Fixed getLocaleCharset() in Windows platform handler, when MLang is not present.

SConstruct

@@ -1124,6 +1124,8 @@ AM_CONFIG_HEADER([config.h])
 
 AM_MAINTAINER_MODE
 
+VMIME_ADDITIONAL_DEFINES=""
+
 
 #
 # Check compilers, processors, etc
@@ -1429,6 +1431,16 @@ x*)
 	;;
 esac
 
+
+#
+# Detect some platform-specific stuff
+#
+
+if test "x$VMIME_DETECT_PLATFORM" = "xwindows"; then
+	AC_CHECK_HEADER(mlang.h, [VMIME_ADDITIONAL_DEFINES="$VMIME_ADDITIONAL_DEFINES HAVE_MLANG_H"])
+fi
+
+
 """)
 
 
@@ -1615,10 +1627,21 @@ typedef unsigned ${VMIME_TYPE_INT32} vmime_uint32;
 			+ " $VMIME_BUILTIN_PLATFORM_" + p + " \n")
 
 	configure_in.write("""
-
-#endif // VMIME_CONFIG_HPP_INCLUDED
 " > vmime/config.hpp
 
+
+# Additional defines
+echo "// Additional defines" >> vmime/config.hpp
+
+for d in $VMIME_ADDITIONAL_DEFINES ; do
+	echo "#define VMIME_$d 1" >> vmime/config.hpp
+done
+
+
+echo "" >> vmime/config.hpp
+echo "#endif // VMIME_CONFIG_HPP_INCLUDED" >> vmime/config.hpp
+
+
 AC_MSG_RESULT([
 +=================+
 |  CONFIGURATION  |

src/platforms/windows/windowsHandler.cpp

@@ -18,11 +18,17 @@
 //
 
 #include "vmime/platforms/windows/windowsHandler.hpp"
+#include "vmime/config.hpp"
 
 #include <time.h>
 #include <locale.h>
 #include <process.h>
-#include <mlang.h>
+#include <windows.h>  // for winnls.h
+
+#ifdef VMIME_HAVE_MLANG_H
+#   include <mlang.h>
+#endif
+
 
 namespace vmime {
 namespace platforms {
@@ -62,7 +68,7 @@ const vmime::datetime windowsHandler::getCurrentLocalTime() const
 	const time_t t(::time(NULL));
 
 	// Get the local time
-#ifdef _REENTRANT
+#if defined(_REENTRANT) && defined(localtime_r)
 	tm local;
 	::localtime_r(&t, &local);
 #else
@@ -70,7 +76,7 @@ const vmime::datetime windowsHandler::getCurrentLocalTime() const
 #endif
 
 	// Get the UTC time
-#ifdef _REENTRANT
+#if defined(_REENTRANT) && defined(gmtime_r)
 	tm gmt;
 	::gmtime_r(&t, &gmt);
 #else
@@ -94,8 +100,9 @@ const vmime::datetime windowsHandler::getCurrentLocalTime() const
 
 const vmime::charset windowsHandler::getLocaleCharset() const
 {
+#ifdef VMIME_HAVE_MLANG_H
 	char szCharset[256];
-	
+
 	CoInitialize(NULL);
 	{
 		IMultiLanguage* pMultiLanguage;
@@ -111,15 +118,66 @@ const vmime::charset windowsHandler::getLocaleCharset() const
 		pMultiLanguage->GetCodePageInfo(codePage, &cpInfo);
 
 		int nLengthW = lstrlenW(cpInfo.wszBodyCharset) + 1;
-		
+
 		WideCharToMultiByte(codePage, 0, cpInfo.wszBodyCharset, nLengthW, szCharset, sizeof(szCharset), NULL, NULL );
-		
+
 		pMultiLanguage->Release();
 
-	}  
+	}
 	CoUninitialize();
 
 	return vmime::charset(szCharset);
+#else // VMIME_HAVE_MLANG_H
+	vmime::string ch = vmime::charsets::ISO8859_1; // default
+
+	switch (GetACP())
+	{
+	case 437: ch = vmime::charsets::CP_437; break;
+	case 737: ch = vmime::charsets::CP_737; break;
+	case 775: ch = vmime::charsets::CP_775; break;
+	case 850: ch = vmime::charsets::CP_850; break;
+	case 852: ch = vmime::charsets::CP_852; break;
+	case 853: ch = vmime::charsets::CP_853; break;
+	case 855: ch = vmime::charsets::CP_855; break;
+	case 857: ch = vmime::charsets::CP_857; break;
+	case 858: ch = vmime::charsets::CP_858; break;
+	case 860: ch = vmime::charsets::CP_860; break;
+	case 861: ch = vmime::charsets::CP_861; break;
+	case 862: ch = vmime::charsets::CP_862; break;
+	case 863: ch = vmime::charsets::CP_863; break;
+	case 864: ch = vmime::charsets::CP_864; break;
+	case 865: ch = vmime::charsets::CP_865; break;
+	case 866: ch = vmime::charsets::CP_866; break;
+	case 869: ch = vmime::charsets::CP_869; break;
+	case 874: ch = vmime::charsets::CP_874; break;
+
+	case 1125: ch = vmime::charsets::CP_1125; break;
+	case 1250: ch = vmime::charsets::CP_1250; break;
+	case 1251: ch = vmime::charsets::CP_1251; break;
+	case 1252: ch = vmime::charsets::CP_1252; break;
+	case 1253: ch = vmime::charsets::CP_1253; break;
+	case 1254: ch = vmime::charsets::CP_1254; break;
+	case 1255: ch = vmime::charsets::CP_1255; break;
+	case 1256: ch = vmime::charsets::CP_1256; break;
+	case 1257: ch = vmime::charsets::CP_1257; break;
+
+	case 28591: ch = vmime::charsets::ISO8859_1; break;
+	case 28592: ch = vmime::charsets::ISO8859_2; break;
+	case 28593: ch = vmime::charsets::ISO8859_3; break;
+	case 28594: ch = vmime::charsets::ISO8859_4; break;
+	case 28595: ch = vmime::charsets::ISO8859_5; break;
+	case 28596: ch = vmime::charsets::ISO8859_6; break;
+	case 28597: ch = vmime::charsets::ISO8859_7; break;
+	case 28598: ch = vmime::charsets::ISO8859_8; break;
+	case 28599: ch = vmime::charsets::ISO8859_9; break;
+	case 28605: ch = vmime::charsets::ISO8859_15; break;
+
+	case 65000: ch = vmime::charsets::UTF_7; break;
+	case 65001: ch = vmime::charsets::UTF_8; break;
+	}
+
+	return (vmime::charset(ch));
+#endif
 }
 
 

src/utility/url.cpp

@@ -181,12 +181,12 @@ void url::parse(const string& str)
 
 	if (colonPos == string::npos)
 	{
-		host = hostPart;
+		host = utility::stringUtils::trim(hostPart);
 	}
 	else
 	{
-		host = string(hostPart.begin(), hostPart.begin() + colonPos);
+		host = utility::stringUtils::trim(string(hostPart.begin(), hostPart.begin() + colonPos));
-		port = string(hostPart.begin() + colonPos + 1, hostPart.end());
+		port = utility::stringUtils::trim(string(hostPart.begin() + colonPos + 1, hostPart.end()));
 	}
 
 	// Path
@@ -198,8 +198,12 @@ void url::parse(const string& str)
 	// Some sanity check
 	if (proto.empty())
 		throw exceptions::malformed_url("No protocol specified");
-	else if (host.empty() && path.empty())  // Accept empty host (eg. "file:///home/vincent/mydoc")
+	else if (host.empty())
-		throw exceptions::malformed_url("No host specified");
+	{
+		// Accept empty host (eg. "file:///home/vincent/mydoc")
+		if (proto != PROTOCOL_FILE)
+			throw exceptions::malformed_url("No host specified");
+	}
 
 	bool onlyDigit = true;
 
@@ -217,6 +221,9 @@ void url::parse(const string& str)
 
 	iss >> portNum;
 
+	if (portNum == 0)
+		portNum = UNSPECIFIED_PORT;
+
 	// Now, save URL parts
 	m_protocol = proto;
 

src/utility/urlUtils.cpp

@@ -34,17 +34,18 @@ const string urlUtils::encode(const string& s)
 	{
 		const char_t c = *it;
 
-		if (parserHelpers::isPrint(c) && c != '%')
+		if (parserHelpers::isPrint(c) && !parserHelpers::isSpace(c) && c != '%')
 		{
 			result += c;
 		}
 		else
 		{
 			char hex[4];
+			const unsigned char k = static_cast <unsigned char>(c);
 
 			hex[0] = '%';
-			hex[1] = c / 16;
+			hex[1] = "0123456789ABCDEF"[k / 16];
-			hex[2] = c % 16;
+			hex[2] = "0123456789ABCDEF"[k % 16];
 			hex[3] = 0;
 
 			result += hex;
@@ -71,7 +72,35 @@ const string urlUtils::decode(const string& s)
 			const char_t p = (++it != s.end() ? *it : 0);
 			const char_t q = (++it != s.end() ? *it : 0);
 
-			result += static_cast <string::value_type>(p * 16 + q);
+			unsigned char r = 0;
+
+			switch (p)
+			{
+			case 0: r = '?'; break;
+			case 'a': case 'A': r = 10; break;
+			case 'b': case 'B': r = 11; break;
+			case 'c': case 'C': r = 12; break;
+			case 'd': case 'D': r = 13; break;
+			case 'e': case 'E': r = 14; break;
+			case 'f': case 'F': r = 15; break;
+			default: r = p - '0'; break;
+			}
+
+			r *= 16;
+
+			switch (q)
+			{
+			case 0: r = '?'; break;
+			case 'a': case 'A': r += 10; break;
+			case 'b': case 'B': r += 11; break;
+			case 'c': case 'C': r += 12; break;
+			case 'd': case 'D': r += 13; break;
+			case 'e': case 'E': r += 14; break;
+			case 'f': case 'F': r += 15; break;
+			default: r += q - '0'; break;
+			}
+
+			result += static_cast <string::value_type>(r);
 
 			if (it != s.end())
 				++it;

tests/utility/urlTest.cpp

@@ -0,0 +1,209 @@
+//
+// VMime library (http://www.vmime.org)
+// Copyright (C) 2002-2005 Vincent Richard <vincent@vincent-richard.net>
+//
+// This program is free software; you can redistribute it and/or
+// modify it under the terms of the GNU General Public License as
+// published by the Free Software Foundation; either version 2 of
+// the License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+// General Public License for more details.
+//
+// You should have received a copy of the GNU General Public License
+// along with this program; if not, write to the Free Software
+// Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
+//
+
+#include "../lib/unit++/unit++.h"
+
+#include <iostream>
+#include <ostream>
+
+#include "vmime/vmime.hpp"
+#include "vmime/platforms/posix/posixHandler.hpp"
+
+#include "vmime/utility/url.hpp"
+#include "vmime/utility/urlUtils.hpp"
+
+using namespace unitpp;
+
+
+namespace
+{
+	class urlTest : public suite
+	{
+		static const bool parseHelper(vmime::utility::url& u, const vmime::string& str)
+		{
+			try
+			{
+				u = vmime::utility::url(str);
+			}
+			catch (vmime::exceptions::malformed_url)
+			{
+				return false;
+			}
+
+			return true;
+		}
+
+
+		void testParse1()
+		{
+			// Test some valid constructions
+			vmime::utility::url u1("", "");
+
+			assert_eq("1.1", true, parseHelper(u1, "protocol://user:password@host:12345/path/"));
+			assert_eq("1.2", "protocol", u1.getProtocol());
+			assert_eq("1.3", "user", u1.getUsername());
+			assert_eq("1.4", "password", u1.getPassword());
+			assert_eq("1.5", "host", u1.getHost());
+			assert_eq("1.6", 12345, u1.getPort());
+			assert_eq("1.7", "/path/", u1.getPath());
+
+			vmime::utility::url u2("", "");
+
+			assert_eq("2.1", true, parseHelper(u2, "protocol://user@host:12345/path/"));
+			assert_eq("2.2", "protocol", u2.getProtocol());
+			assert_eq("2.3", "user", u2.getUsername());
+			assert_eq("2.4", "", u2.getPassword());
+			assert_eq("2.5", "host", u2.getHost());
+			assert_eq("2.6", 12345, u2.getPort());
+			assert_eq("2.7", "/path/", u2.getPath());
+
+			vmime::utility::url u3("", "");
+
+			assert_eq("3.1", true, parseHelper(u3, "protocol://host:12345/path/"));
+			assert_eq("3.2", "protocol", u3.getProtocol());
+			assert_eq("3.3", "", u3.getUsername());
+			assert_eq("3.4", "", u3.getPassword());
+			assert_eq("3.5", "host", u3.getHost());
+			assert_eq("3.6", 12345, u3.getPort());
+			assert_eq("3.7", "/path/", u3.getPath());
+
+			vmime::utility::url u4("", "");
+
+			assert_eq("4.1", true, parseHelper(u4, "protocol://host/path/"));
+			assert_eq("4.2", "protocol", u4.getProtocol());
+			assert_eq("4.3", "", u4.getUsername());
+			assert_eq("4.4", "", u4.getPassword());
+			assert_eq("4.5", "host", u4.getHost());
+			assert_eq("4.6", vmime::utility::url::UNSPECIFIED_PORT, u4.getPort());
+			assert_eq("4.7", "/path/", u4.getPath());
+
+			vmime::utility::url u5("", "");
+
+			assert_eq("5.1", true, parseHelper(u5, "protocol://host/"));
+			assert_eq("5.2", "protocol", u5.getProtocol());
+			assert_eq("5.3", "", u5.getUsername());
+			assert_eq("5.4", "", u5.getPassword());
+			assert_eq("5.5", "host", u5.getHost());
+			assert_eq("5.6", vmime::utility::url::UNSPECIFIED_PORT, u4.getPort());
+			assert_eq("5.7", "", u5.getPath());
+
+			vmime::utility::url u6("", "");
+
+			assert_eq("6.1", true, parseHelper(u4, "protocol://host/path/file"));
+			assert_eq("6.2", "protocol", u4.getProtocol());
+			assert_eq("6.3", "", u4.getUsername());
+			assert_eq("6.4", "", u4.getPassword());
+			assert_eq("6.5", "host", u4.getHost());
+			assert_eq("6.6", vmime::utility::url::UNSPECIFIED_PORT, u4.getPort());
+			assert_eq("6.7", "/path/file", u4.getPath());
+		}
+
+		void testParse2()
+		{
+			// Now, test some ill-formed URLs
+
+			// -- missing protocol
+			vmime::utility::url u1("", "");
+			assert_eq("1", false, parseHelper(u1, "://host"));
+
+			// -- port can contain only digits
+			vmime::utility::url u2("", "");
+			assert_eq("2", false, parseHelper(u2, "proto://host:abc123"));
+
+			// -- no host specified
+			vmime::utility::url u3("", "");
+			assert_eq("3", false, parseHelper(u3, "proto:///path"));
+
+			// -- no protocol separator (://)
+			vmime::utility::url u4("", "");
+			assert_eq("4", false, parseHelper(u4, "protohost/path"));
+		}
+
+		void testParse3()
+		{
+			// Test decoding
+			vmime::utility::url u1("", "");
+
+			assert_eq("1.1", true, parseHelper(u1, "pro%12to://user%34:pass%56word@ho%78st:12345/pa%abth/"));
+			assert_eq("1.2", "pro%12to", u1.getProtocol());  // protocol should not be decoded
+			assert_eq("1.3", "user\x34", u1.getUsername());
+			assert_eq("1.4", "pass\x56word", u1.getPassword());
+			assert_eq("1.5", "ho\x78st", u1.getHost());
+			assert_eq("1.6", 12345, u1.getPort());
+			assert_eq("1.7", "/pa\xabth/", u1.getPath());
+		}
+
+		void testGenerate()
+		{
+			vmime::utility::url u1("proto", "host", 12345, "path", "user", "password");
+			assert_eq("1", "proto://user:password@host:12345/path",
+				static_cast <vmime::string>(u1));
+
+			// TODO: more tests
+		}
+
+		void testUtilsEncode()
+		{
+			assert_eq("1", "%01", vmime::utility::urlUtils::encode("\x01"));
+			assert_eq("2", "%20", vmime::utility::urlUtils::encode(" "));
+			assert_eq("3", "%FF", vmime::utility::urlUtils::encode("\xff"));
+			assert_eq("4", "a", vmime::utility::urlUtils::encode("a"));
+		}
+
+		void testUtilsDecode()
+		{
+			for (int i = 0 ; i < 255 ; ++i)
+			{
+				std::ostringstream ossTest;
+				ossTest << "%" << "0123456789ABCDEF"[i / 16]
+				               << "0123456789ABCDEF"[i % 16];
+
+				std::ostringstream ossNum;
+				ossNum << i;
+
+				vmime::string res;
+				res += static_cast <unsigned char>(i);
+
+				assert_eq(ossNum.str(), res,
+					vmime::utility::urlUtils::decode(ossTest.str()));
+			}
+
+		}
+
+	public:
+
+		urlTest() : suite("vmime::utility::url")
+		{
+			// VMime initialization
+			vmime::platformDependant::setHandler<vmime::platforms::posix::posixHandler>();
+
+			add("Parse1", testcase(this, "Parse1", &urlTest::testParse1));
+			add("Parse2", testcase(this, "Parse2", &urlTest::testParse2));
+			add("Parse3", testcase(this, "Parse3", &urlTest::testParse3));
+			add("Generate", testcase(this, "Generate", &urlTest::testGenerate));
+			add("UtilsEncode", testcase(this, "UtilsEncode", &urlTest::testUtilsEncode));
+			add("UtilsDecode", testcase(this, "UtilsDecode", &urlTest::testUtilsDecode));
+
+			suite::main().add("vmime::utility::url", this);
+		}
+
+	};
+
+	urlTest* theTest = new urlTest();
+}