| |
Outline
Based on Java, describes the type of decimal floating point format of the program.
text
In Java, a class is used to format decimal java.text.DecimalFormat, for example, you can use:
double data = 3334590479.22d;
System.out.println (data); // 3.33459047922E9
java.text.DecimalFormat df = new java.text.DecimalFormat ( "$ #, ### ##.");
String result = df.format (data);
System.out.println (result); // $ 3,334,590,479.22
For larger values, mode uses scientific notation (scientific notation); use java.text.DecimalFormat specified format, with the specified format.
Format. "$ #, ### ##" Contains the following semantics:
1. $ prefix
2. Every 3-digit integer part of a group
3. Up to 2 decimal places
About format
A general structure of the format is: Prefixopt Number Suffixopt, namely the prefix + number + suffix; wherein the prefix and suffix are optional.
About Group
Group delimiter, and only the integer part; no matter how many there are delimiters mode, the rightmost that a valid; the number of each group is the rightmost delimiter right hand of integer digits.
Therefore, the "#, ##, ###, ####" == "###### ####" == "##, ####, ####."
About decimals
Decimal separator., Used to separate the integer and fractional parts.
About digital special characters
1. #
# Is a special character for the digital section; it represents a digital display, but the foremost and rearmost 0 not displayed.
data: 02222.20
pattern: "., ### ##"
result: 2,222.2
# There is no limit for the number of bits in the integer part, but for the fractional part, there are "many decimal places can only have the meaning of" excess rounding. such as:
data: 2222.222
pattern: "., ### ##"
result: 2,222.22
2.0
# 0 and usage is similar, but slightly different; except that:
A 0-digit integer part of the least restrictions arise;. Less than the median, preceded by 0
data: 22.26
pattern: "0,000.00"
result: 0,022.26
. B 0 limit number of fractional digits appear; if less than 0, more rounded
data: 22.26
-----------------
pattern: "00.000"
result: 22.260
-----------------
pattern: "00.0"
result: 22.3
3. E
Scientific notation, mantissa and exponent separator.
Details, see below.
About prefixes, suffixes
In addition to \ uFFFE, \ uFFFF special characters and any Unicode character can be used as a prefix, suffix; there are four special characters:
1. %
The number is multiplied by 100
data: 1234.567
pattern: "., ### ##%"
result: 123,456.7%
2. \ u2030
The number by 1000
data: 1234.567
pattern: "., ### ## \ u2030"
result: 1,234,567
3. (\ u00A4)
Localized currency symbol, if two consecutive, on behalf of the international currency symbol codes.
data: 1234.567
---------------------
pattern: "., ### ## "
result: 1,234.57
---------------------
pattern: "., ### ## "
result: 1,234.57
4. '
Used to refer to special characters, as a prefix or suffix.
data: 1234.567
pattern: ". '#', ### ##"
result: # 1,234.57
If you want to use the 'itself as a prefix or suffix, input twice in succession:
data: 1234.567
pattern: ". '', ### ##"
result: '1,234.57
Scientific notation
Scientific notation is added as E0 digital content in the tail section, where E is the mantissa and exponent separator, the number behind the E 0 represents the minimum number of indices (less than 0 up front).
data: 1E5
pattern: 0. # E00
result: 1E05 // Note that there are two indices
E content must be in front of, or at least a # 0, otherwise it is treated as a prefix E handled:
data: 1E5
---------------
pattern: "E0"
result: E100000
---------------
pattern: "0E0"
result: 1E5
---------------
pattern: "# E0"
result: .1E6
Exponent part is not much content, discussed below mantissa part.
Mantissa generally includes integers and fractions, if the integer part no content (no content front.), Valid on all the decimal numbers:
data: 12345E5
pattern: ". ### E0"
result: .123E10
If the integer part of the need to retain few, on the use of several 0 (# use the results of a little bizarre, all of you to try):
data: 12345E5
----------------------
pattern: ". 0 ### E0"
result: 1.234E9
----------------------
pattern: ". 00 ### E0"
result: 12.345E8
---------------------
pattern: ". 000 ### E0"
result: 123.45E7
Fractional part, if you want to achieve "up to several decimal places reserved" effect, in the pattern. After adding a # corresponding. This excess will be rounded up; if you want to achieve "just a few decimal reserved" effect, after the addition of the pattern of the corresponding zeros. Thus, the lack of will in the back fill 0, the excess will be rounded.
data: 123E5
------------------
pattern: "0. # E0"
result: 1.2E7
------------------
pattern: ". 0 ## E0"
result: 1.23E7
------------------
pattern: ". 0 ### E0"
result: 1.23E7
==================
pattern: "0.0E0"
result: 1.2E7
------------------
pattern: "0.00E0"
result: 1.23E7
------------------
pattern: "0.000E0"
result: 1.230E7 |
|
|
|