPR_END;
 
 	void include(const std::string& path);
 	void parseAttributes();
 	void nextToken(std::istream& istr, std::string& token);
 	void handleAttribute(const std::string& name, const std::string& value);
 	std::string where() const;
+	bool escape() const;
 
 protected:
 	void generateLineDirective(std::ostream& ostr);
-	
+
 private:
 	PageReader();
 	PageReader(const PageReader&);