.net你怎么判断字符串是否全是数字

今天在写代码时突然想起测试经常用Microsoft.VisualBasic.Information.IsNumeric判断 url参数是否为数字时的这个方法的效率

因为数字是字符串是直接使用的，所以不需要转型，也就没有用tryparse

结果一测试吓一跳，这个方法的效率是如此的低，再测试了下tryparse还不错，正则的也比较差，

没什么技术含量，看结果吧：

 

先拓展下字符串：


public static class Common { //isDigit public static bool isNumberic1(this string _string) { if (string.IsNullOrEmpty(_string)) return false; foreach (char c in _string) { if (!char.IsDigit(c)) return false; } return true; } //vb isnumberic public static bool isNumberic2(this string _string) { return !string.IsNullOrEmpty(_string) && Microsoft.VisualBasic.Information.IsNumeric(_string); } //try parese public static bool isNumberic3(this string _string) { if (string.IsNullOrEmpty(_string)) return false; int i = 0; return int.TryParse(_string, out i); } //try catch public static bool isNumberic4(this string _string) { if (string.IsNullOrEmpty(_string)) return false; int i = 0; try { int.Parse(_string); } catch { return false; } return true; } //regex public static bool isNumberic5(this string _string) { return !string.IsNullOrEmpty(_string) && Regex.IsMatch(_string, "^//d+$"); } }

 

 

 

测试的代码：

 

 class Program { static void Main(string[] args) { Test("1234"); Test("1234a"); Test("a1234"); Test(""); Test(null); } static void Test(string str) { bool res1 = false, res2 = false, res3 = false, res4 = false, res5 = false; Stopwatch wat = new Stopwatch(); wat.Start(); for (int i = 1; i < 100000; i++) { res1 = str.isNumberic1(); } wat.Stop(); Console.WriteLine("isDigit {0}:{1},{2}", str, wat.ElapsedMilliseconds, res1); wat.Reset(); wat.Start(); for (int i = 1; i < 100000; i++) { res2= str.isNumberic2(); } wat.Stop(); Console.WriteLine("isNumberic {0}:{1},{2}", str, wat.ElapsedMilliseconds, res2); wat.Reset(); wat.Start(); for (int i = 1; i < 100000; i++) { res3 = str.isNumberic3(); } wat.Stop(); Console.WriteLine("try parse {0}:{1},{2}", str, wat.ElapsedMilliseconds, res3); wat.Reset(); wat.Start(); for (int i = 1; i < 100000; i++) { res4 = str.isNumberic4(); } wat.Stop(); Console.WriteLine("try catch {0}:{1},{2}", str, wat.ElapsedMilliseconds, res4); wat.Reset(); wat.Start(); for (int i = 1; i < 100000; i++) { res5 = str.isNumberic5(); } wat.Stop(); Console.WriteLine("regex {0}:{1},{2}", str, wat.ElapsedMilliseconds, res5); Console.WriteLine(); } }

 

下面是我本机的测试结果



isDigit      1234:0,True
isNumberic  1234:173,True
try parse    1234:21,True
try catch    1234:23,True
regex        1234:138,True

isDigit      1234a:0,False
isNumberic  1234a:204,False
try parse    1234a:20,False
try catch    1234a:5309,False
regex        1234a:151,False

isDigit      a1234:0,False
isNumberic  a1234:191,False
try parse    a1234:16,False
try catch    a1234:5229,False
regex        a1234:109,False

isDigit      :0,False
isNumberic  :0,False
try parse    :0,False
try catch    :1,False
regex        :0,False

isDigit      :0,False
isNumberic  :1,False
try parse    :0,False
try catch    :1,False
regex        :0,False

结果：循环判断是否是数字字符效率是最高的

而vbscript的方法效率比较低了

顺便测试了下vbscript里的left和right方法效率也一样的低，还不及substring的十分之一

所以vbscript里虽然有几个方法从名字看来比较好用，实际却比较杯具。