Objective-C拓展了C,自然很多用法是和C一致的。比如浮点数转化成整数，就有以下四种情况。
1.简单粗暴，直接转化
float f = 1.5; int a; a = (int)f; NSLog("a = %d",a);输出结果是1。（int）是强制类型转化，丢弃浮点数的小数部分。2.高斯函数，向下取整
float f = 1.6; int a; a = floor(f); NSLog("a = %d",a);输出结果是1。floor()方法是向下取整，类似于数学中的高斯函数 [].取得不大于浮点数的最大整数，对于正数来说是舍弃浮点数部分，对于复数来说，舍弃浮点数部分后再减1.3.ceil函数，向上取整。
float f = 1.5; int a; a = ceil(f); NSLog("a = %d",a);输出结果是2。ceil()方法是向上取整，取得不小于浮点数的最小整数，对于正数来说是舍弃浮点数部分并加1，对于复数来说就是舍弃浮点数部分.4.通过强制类型转换四舍五入。
float f = 1.5; int a; a = (int)(f+0.5); NSLog("a = %d",a);其中原理非常简单，所以就不做详细说明了。在iOS开发中，和货币价格计算相关的，需要注意计算精度的问题。即使只是两位小数，也会出现误差。使用float类型运算，是完全不够的。经过一番测试，最后选择使用系统提供的API的NSDecimalNumber来进行更好的解决。作为一个对外的库，鉴于版本延续，我们保留对外的flaot的类型，不改变接口，选择进行内部适配。以下是一些基本的测试，原始数据
| float a = 0.01;int b = 99999999;double c = 0.0; |
1：使用浮点运算，
| c = a*b;NSLog(@"%f",c);NSLog(@"%.2f",c); |
使用double类型存储没有触及问题的实质，完全不能解决。
| 2011-12-30 11:04:00.121 Untitled[2912:207] 1000000.0000002011-12-30 11:04:00.123 Untitled[2912:207] 1000000.00 |
2：使用类型转换，提高精度
| c = a*(double)b;NSLog(@"%f",c);NSLog(@"%.2f",c); |
Double运算的精度是提高了，可是浮点数的数值早已经出现了精度的不准确，即使存储空间足够，同样还是不准确的数值。
| 2011-12-30 11:04:00.123 Untitled[2912:207] 999999.9676482011-12-30 11:04:00.124 Untitled[2912:207] 999999.97 |
3：通过和NSString的转换，将计算的原始数据转换为纯粹的double类型的数据，这样的计算精度就可以达到要求了。
| NSString *objA = [NSString stringWithFormat:@"%.2f", a];NSString *objB = [NSString stringWithFormat:@"%.2f", (double)b];c = [objA doubleValue] * [objB doubleValue];NSLog(@"%.2f",c); |
计算的结果还是比较准确的，不过需要做格式化输入和格式化输出的处理。同时使用NSString来转换，这样的写法看起来比较奇怪。
| 2011-12-30 11:04:00.190 Untitled[2912:207] 999999.99 |
4：个人还是比较喜欢使用系统提供的类型来进行计算。通过NSDecimalNumber提供的计算方式，可以很好的计算出准确的精度的数据，同时不需要使用格式化输出等。其计算的精度是比较高，这是官方建议的货币计算的API，对乘除等计算都有单独的API接口来提供。
| NSString *decimalNumberMutiplyWithString(NSString *multiplierValue,NSString *multiplicandValue){ NSDecimalNumber *multiplierNumber = [NSDecimalNumber decimalNumberWithString:multiplierValue]; NSDecimalNumber *multiplicandNumber = [NSDecimalNumber decimalNumberWithString:multiplicandValue]; NSDecimalNumber *product = [multiplicandNumber decimalNumberByMultiplyingBy:multiplierNumber]; return [product stringValue];} NSLog(@"%@",decimalNumberMutiplyWithString([NSString stringWithFormat:@"%f",a], [NSString stringWithFormat:@"%d",b])); |
| 2011-12-30 11:04:00.251 Untitled[2912:207] 999999.99 |
只是测试，所以接口名大致写写，名字取得比较不那么讲究，希望可以表达清楚。 总的来说，对于货币计算，应该需要注意精度的问题。同时在运算的时候，应该优先选用框架提供的API，否则，就应该使用足够精度的类型运算，同时对自己写的接口进行足够的说明，要求开发者按照规范来使用。在自己不能保证足够准确的情况下，用适当的说明的要求来规避责任还是可以接受的。至少被人抱怨两句总比出错强。
elf.orderCost.text = [NSStringstringWithFormat:@"%.1f元",self.order.cost.floatValue];%.1f 表示小数点一位，%.2f 表示小数点2位，依次类推.
格式定义The format specifiers supported by the NSString formatting methods and CFString formatting functions follow the IEEE printf specification; the specifiers are summarized in Table 1. Note that you can also use the “n$” positional specifiers such as %1$@ %2$s. For more details, see the IEEE printf specification. You can also use these format specifiers with the NSLog function.
Table 1 Format specifiers supported by the NSString formatting methods and CFString formatting functions| 定义 | 说明 |
| %@ | Objective-C object, printed as the string returned by descriptionWithLocale: if available, or description otherwise. Also works with CFTypeRef objects, returning the result of the CFCopyDescription function. |
| %% | ‘%’ character |
| %d, %D, %i | Signed 32-bit integer (int) |
| %u, %U | Unsigned 32-bit integer (unsigned int) |
| %hi | Signed 16-bit integer (short) |
| %hu | Unsigned 16-bit integer (unsigned short) |
| %qi | Signed 64-bit integer (long long) |
| %qu | Unsigned 64-bit integer (unsigned long long) |
| %x | Unsigned 32-bit integer (unsigned int), printed in hexadecimal using the digits 0–9 and lowercase a–f |
| %X | Unsigned 32-bit integer (unsigned int), printed in hexadecimal using the digits 0–9 and uppercase A–F |
| %qx | Unsigned 64-bit integer (unsigned long long), printed in hexadecimal using the digits 0–9 and lowercase a–f |
| %qX | Unsigned 64-bit integer (unsigned long long), printed in hexadecimal using the digits 0–9 and uppercase A–F |
| %o, %O | Unsigned 32-bit integer (unsigned int), printed in octal |
| %f | 64-bit floating-point number (double) |
| %e | 64-bit floating-point number (double), printed in scientific notation using a lowercase e to introduce the exponent |
| %E | 64-bit floating-point number (double), printed in scientific notation using an uppercase E to introduce the exponent |
| %g | 64-bit floating-point number (double), printed in the style of %e if the exponent is less than –4 or greater than or equal to the precision, in the style of %f otherwise |
| %G | 64-bit floating-point number (double), printed in the style of %E if the exponent is less than –4 or greater than or equal to the precision, in the style of %f otherwise |
| %c | 8-bit unsigned character (unsigned char), printed by NSLog() as an ASCII character, or, if not an ASCII character, in the octal format \\ddd or the Unicode hexadecimal format \\udddd, where d is a digit |
| %C | 16-bit Unicode character (unichar), printed by NSLog() as an ASCII character, or, if not an ASCII character, in the octal format \\ddd or the Unicode hexadecimal format \\udddd, where d is a digit |
| %s | Null-terminated array of 8-bit unsigned characters. %s interprets its input in the system encoding rather than, for example, UTF-8. |
| %S | Null-terminated array of 16-bit Unicode characters |
| %p | Void pointer (void *), printed in hexadecimal with the digits 0–9 and lowercase a–f, with a leading 0x |
| %L | Length modifier specifying that a following a, A, e, E, f, F, g, or G conversion specifier applies to a long double argument |
| %a | 64-bit floating-point number (double), printed in scientific notation with a leading 0x and one hexadecimal digit before the decimal point using a lowercase p to introduce the exponent |
| %A | 64-bit floating-point number (double), printed in scientific notation with a leading 0X and one hexadecimal digit before the decimal point using a uppercase P to introduce the exponent |
| %F | 64-bit floating-point number (double), printed in decimal notation |
| %z | Length modifier specifying that a following d, i, o, u, x, or X conversion specifier applies to a size_t or the corresponding signed integer type argument |
| %t | Length modifier specifying that a following d, i, o, u, x, or X conversion specifier applies to a ptrdiff_t or the corresponding unsigned integer type argument |
| %j | Length modifier specifying that a following d, i, o, u, x, or X conversion specifier applies to a intmax_t or uintmax_t argument |
平台依赖Mac OS X uses several data types—NSInteger, NSUInteger,CGFloat, and CFIndex—to provide a consistent means of representing values in 32- and 64-bit environments. In a 32-bit environment, NSInteger and NSUInteger are defined as int and unsigned int, respectively. In 64-bit environments, NSInteger and NSUInteger are defined as long and unsigned long, respectively. To avoid the need to use different printf-style type specifiers depending on the platform, you can use the specifiers shown in Table 2. Note that in some cases you may have to cast the value.
Table 2 Format specifiers for data types| 类型 | 定义 | 建议 |
| NSInteger | %ld or %lx | Cast the value to long |
| NSUInteger | %lu or %lx | Cast the value to unsigned long |
| CGFloat | %f or %g | %f works for floats and doubles when formatting; but see below warning when scanning |
| CFIndex | %ld or %lx | The same as NSInteger |
| pointer | %p | %p adds 0x to the beginning of the output. If you don’t want that, use %lx and cast to long. |
| long long | %lld or %llx | long long is 64-bit on both 32- and 64-bit platforms |
| unsigned long long | %llu or %llx | unsigned long long is 64-bit on both 32- and 64-bit platforms |
The following example illustrates the use of %ld to format an NSInteger and the use of a cast.
1
2 | NSInteger i = 42;
printf("%ld\n", (long)i);
|
In addition to the considerations mentioned in Table 2, there is one extra case with scanning: you must distinguish the types for float and double. You should use %f for float, %lf for double. If you need to use scanf (or a variant thereof) with CGFloat, switch to double instead, and copy the double to CGFloat.
1
2
3
4 | CGFloat imageWidth;
double tmp;
sscanf (str, "%lf", &tmp);
imageWidth = tmp;
|
It is important to remember that %lf does not represent CGFloat correctly on either 32- or 64-bit platforms. This is unlike %ld, which works for long in all cases.
