我想将字符数组micPointsChar[]
传递给函数initMicPoints()
并将其解析为多维数组micPoints
。我能够使用一维数组成功地做到这一点:
char micPointsChar[30 + 1] = {};
float *initMicPoints(char micPointsChar[], float micPoints[3]);
int main()
{
// Read in mic points from file
char micPointsChar[40] = "2,3.343,4.432\n";
float micPoints[3] = {};
float *newMicPoints = initMicPoints(micPointsChar, micPoints);
for (int i = 1; i <= 3; i++)
{
Serial.print(newMicPoints[i]);
Serial.print("\n");
}
return 0;
}
float *initMicPoints(char micPointsChar[], float micPoints[3])
{
static int i = 1;
static int micNum = 1;
static int numMics = 1;
float coordinateDec = 0;
char *coordinate = strtok(micPointsChar, ",\n");
coordinateDec = atof(coordinate);
while (micNum <= numMics)
{
while (i <= ((micNum * 3)) && (coordinate != NULL))
{
if (i == ((micNum * 3) - 2))
{
micPoints[1] = coordinateDec;
}
else if (i == ((micNum * 3) - 1))
{
micPoints[2] = coordinateDec;
}
else if (i == ((micNum * 3) - 0))
{
micPoints[3] = coordinateDec;
}
coordinate = strtok(NULL, ",\n");
coordinateDec = atof(coordinate);
i++;
}
micNum++;
}
return micPoints;
}
这会输出预期的:
2.00
3.34
4.43
但是,当我更改代码以处理多维数组时,micPoints[360][3]
:
char micPointsChar[30 + 1] = {};
float *initMicPoints(char micPointsChar[], float micPoints[360][3]);
int main()
{
// Read in mic points from file
char micPointsChar[40] = "2,3.343,4.432\n";
float micPoints[360][3] = {};
float *newMicPoints = initMicPoints(micPointsChar, micPoints);
static int i = 0;
for (i = 1; i <= 3; i++)
{
Serial.print(*newMicPoints[i][0]);
Serial.print("\n");
Serial.print(*newMicPoints[i][1]);
Serial.print("\n");
Serial.print(*newMicPoints[i][2]);
Serial.print("\n");
}
return 0;
}
float *initMicPoints(char micPointsChar[], float micPoints[360][3])
{
static int i = 1;
static int micNum = 1;
static int numMics = 1;
float coordinateDec = 0;
char *coordinate = strtok(micPointsChar, ",\n");
coordinateDec = atof(coordinate);
while (micNum <= numMics)
{
while (i <= ((micNum * 3)) && (coordinate != NULL))
{
if (i == ((micNum * 3) - 2))
{
micPoints[i][0] = coordinateDec;
}
else if (i == ((micNum * 3) - 1))
{
micPoints[i][1] = coordinateDec;
}
else if (i == ((micNum * 3) - 0))
{
micPoints[i][2] = coordinateDec;
}
coordinate = strtok(NULL, ",\n");
coordinateDec = atof(coordinate);
i++;
}
micNum++;
}
return micPoints;
}
我收到编译时错误:
cannot convert 'float (*)[3]' to 'float*' in return
我这太复杂了吗?返回多维数组的最佳方法是什么?
答案 0 :(得分:2)
首先,不幸的是
float *initMicPoints(char micPointsChar[], float micPoints[360][3])
被视为
float *initMicPoints(char* micPointsChar, float (*micPoints)[3])
您可以通过引用传递以保持大小:
float *initMicPoints(char* micPointsChar, float (&micPoints)[360][3])
然后返回micPoints
返回类型应为float (&)[360][3]
或float (&)[360][3]
这给了一个丑陋的
float (&initMicPoints(char* micPointsChar, float (&micPoints)[360][3]))[360][3]
并在通话现场:
float (&newMicPoints)[360][3] = initMicPoints(micPointsChar, micPoints);
首选std::array
或std::vector
语法清晰。
答案 1 :(得分:0)
在这两种情况下,您只是返回参数。所以这个返回值是多余的。相反,请返回void
:
void initMicPoints(char micPointsChar[], float micPoints[360][3])
调用代码如下所示:
float micPoints[360][3] = {};
initMicPoints(micPointsChar, micPoints);
for (int i = 1; i <= 3; i++)
{
Serial.print(micPoints[i][0]);
Serial.print("\n");
等。如果需要,您可以创建另一个变量float (*newMicPoints)[3] = micPoints;
,但这也是多余的。
答案 2 :(得分:0)
返回数组没什么意义,因为你的函数没有构造它。你只是给调用者一个参数值的副本。
标准C库中的一些传统函数可以执行此操作,例如strcpy
。我不记得上次我看到一段使用strcpy
返回值的代码,它只是传入的目标指针。
// redeclare and redefine to return nothing!
void initMicPoints(char micPointsChar[], float micPoints[3]);
int main()
{
// Read in mic points from file
char micPointsChar[40] = "2,3.343,4.432\n";
float micPoints[3] = {};
initMicPoints(micPointsChar, micPoints);
for (int i = 1; i <= 3; i++)
{
Serial.print(micPoints[i]); // refer to original array, now initialized
Serial.print("\n");
}
return 0;
}
事实是,initMicPoints
会传入传入的数组,这就是你称之为init
的原因。捕获指针几乎没有用处,然后忽略范围内的原始数组。这只是将命令式代码打扮成功能性的,而没有潜在的语义。
在上面的代码中我们可以将数组转换为二维,而不会出现返回值类型问题;我们消除了它。