fix(Net::MessageHeader): Correct RFC 2047 decoding for two consecutiv…

…e encoded words (pocoproject#4542)

* fix bug pocoproject#4535
RFC 2047 decodeWord

* modify logic
uses separate string which contains simbols between chunks and if this
string contains only space ot \n or \t or \v than trim it

Net/src/MessageHeader.cpp

@@ -413,10 +413,10 @@ void MessageHeader::decodeRFC2047(const std::string& ins, std::string& outs, con
 std::string MessageHeader::decodeWord(const std::string& text, const std::string& charset)
 {
 	std::string outs, tmp = text;
+	size_t pos = tmp.find("=?");
 	do {
 		std::string tmp2;
 		// find the begining of the next rfc2047 chunk
-		size_t pos = tmp.find("=?");
 		if (pos == std::string::npos) {
 			// No more found, return
 			outs += tmp;
@@ -453,14 +453,24 @@ std::string MessageHeader::decodeWord(const std::string& text, const std::string
 			// not found.
 			outs += tmp;
 			break;
-
 		}
+
 		// At this place, there are a valid rfc2047 chunk, so decode and copy the result.
 		decodeRFC2047(tmp.substr(0, pos3), tmp2, charset);
 		outs += tmp2;
 
 		// Jump at the rest of the string and repeat the whole process.
 		tmp = tmp.substr(pos3 + 2);
+		pos = tmp.find("=?");
+		if (pos != std::string::npos)
+		{
+			std::string betweenChunks = tmp.substr(0, pos);
+			if (betweenChunks.find_first_not_of(" \t\v\n") == std::string::npos)
+			{
+				tmp = tmp.substr(pos);
+				pos = 0;
+			}
+		}
 	} while (true);
 
 	return outs;

Net/testsuite/src/MessageHeaderTest.cpp

@@ -400,11 +400,19 @@ void MessageHeaderTest::testDecodeWord()
 
 	coded = "(=?ISO-8859-1?Q?a?= =?ISO-8859-1?Q?b?=)";
 	decoded = MessageHeader::decodeWord(coded, "ISO-8859-1");
-	assertTrue (decoded == "(a b)");
+	assertTrue (decoded == "(ab)");
+
+	coded = "(=?ISO-8859-1?Q?a?= <0> =?ISO-8859-1?Q?b?=)";
+	decoded = MessageHeader::decodeWord(coded, "ISO-8859-1");
+	assertTrue (decoded == "(a <0> b)");
 
 	coded = "Hello =?UTF-8?B?RnJhbmNpcw==?=, good bye";
 	decoded = MessageHeader::decodeWord(coded, "ISO-8859-1");
 	assertTrue (decoded == "Hello Francis, good bye");
+
+	coded = "application/pdf; name=\"=?utf-8?Q?RUG_Regler-_und_Ger=C3=A4tebau_Gm?= =?utf-8?Q?bH_Angebot_Erneuerung_=C3=9CE.pdf?=\"";
+	decoded = MessageHeader::decodeWord(coded, "UTF-8");
+	assertTrue (decoded == "application/pdf; name=\"RUG Regler- und Gerätebau GmbH Angebot Erneuerung ÜE.pdf\"");
 }
 
 // Sample HTTP reuest header
@@ -430,10 +438,10 @@ void MessageHeaderTest::testAutoDecode()
 		MessageHeader mh;
 		mh.read(istr);
 
-		assertEquals(mh.get("X-Encoded-Header-A"), "(a b)");
+		assertEquals(mh.get("X-Encoded-Header-A"), "(ab)");
 		assertEquals(mh.get("X-Encoded-Header-B"), "Hello Francis, good bye");
 
-		assertEquals(mh.getDecoded("X-Encoded-Header-A"), "(a b)");
+		assertEquals(mh.getDecoded("X-Encoded-Header-A"), "(ab)");
 		assertEquals(mh.getDecoded("X-Encoded-Header-B"), "Hello Francis, good bye");
 	}
 	{
@@ -445,7 +453,7 @@ void MessageHeaderTest::testAutoDecode()
 		assertEquals(mh.get("X-Encoded-Header-A"), "(=?ISO-8859-1?Q?a?= =?ISO-8859-1?Q?b?=)");
 		assertEquals(mh.get("X-Encoded-Header-B"), "Hello =?UTF-8?B?RnJhbmNpcw==?=, good bye");
 
-		assertEquals(mh.getDecoded("X-Encoded-Header-A"), "(a b)");
+		assertEquals(mh.getDecoded("X-Encoded-Header-A"), "(ab)");
 		assertEquals(mh.getDecoded("X-Encoded-Header-B"), "Hello Francis, good bye");
 	}
 }