[10715] UTF-8の冗長な表現を受け入れないようにした (ttssh2-commit) - Tera Term

Revision: 10715
          https://osdn.net/projects/ttssh2/scm/svn/commits/10715
Author:   zmatsuo
Date:     2023-05-21 00:28:54 +0900 (Sun, 21 May 2023)
Log Message:
-----------
UTF-8の冗長な表現を受け入れないようにした

- すべて "!" U+0021 (冗長なUTF-8表現)
  - echo -e "\x21"
  - echo -e "\xc0\xa1"
  - echo -e "\xe0\x80\xa1"
  - echo -e "\xf0\x80\x80\xa1"
  - 修正前はすべて U+0021 として扱っていた
- 0x80以上の1byteは、直接Unicodeとして扱うようにした
  - UTF-8 では1byte目に 0x80以上の値は存在しない
  - Tera Term の入力文字コードが UTF-8 のとき、そのまま ISO8059-1 として使用できる
- 上の例の場合は次の入力となる
  - U+0021
  - U+00C0 U+00A1
  - U+00E0 U+0080 U+00A1
  - U+00F0 U+0080 U+0080 U+00A1

ticket #17226

Ticket Links:
------------
    https://osdn.net/projects/ttssh2/tracker/detail/17226

Modified Paths:
--------------
    trunk/teraterm/teraterm/vtterm.c

-------------- next part --------------
Modified: trunk/teraterm/teraterm/vtterm.c
===================================================================

--- trunk/teraterm/teraterm/vtterm.c	2023-05-20 14:14:27 UTC (rev 10714)
+++ trunk/teraterm/teraterm/vtterm.c	2023-05-20 15:28:54 UTC (rev 10715)
@@ -4319,7 +4319,7 @@
 	FirstPrm = FALSE;
 }
 
-int CheckUTF8Seq(BYTE b, int utf8_stat)
+static int CheckUTF8Seq(BYTE b, int utf8_stat)
 {
 	if (ts.Language == IdUtf8 || (ts.Language==IdJapanese && ts.KanjiCode==IdUTF8)) {
 		if (utf8_stat > 0) {
@@ -5973,6 +5973,7 @@
 	static int count = 0;
 
 	unsigned int code;
+	int i;
 
 	if (ts.FallbackToCP932 && Fallbacked) {
 		return ParseFirstJP(b);
@@ -5991,94 +5992,125 @@
 	//		- 0xf5 - 0xff
 	//	- 2byte\x96ڈȍ~
 	//		- 0x00 - 0x7f
-	//		--0xc0 - 0xff
+	//		- 0xc0 - 0xff
 
-	if ((b & 0x80) != 0x80 || ((b & 0xe0) == 0x80 && count == 0)) {
-		// 1\x83o\x83C\x83g\x96ڂ\xA8\x82\xE6\x82\xD12\x83o\x83C\x83g\x96ڂ\xAAASCII\x82̏ꍇ\x82́A\x82\xB7\x82ׂ\xC4ASCII\x8Fo\x97͂Ƃ\xB7\x82\xE9\x81B
-		// 1\x83o\x83C\x83g\x96ڂ\xAAC1\x90\xA7\x8C䕶\x8E\x9A(0x80-0x9f)\x82̏ꍇ\x82\xE0\x93\xAF\x97l\x81B
-
-		// \x93\xFC\x97͕\xB6\x8E\x9A\x82\xAA 0x00 ... 0x7f
-		if (count == 0 || count == 1) {
-			if (count == 1) {
-				ParseASCII(buf[0]);
-			}
+	// 1byte(7bit)
+	if (count == 0) {
+		if ((b & 0x80) == 0x00) {
+			// 1byte(7bit)
+			//		0x7f\x88ȉ\xBA, \x82̂Ƃ\xAB\x81A\x82\xBB\x82̂܂܏o\x97\xCD
 			ParseASCII(b);
-			count = 0;  // reset counter
 			return TRUE;
 		}
-	}
-
-	buf[count++] = b;
-	if (count < 2) {
+		if ((b & 0x40) == 0x00 || b >= 0xf6 ) {
+			// UTF-8\x82\xC51byte\x82ɏo\x8C\xBB\x82\xB5\x82Ȃ\xA2\x83R\x81[\x83h\x82̂Ƃ\xAB\x81A\x82\xBB\x82̂܂܏o\x97\xCD
+			//	0x40 = 0b1011_1111, 0b10xx_xxxx\x82Ƃ\xA2\x82\xA4bit\x83p\x83^\x81[\x83\x93\x82ɂ͂Ȃ\xE7\x82Ȃ\xA2
+			//  0xf6 \x88ȏ\xE3\x82̂Ƃ\xAB U+10FFFF\x82\xE6\x82\xE8\x91傫\x82\xAD\x82Ȃ\xE9
+			PutU32(b);
+			return TRUE;
+		}
+		// 1byte\x96ڕۑ\xB6
+		buf[count++] = b;
 		return TRUE;
 	}
 
-	// 2\x83o\x83C\x83g\x83R\x81[\x83h\x82̏ꍇ
+	// 2byte(11bit)
 	if ((buf[0] & 0xe0) == 0xc0) {
-		if ((buf[1] & 0xc0) == 0x80) {
-
-			code = ((buf[0] & 0x1f) << 6);
-			code |= ((buf[1] & 0x3f));
-
-			PutU32(code);
+		code = 0;
+		if((b & 0xc0) == 0x80) {
+			// 5bit + 6bit
+			code = ((buf[0] & 0x1f) << 6) | (b & 0x3f);
+			if (code < 0x80) {
+				// 11bit\x8Eg\x82\xC1\x82\xC47bit\x88ȉ\xBA\x82̎\x9E\x81AUTF-8\x82̏璷\x82ȕ\\x8C\xBB
+				code = 0;
+			}
 		}
+		if (code == 0){
+			// \x82\xBB\x82̂܂܏o\x97\xCD
+			PutU32(buf[0]);
+			PutU32(b);
+			count = 0;
+			return TRUE;
+		}
 		else {
-			ParseASCII(buf[0]);
-			ParseASCII(buf[1]);
+			PutU32(code);
+			count = 0;
+			return TRUE;
 		}
-		count = 0;
-		return TRUE;
 	}
 
-	if (count < 3) {
-		return TRUE;
-	}
+	// 2byte\x96ڈȍ~\x95ۑ\xB6
+	buf[count++] = b;
 
-	if ((buf[0] & 0xf0) == 0xe0 &&
-		(buf[1] & 0xc0) == 0x80 &&
-		(buf[2] & 0xc0) == 0x80) { // 3\x83o\x83C\x83g\x83R\x81[\x83h\x82̏ꍇ
-
-		// UTF-8 BOM(Byte Order Mark)
-		if (buf[0] == 0xef && buf[1] == 0xbb && buf[2] == 0xbf) {
-			goto skip;
+	// 3byte(16bit)
+	if ((buf[0] & 0xf0) == 0xe0) {
+		if(count < 3) {
+			return TRUE;
 		}
-
-		code = ((buf[0] & 0xf) << 12);
-		code |= ((buf[1] & 0x3f) << 6);
-		code |= ((buf[2] & 0x3f));
-
-		PutU32(code);
-
-skip:
-		count = 0;
-
+		code = 0;
+		if ((buf[1] & 0xc0) == 0x80 && (buf[2] & 0xc0) == 0x80) {
+			// 4bit + 6bit + 6bit
+			code = ((buf[0] & 0xf) << 12);
+			code |= ((buf[1] & 0x3f) << 6);
+			code |= ((buf[2] & 0x3f));
+			if (code < 0x800) {
+				// 16bit\x8Eg\x82\xC1\x82\xC411bit\x88ȉ\xBA\x82̂Ƃ\xAB\x81AUTF-8\x82̏璷\x82ȕ\\x8C\xBB
+				code = 0;
+			}
+		}
+		if (code == 0) {
+			// \x82\xBB\x82̂܂܏o\x97\xCD
+			PutU32(buf[0]);
+			PutU32(buf[1]);
+			PutU32(buf[2]);
+			count = 0;
+			return TRUE;
+		} else {
+			PutU32(code);
+			count = 0;
+			return TRUE;
+		}
 	}
 
-	if (count < 4) {
-		return TRUE;
+	// 4byte(21bit)
+	if ((buf[0] & 0xf8) == 0xf0) {
+		if(count < 4) {
+			return TRUE;
+		}
+		code = 0;
+		if ((buf[1] & 0xc0) == 0x80 && (buf[2] & 0xc0) == 0x80 && (buf[3] & 0xc0) == 0x80) {
+			// 3bit + 6bit + 6bit + 6bit
+			code = ((buf[0] & 0x07) << 18);
+			code |= ((buf[1] & 0x3f) << 12);
+			code |= ((buf[2] & 0x3f) << 6);
+			code |= (buf[3] & 0x3f);
+			if (code < 0x10000) {
+				// 21bit\x8Eg\x82\xC1\x82\xC416bit\x88ȉ\xBA\x82̂Ƃ\xAB\x81AUTF-8\x82̏璷\x82ȕ\\x8C\xBB
+				code = 0;
+			}
+		}
+		if (code == 0) {
+			// \x82\xBB\x82̂܂܏o\x97\xCD
+			PutU32(buf[0]);
+			PutU32(buf[1]);
+			PutU32(buf[2]);
+			PutU32(buf[3]);
+			count = 0;
+			return TRUE;
+		} else {
+			PutU32(code);
+			count = 0;
+			return TRUE;
+		}
 	}
 
-	if ((buf[0] & 0xf8) == 0xf0 &&
-		(buf[1] & 0xc0) == 0x80 &&
-		(buf[2] & 0xc0) == 0x80 &&
-		(buf[3] & 0xc0) == 0x80)
-	{	// 4\x83o\x83C\x83g\x83R\x81[\x83h\x82̏ꍇ
-		code = ((buf[0] & 0x07) << 18);
-		code |= ((buf[1] & 0x3f) << 12);
-		code |= ((buf[2] & 0x3f) << 6);
-		code |= (buf[3] & 0x3f);
+	// \x82\xB1\x82\xB1\x82ɂ͗\x88\x82Ȃ\xA2
+	assert(FALSE);
 
-		PutU32(code);
-		count = 0;
-		return TRUE;
-	} else {
-		ParseASCII(buf[0]);
-		ParseASCII(buf[1]);
-		ParseASCII(buf[2]);
-		ParseASCII(buf[3]);
-		count = 0;
+	for (i = 0; i < count; i++) {
+		ParseASCII(buf[i]);
 	}
-
+	count = 0;
 	return TRUE;
 }
 


Tera Term

[ttssh2-commit] [10715] UTF-8の冗長な表現を受け入れないようにした