关于编译器无法识别字符串中十六进制转议字符(\xff)和普通字符的问题

Really_want 2017-09-20 07:08:22

无意间发现一个有趣的情况，也不知这算不算一个问题。在网上还没找到相关的令我比较信服的解释。如下图所示：

预期，3行输出的结果应该是相同的，都将解释为两个字节的数据: 0x5,0x31
但是，实际运行结果，并不如此。如果将这两个字节的数据用普通字符('1')和转义字符(\x05)混合存到字符串(p2)中，将不能得到预期结果。十六进制转义字符，我采用的是\xff格式；我也用过 \xddd,\xdddd 格式，结果均相同。
参考汇编后的代码，问题出在把"\051"解释成了"Q\0" !
参考：
https://msdn.microsoft.com/zh-cn/library/edsza5ck.aspx
http://blog.csdn.net/shangboerds/article/details/7543832
http://www.cnblogs.com/graphics/archive/2011/07/08/2100909.html

//helloworld.cpp

#include <cstdio>



int main(int argc,char*argv[])

{

   char ch[2] = {'\x05', '1'};

   const char* p = "\x05\x31";

   const char* p2 = "\x051";

   printf("ch:%x,%x\n",ch[0],ch[1]);

   printf("p :%x,%x\n",p[0],p[1]);

   printf("p2:%x,%x\n",p2[0],p2[1]);

   return 0;

}

//makefile

exes = a.exe a.s



target: $(exes)



a.exe: helloworld.cpp dump.h

	g++ -Wall -O0 -o a.exe helloworld.cpp



a.s: helloworld.cpp

	g++ -S -o a.s helloworld.cpp



clean:

	cmd //C del $(exes)

//a.s

	.file	"helloworld.cpp"

	.def	___main;	.scl	2;	.type	32;	.endef

	.section .rdata,"dr"

LC0:

	.ascii "\5"

	.ascii "1\0"

LC1:

	.ascii "Q\0"

LC2:

	.ascii "ch:%x,%x\12\0"

LC3:

	.ascii "p :%x,%x\12\0"

LC4:

	.ascii "p2:%x,%x\12\0"

	.text

	.globl	_main

	.def	_main;	.scl	2;	.type	32;	.endef

_main:

LFB12:

	.cfi_startproc

	pushl	%ebp

	.cfi_def_cfa_offset 8

	.cfi_offset 5, -8

	movl	%esp, %ebp

	.cfi_def_cfa_register 5

	andl	$-16, %esp

	subl	$32, %esp

	call	___main

	movb	$5, 22(%esp)

	movb	$49, 23(%esp)

	movl	$LC0, 28(%esp)

	movl	$LC1, 24(%esp)

	movzbl	23(%esp), %eax

	movsbl	%al, %edx

	movzbl	22(%esp), %eax

	movsbl	%al, %eax

	movl	%edx, 8(%esp)

	movl	%eax, 4(%esp)

	movl	$LC2, (%esp)

	call	_printf

	movl	28(%esp), %eax

	addl	$1, %eax

	movzbl	(%eax), %eax

	movsbl	%al, %edx

	movl	28(%esp), %eax

	movzbl	(%eax), %eax

	movsbl	%al, %eax

	movl	%edx, 8(%esp)

	movl	%eax, 4(%esp)

	movl	$LC3, (%esp)

	call	_printf

	movl	24(%esp), %eax

	addl	$1, %eax

	movzbl	(%eax), %eax

	movsbl	%al, %edx

	movl	24(%esp), %eax

	movzbl	(%eax), %eax

	movsbl	%al, %eax

	movl	%edx, 8(%esp)

	movl	%eax, 4(%esp)

	movl	$LC4, (%esp)

	call	_printf

	movl	$0, %eax

	leave

	.cfi_restore 5

	.cfi_def_cfa 4, 4

	ret

	.cfi_endproc

LFE12:

	.ident	"GCC: (GNU) 5.3.0"

	.def	_printf;	.scl	2;	.type	32;	.endef

...全文

1097 20 打赏收藏转发到动态举报

写回复

用AI写文章

20 条回复

切换为时间正序

请发表友善的回复…

发表回复

Really_want 2017-09-23

打赏
举报

“标准规定的”，这种回答我没有理由不接受。我也不质疑标准的权威。我现在只想知道，为什么"\x05five"不定义成'\x05'加上"five"，而"\005five"可以？如果\ooo可以固定长度为3个八进制数，那为什么\xhh为什么就不能固定长度为2个十六进制数？这其中有什么故事吗？

paschen 2017-09-22

打赏
举报

引用 5 楼 ccssddnn218 的回复:

你们都试验过"\0x0051"能行？在我的测试里表现并不是你们说的那样呢，还是“翻译”成了'Q'。
引用 4 楼 ForestDB 的回复:
LZ给的链接中https://msdn.microsoft.com/zh-cn/library/edsza5ck.aspx不是写得很清楚么：\xhhh，是3个数字，不是2个数字，既然0、5、1都是合法的h，那么自然是凑到一起来当作Q被识别。链接中也给出了"\x05""1"这样的技巧。感觉"\x0051"这样比较符合LZ的审美。
引用 3 楼 paschen 的回复:
1楼正解，"\x05""1"; 要分隔开，否则会连在一起判断，再或者写成："\x0051"

"\x0051"可以，前面要多两个0

ForestDB 2017-09-21

打赏
举报

LZ给的链接中https://msdn.microsoft.com/zh-cn/library/edsza5ck.aspx不是写得很清楚么：\xhhh，是3个数字，不是2个数字，既然0、5、1都是合法的h，那么自然是凑到一起来当作Q被识别。链接中也给出了"\x05""1"这样的技巧。感觉"\x0051"这样比较符合LZ的审美。

ForestDB 2017-09-21

打赏
举报

老赵的资料表示就是这样定义的： octal-escape-sequence : \octal-digit \octal-digit octal-digit \octal-digit octal-digit octal-digit // \ooo最多三个oct hexadecimal-escape-sequence : \xhexadecimal-digit hexadecimal-escape-sequence hexadecimal-digit // \xhh是个递归的定义，多长都可以所以： 1、使用""连接 2、使用\ooo的方式

ForestDB 2017-09-21

打赏
举报

是我错了，我试了下，如果是\x打头的话，那么只要是个小于0xff的合法的hex数的话，前面有多少0都是无所谓的。那么只有用""连接的方式了。然后\ooo是有长度限制的，最多就3个合法的oct。

赵4老师 2017-09-21

打赏
举报

赵4老师 2017-09-21

打赏
举报

C++ String Literals
A string literal consists of zero or more characters from the source character set surrounded by double quotation marks ("). A string literal represents a sequence of characters that, taken together, form a null-terminated string.

Syntax

string-literal :

"s-char-sequenceopt"
L"s-char-sequenceopt"

s-char-sequence :

s-char
s-char-sequence s-char

s-char :

any member of the source character set except the double quotation mark ("), backslash (\), or newline character
escape-sequence

C++ strings have these types:

Array of char[n], where n is the length of the string (in characters) plus 1 for the terminating '\0' that marks the end of the string

Array of wchar_t, for wide-character strings
The result of modifying a string constant is undefined. For example:

char *szStr = "1234";
szStr[2] = 'A'; // Results undefined

Microsoft Specific

In some cases, identical string literals can be “pooled” to save space in the executable file. In string-literal pooling, the compiler causes all references to a particular string literal to point to the same location in memory, instead of having each reference point to a separate instance of the string literal. The/Gf compiler option enables string pooling.

END Microsoft Specific

When specifying string literals, adjacent strings are concatenated. Therefore, this declaration:

char szStr[] = "12" "34";

is identical to this declaration:

char szStr[] = "1234";

This concatenation of adjacent strings makes it easy to specify long strings across multiple lines:

cout << "Four score and seven years "
"ago, our forefathers brought forth "
"upon this continent a new nation.";

In the preceding example, the entire string Four score and seven years ago, our forefathers brought forth upon this continent a new nation. is spliced together. This string can also be specified using line splicing as follows:

cout << "Four score and seven years \
ago, our forefathers brought forth \
upon this continent a new nation.";

After all adjacent strings in the constant have been concatenated, the NULL character, '\0', is appended to provide an end-of-string marker for C string-handling functions.

When the first string contains an escape character, string concatenation can yield surprising results. Consider the following two declarations:

char szStr1[] = "\01" "23";
char szStr2[] = "\0123";

Although it is natural to assume that szStr1 and szStr2 contain the same values, the values they actually contain are shown in Figure 1.1.

Figure 1.1 Escapes and String Concatenation

Microsoft Specific

The maximum length of a string literal is approximately 2,048 bytes. This limit applies to strings of type char[] and wchar_t[]. If a string literal consists of parts enclosed in double quotation marks, the preprocessor concatenates the parts into a single string, and for each line concatenated, it adds an extra byte to the total number of bytes.

For example, suppose a string consists of 40 lines with 50 characters per line (2,000 characters), and one line with 7 characters, and each line is surrounded by double quotation marks. This adds up to 2,007 bytes plus one byte for the terminating null character, for a total of 2,008 bytes. On concatenation, an extra character is added to the total number of bytes for each of the first 40 lines. This makes a total of 2,048 bytes. (The extra characters are not actually written to the string.) Note, however, that if line continuations (\) are used instead of double quotation marks, the preprocessor does not add an extra character for each line.

END Microsoft Specific

Determine the size of string objects by counting the number of characters and adding 1 for the terminating '\0' or 2 for type wchar_t.

Because the double quotation mark (") encloses strings, use the escape sequence (\") to represent enclosed double quotation marks. The single quotation mark (') can be represented without an escape sequence. The backslash character (\) is a line-continuation character when placed at the end of a line. If you want a backslash character to appear within a string, you must type two backslashes (\\). (SeePhases of Translation in the Preprocessor Reference for more information about line continuation.)

To specify a string of type wide-character (wchar_t[]), precede the opening double quotation mark with the character L. For example:

wchar_t wszStr[] = L"1a1g";

All normal escape codes listed in Character Constants are valid in string constants. For example:

cout << "First line\nSecond line";
cout << "Error! Take corrective action\a";

Because the escape code terminates at the first character that is not a hexadecimal digit, specification of string constants with embedded hexadecimal escape codes can cause unexpected results. The following example is intended to create a string literal containing ASCII 5, followed by the characters five:

\x05five"

The actual result is a hexadecimal 5F, which is the ASCII code for an underscore, followed by the characters ive. The following example produces the desired results:

"\005five" // Use octal constant.
"\x05" "five" // Use string splicing.

赵4老师 2017-09-21

打赏
举报

查MSDN是Windows程序员必须掌握的技能之一。 C++ Character Constants Character constants are one or more members of the “source character set,” the character set in which a program is written, surrounded by single quotation marks ('). They are used to represent characters in the “execution character set,” the character set on the machine where the program executes. Microsoft Specific For Microsoft C++, the source and execution character sets are both ASCII. END Microsoft Specific There are three kinds of character constants: Normal character constants Multicharacter constants Wide-character constants Note Use wide-character constants in place of multicharacter constants to ensure portability. Character constants are specified as one or more characters enclosed in single quotation marks. For example: char ch = 'x'; // Specify normal character constant. int mbch = 'ab'; // Specify system-dependent // multicharacter constant. wchar_t wcch = L'ab'; // Specify wide-character constant. Note that mbch is of type int. If it were declared as type char, the second byte would not be retained. A multicharacter constant has four meaningful characters; specifying more than four generates an error message. Syntax character-constant : 'c-char-sequence' L'c-char-sequence' c-char-sequence : c-char c-char-sequence c-char c-char : any member of the source character set except the single quotation mark ('), backslash (\), or newline character escape-sequence escape-sequence : simple-escape-sequence octal-escape-sequence hexadecimal-escape-sequence simple-escape-sequence : one of \' \" \? \\ \a \b \f \n \r \t \v octal-escape-sequence : \octal-digit \octal-digit octal-digit \octal-digit octal-digit octal-digit hexadecimal-escape-sequence : \xhexadecimal-digit hexadecimal-escape-sequence hexadecimal-digit Microsoft C++ supports normal, multicharacter, and wide-character constants. Use wide-character constants to specify members of the extended execution character set (for example, to support an international application). Normal character constants have type char, multicharacter constants have type int, and wide-character constants have type wchar_t. (The type wchar_t is defined in the standard include files STDDEF.H, STDLIB.H, and STRING.H. The wide-character functions, however, are prototyped only in STDLIB.H.) The only difference in specification between normal and wide-character constants is that wide-character constants are preceded by the letter L. For example: char schar = 'x'; // Normal character constant wchar_t wchar = L'\x81\x19'; // Wide-character constant Table 1.2 shows reserved or nongraphic characters that are system dependent or not allowed within character constants. These characters should be represented with escape sequences. Table 1.2 C++ Reserved or Nongraphic Characters Character ASCII Representation ASCII Value Escape Sequence Newline NL (LF) 10 or 0x0a \n Horizontal tab HT 9 \t Vertical tab VT 11 or 0x0b \v Backspace BS 8 \b Carriage return CR 13 or 0x0d \r Formfeed FF 12 or 0x0c \f Alert BEL 7 \a Backslash \ 92 or 0x5c \\ Question mark ? 63 or 0x3f \? Single quotation mark ' 39 or 0x27 \' Double quotation mark " 34 or 0x22 \" Octal number ooo — \ooo Hexadecimal number hhh — \xhhh Null character NUL 0 \0 If the character following the backslash does not specify a legal escape sequence, the result is implementation defined. In Microsoft C++, the character following the backslash is taken literally, as though the escape were not present, and a level 1 warning (“unrecognized character escape sequence”) is issued. Octal escape sequences, specified in the form \ooo, consist of a backslash and one, two, or three octal characters. Hexadecimal escape sequences, specified in the form \xhhh, consist of the characters \x followed by a sequence of hexadecimal digits. Unlike octal escape constants, there is no limit on the number of hexadecimal digits in an escape sequence. Octal escape sequences are terminated by the first character that is not an octal digit, or when three characters are seen. For example: wchar_t och = L'\076a'; // Sequence terminates at a char ch = '\233'; // Sequence terminates after 3 characters Similarly, hexadecimal escape sequences terminate at the first character that is not a hexadecimal digit. Because hexadecimal digits include the letters a through f (and A through F), make sure the escape sequence terminates at the intended digit. Because the single quotation mark (') encloses character constants, use the escape sequence \' to represent enclosed single quotation marks. The double quotation mark (") can be represented without an escape sequence. The backslash character (\) is a line-continuation character when placed at the end of a line. If you want a backslash character to appear within a character constant, you must type two backslashes in a row (\\). (SeePhases of Translation in the Preprocessor Reference for more information about line continuation.)

ForestDB 2017-09-21

打赏
举报

引用 13 楼 ccssddnn218 的回复:

谢谢zhao4zhong1老师和其他网友的耐心查阅和回答。八进制的写法,\000: "\005five" 十六进制写法,\xhh: "\x05five" 怎么看，我都觉得这两种写法表示的是同一个意思。我找不出任何不支持第2种写法的理由。 MSDN? 谁写的MSDN？！画个圈圈诅咒你：

八进制的写法没跑，\005 f i v e 十六进制的写法就比较坑，\x05f i v e，为什么呢？f是个有效的十六进制数啊，\x后面可以跟若干个十六进制数啊，标准规定的啊，12L就是标准的描述啊。

Really_want 2017-09-21

打赏
举报

"H\x01ello World!\n" 这个字符串我是希望在"Hello World!"中插入一个'\x01'字节，看看程序的表现。如果采用多个引号的方法就只表表示成下面这样了： "H""\x01""ello World!\n" 我倒是想用第一种形式，无论是\x001还是\x0001都无所谓，只要格式能统一，但测试表明"...\x001e..."不是我想要的结果。沮丧

Really_want 2017-09-21

打赏
举报

更正一下！"\x0051"

引用 5 楼 ccssddnn218 的回复:

你们都试验过"\0x0051"能行？在我的测试里表现并不是你们说的那样呢，还是“翻译”成了'Q'。 [quote=引用 4 楼 ForestDB 的回复:] LZ给的链接中https://msdn.microsoft.com/zh-cn/library/edsza5ck.aspx不是写得很清楚么：\xhhh，是3个数字，不是2个数字，既然0、5、1都是合法的h，那么自然是凑到一起来当作Q被识别。链接中也给出了"\x05""1"这样的技巧。感觉"\x0051"这样比较符合LZ的审美。

引用 3 楼 paschen 的回复:

1楼正解，"\x05""1"; 要分隔开，否则会连在一起判断，再或者写成："\x0051"

[/quote]

Really_want 2017-09-21

打赏
举报

你们都试验过"\0x0051"能行？在我的测试里表现并不是你们说的那样呢，还是“翻译”成了'Q'。

引用 4 楼 ForestDB 的回复:

LZ给的链接中https://msdn.microsoft.com/zh-cn/library/edsza5ck.aspx不是写得很清楚么：\xhhh，是3个数字，不是2个数字，既然0、5、1都是合法的h，那么自然是凑到一起来当作Q被识别。链接中也给出了"\x05""1"这样的技巧。感觉"\x0051"这样比较符合LZ的审美。

引用 3 楼 paschen 的回复:

1楼正解，"\x05""1"; 要分隔开，否则会连在一起判断，再或者写成："\x0051"

赵4老师 2017-09-21

打赏
举报

仅供参考：

#pragma comment(lib,"user32")
#pragma comment(lib,"gdi32")
#include <conio.h>
#include <stdio.h>
#include <stdlib.h>
#include <windows.h>
extern "C" HWND WINAPI GetConsoleWindow();
void HideTheCursor() {
    CONSOLE_CURSOR_INFO cciCursor;
    HANDLE hStdOut = GetStdHandle(STD_OUTPUT_HANDLE);

    if(GetConsoleCursorInfo(hStdOut, &cciCursor)) {
        cciCursor.bVisible = FALSE;
        SetConsoleCursorInfo(hStdOut, &cciCursor);
    }
}
void ShowTheCursor() {
    CONSOLE_CURSOR_INFO cciCursor;
    HANDLE hStdOut = GetStdHandle(STD_OUTPUT_HANDLE);

    if(GetConsoleCursorInfo(hStdOut, &cciCursor)) {
        cciCursor.bVisible = TRUE;
        SetConsoleCursorInfo(hStdOut, &cciCursor);
    }
}
int main() {
    HWND  hwnd;
    HDC   hdc;
    HFONT hfont;
    wchar_t wc[2];

    system("color F0");
    system("cls");
    HideTheCursor();
    hwnd  = GetConsoleWindow();
    hdc   = GetDC(hwnd);
    hfont = CreateFont(48,0,0,0,0,0,0,0,GB2312_CHARSET ,0,0,0,0,"宋体-方正超大字符集");
    SelectObject(hdc,hfont);
    wc[0]=0xD854u;
    wc[1]=0xDC00u;
    TextOutW(hdc,10,10,wc,2);
    DeleteObject(hfont);
    ReleaseDC(hwnd,hdc);
    getch();
    system("color 07");
    system("cls");
    ShowTheCursor();
    return 0;
}
#if 0
代理项或代理项对是一对共同表示单个字符的 16 位 Unicode 编码值。需要记住的关键一点是：
代理项对实际上是 32 位单个字符，不能再假定一个 16 位 Unicode 编码值正好映射到一个字符。

使用代理项对
代理项对的第一个值是高代理项，包含介于 U+D800 到 U+DBFF 范围内的 16 位代码值。
该对的第二个值是低代理项，包含介于 U+DC00 到 U+DFFF 范围内的值。通过使用代理项对，
16 位 Unicode 编码系统可以对已由 Unicode 标准定义的一百多万个其他字符 (220) 进行寻址。

在传递给 XmlTextWriter 方法的任何字符串中都可以使用代理项字符。不过，代理项字符在编写的
XML 中应该有效。例如，万维网联合会 (W3C) 建议不允许在元素或属性的名称中使用代理项字符。
如果字符串包含无效的代理项对，则引发异常。

另外，可以使用 WriteSurrogateCharEntity 写出与代理项对相对应的字符实体。字符实体以十六
进制格式写出，并用以下公式生成：

(highChar -0xD800) * 0x400 + (lowChar -0xDC00) + 0x10000

如果字符串包含无效的代理项对，则引发异常。下面的示例显示将代理项对作为输入的 WriteSurrogateCharEntity 方法。

C#复制
 // The following line writes 𐀀.
WriteSurrogateCharEntity ('\uDC00', '\uD800');
下面的示例生成一个代理项对文件，将其加载到 XmlReader 中，并用新的文件名保存文件。
然后，原始文件和新文件被加载回应用程序的 XML 文档对象模型 (DOM) 结构中以进行比较。

C#复制
 char lowChar, highChar;
char [] charArray = new char[10];
FileStream targetFile = new FileStream("SurrogatePair.xml",
      FileMode.Create, FileAccess.ReadWrite, FileShare.ReadWrite);

lowChar = Convert.ToChar(0xDC00);
highChar = Convert.ToChar(0xD800);
XmlTextWriter tw = new XmlTextWriter(targetFile, null);
tw.Formatting = Formatting.Indented;
tw.WriteStartElement("root");
tw.WriteStartAttribute("test", null);
tw.WriteSurrogateCharEntity(lowChar, highChar);
lowChar = Convert.ToChar(0xDC01);
highChar = Convert.ToChar(0xD801);
tw.WriteSurrogateCharEntity(lowChar, highChar);
lowChar = Convert.ToChar(0xDFFF);
highChar = Convert.ToChar(0xDBFF);
tw.WriteSurrogateCharEntity(lowChar, highChar);

// Add 10 random surrogate pairs.
// As Unicode, the high bytes are in lower
// memory; for example, word 6A21 as 21 6A.
// The high or low is in the logical sense.
Random random = new Random();
for (int i = 0; i < 10; ++i) {
      lowChar = Convert.ToChar(random.Next(0xDC00, 0xE000));
      highChar = Convert.ToChar(random.Next(0xD800, 0xDC00));
      charArray[i] = highChar;
      charArray[++i] = lowChar;
}
tw.WriteChars(charArray, 0, charArray.Length);

for (int i = 0; i < 10; ++i) {
      lowChar = Convert.ToChar(random.Next(0xDC00, 0xE000));
      highChar = Convert.ToChar(random.Next(0xD800, 0xDC00));
      tw.WriteSurrogateCharEntity(lowChar, highChar);
}

tw.WriteEndAttribute();
tw.WriteEndElement();
tw.Flush();
tw.Close();

XmlTextReader r = new XmlTextReader("SurrogatePair.xml");

r.Read();
r.MoveToFirstAttribute();
targetFile = new FileStream("SurrogatePairFromReader.xml",
       FileMode.Create, FileAccess.ReadWrite, FileShare.ReadWrite);

tw = new XmlTextWriter(targetFile, null);
tw.Formatting = Formatting.Indented;
tw.WriteStartElement("root");
tw.WriteStartAttribute("test", null);
tw.WriteString(r.Value);
tw.WriteEndAttribute();
tw.WriteEndElement();
tw.Flush();
tw.Close();

// Load both result files into the DOM and compare.
XmlDocument doc1 = new XmlDocument();
XmlDocument doc2 = new XmlDocument();
doc1.Load("SurrogatePair.xml");
doc2.Load("SurrogatePairFromReader.xml");
if (doc1.InnerXml != doc2.InnerXml) {
      Console.WriteLine("Surrogate Pair test case failed");
}
在使用 WriteChars 方法（一次写出一个缓冲区的数据）写出时，输入中的代理项对可能
会在一个缓冲区内被意外拆分。由于代理项值是定义完善的，如果 WriteChars 遇到来自
较低范围或者较高范围的 Unicode 值，它将该值标识为代理项对的一半。当遇到
 WriteChars 将导致从拆分代理项对的缓冲区写入的情况时，将引发异常。使用
  IsHighSurrogate 方法检查缓冲区是否以高代理项字符结束。如果缓冲区中的最后一个
  字符不是高代理项，可以将该缓冲区传递给 WriteChars 方法。

请参见
概念
使用 XmlTextWriter 创建格式正确的 XML
XmlTextWriter 的 XML 输出格式设置
XmlTextWriter 的命名空间功能

#endif

赵4老师 2017-09-21