[英]Compare string with uint32_t by dereference
Say for instance I wanted to parse input from the user/stdin, this input was a string ( char*
/ char []
).比如说我想解析来自用户/标准输入的输入,这个输入是一个字符串( char*
/ char []
)。 I know I can typecast a string's memory address pointer to a uint32_t
pointer and dereference said uint32_t
pointer to extract a unsigned int
value which is the same across most architectures (that I've tested).我知道我可以将字符串的内存地址指针类型转换为uint32_t
指针并取消引用所述uint32_t
指针以提取一个unsigned int
值,该值在大多数体系结构中都是相同的(我已经测试过)。 Is this safe?这安全吗? Can it lead to bugs/undefined behavior?它会导致错误/未定义的行为吗? ie just in general a bad idea?即一般来说是个坏主意?
Example:例子:
#include <stdio.h>
#include <stdint.h>
int main(void) {
char data[] = "status";
printf("%u\n", *(uint32_t*)data);
return 0;
}
Will result in 1952543859
on 64bit and 32bit (windows & linux, not tested on MAC/unix system)将导致1952543859
在 64 位和 32 位(windows 和 linux,未在 MAC/unix 系统上测试)
Is it safe to do the below operation?:进行以下操作是否安全?:
char str_from_input[] = "status";
if (*(uint32_t*)str_from_input == 1952543859)
//... execute some code oriented around "status".
My obvious aim is to reduce computations by not repeatedly calling string parsing functions like strcmp
and strncmp
by a direct uint32_t
check/if statement.我的明显目标是通过不通过直接uint32_t
check/if 语句重复调用strcmp
和strncmp
等字符串解析函数来减少计算。
… typecast a string's memory address pointer to a uint32_t pointer… Is this safe? ……将字符串的内存地址指针类型转换为 uint32_t 指针……这样安全吗?
No. If the string is not properly aligned for a uint32_t
, the behavior of the cast is not defined, per C 2018 6.3.2.3 7: “… If the resulting pointer is not correctly aligned for the referenced type, the behavior is undefined…”否。如果uint32_t
的字符串未正确对齐,则未定义uint32_t
的行为,根据 C 2018 6.3.2.3 7:“......如果结果指针未正确对齐引用类型,则行为未定义...... ”
Dereferencing the resulting pointer accesses the string using a type that is inappropriate for it, violating the aliasing rule in 6.5 7, and the behavior is not defined by the C standard.取消引用结果指针会使用不适合它的类型访问字符串,违反了 6.5 7 中的别名规则,并且 C 标准没有定义该行为。
Can it lead to bugs/undefined behavior?它会导致错误/未定义的行为吗?
Yes.是的。
ie just in general a bad idea?即一般来说是个坏主意?
Yes.是的。
Is it safe to do the below operation?:进行以下操作是否安全?:
No.不。
To reinterpret the bytes in a string as a uint32_t
, use:要将字符串中的字节重新解释为uint32_t
,请使用:
uint32_t t;
memcpy(&t, string, sizeof t);
printf("%" PRIu32 "\n", t);
( memcpy
is declared in <string.h>
, and PRIu32
is declared in <inttypes.h>
.) ( memcpy
在<string.h>
中声明, PRIu32
在<inttypes.h>
声明。)
To copy the bytes of a uint32_t
into a string, use memcpy(string, &t, sizeof t);
要将uint32_t
的字节复制到字符串中,请使用memcpy(string, &t, sizeof t);
. .
A good C compiler is likely to replace the memcpy
calls with direct data move instructions, if feasible.如果可行,一个好的 C 编译器很可能会用直接数据移动指令替换memcpy
调用。
My obvious aim is to reduce computations by not repeatedly calling string parsing functions like
strcmp
andstrncmp
by a directuint32_t
check/if statement.我的明显目标是通过不通过直接uint32_t
check/if 语句重复调用strcmp
和strncmp
等字符串解析函数来减少计算。
Avoid using kludges.避免使用杂物。 Write clear code and let the compiler optimize.编写清晰的代码并让编译器进行优化。
声明:本站的技术帖子网页,遵循CC BY-SA 4.0协议,如果您需要转载,请注明本站网址或者原文地址。任何问题请咨询:yoyou2525@163.com.