C# 用千位分隔符格式化数字，但保留小数位数

Question

I want to add a thousands separator to a double number but want to keep the decimal places as is ie dont want any rounding.

#,# is solving my problem of adding the thousand separator but how do I preserve the decimal places? #,# strips off the part after . .

I cannot use any culture or something like that & the developer whose function I am calling has only given me a way of changing the format by passing as parameter strFormat .

I did check other posts & even the docs but somehow not able to figure this out.

string strFormat = "#,#";
string str = double.parse("912123456.1123465789").ToString(strFormat);
 //Expected here 912,123,456.1123465789 
//Actual Output 912,123,456

//912123456.123 should give 912,123,456.123
//912123456.1 should give 912,123,456.1
//912123456.1123465789 should give 912,123,456.1123465789
//912123456 should give 912,123,456

Answer 1

If you know the max number of decimal places, eg 10, then use:

string strFormat = "#,#0.##########";

Update:

This max number is known. According to Microsoft documentation a Double value has up to 15 decimal digits of precision (including both before and after the decimal point). More than 15 digits will be rounded.

So if you must invoke that method of 'double.parse' and can only send the format, this is the best you can do:

string strFormat = "#,#0.###############";

Answer 2

Well, after looking at the documentation on Microsoft , it would appear that there is no particular way to allow a floating point position in a number - all characters in a format string are character placeholders.

I would recommend that you either use a very nasty predetermined number of # s to set the width of the decimal position, or the slightly less (or possibly more, depending on your outlook) nasty option of reading all numbers into an array, determining the longest decimal position, then building a format of # s using the result.

At the end of the day, this is a single format string that you can put into place, test and ensure it works, then come back later and fix if you find a better alternative.

Also, this is one of those things where you could put the string into a configuration setting and change as and when you need to - far more flexible.

To be honest, this is a very slight thing to be worried about in the grand scheme of performance and writing a program.

Technically, Udi Y gets my vote!

Answer 3

You can calculate the formatting dynamically for each number:

public static void Main()
{
    var number = 1234.12312323123;

    var format = GetNumberFormat(number);
    Console.WriteLine(number.ToString(format));
}

public static string GetNumberFormat(double number)
{
    var numberAsString = number.ToString();

    var decimalPartSize = numberAsString.Substring(numberAsString.LastIndexOf('.') + 1).Length;

    return $"N{decimalPartSize}";
}

So

number = 1234.12312323123

will give you 1,234.12312323123 . Works for negative numbers as well. Also, as we work with strings, there won't be any rounding errors or precision artifacts.