冷算法:自动生成代码标识符(类名、方法名、变量名)
竟然有小伙伴喜欢在编写代码时使用随机字符当作类名、方法名、变量名,例如这一篇博客里的代码:使用 Resharper 特性 - 林德熙。既然随机,那也随机得像一些啊!于是我改进了标识符的随机算法,使得生成的标识符更像真实单词的组合。
看看标识符的生成效果吧!0、2、4……行是 PascalCase,即首字母大写的;1、3、5……行是 camelCase 即首字母小写的。
▲ 是不是感觉甚至能读出来?
嗯嗯,因为生成规则中考虑到了辅音和元音的组合,而且……嗯……还考虑到了部件出现的概率。
比如一个单词中的音节数,单音节概率 44%,双音节概率 31%,三音节概率 19%,四音节概率 6%。而这样的概率是通过一个幂函数来实现的。具体来说,是下面这个函数:
好吧,把我的源码放出来:
public class RandomIdentifier
{
private readonly Random _random = new Random();
public string Generate(bool pascal)
{
var builder = new StringBuilder();
var wordCount = _random.Next(2, 4);
for (var i = 0; i < wordCount; i++)
{
var syllableCount = 4 - (int) Math.Sqrt(_random.Next(0, 16));
for (var j = 0; j < syllableCount; j++)
{
var consonant = Consonants[_random.Next(Consonants.Count)];
var vowel = Vowels[_random.Next(Vowels.Count)];
if ((pascal || i != 0) && j == 0)
{
consonant = CultureInfo.CurrentCulture.TextInfo.ToTitleCase(consonant);
}
builder.Append(consonant);
builder.Append(vowel);
}
}
return builder.ToString();
}
private static readonly List<string> Consonants = new List<string>
{
"q","w","r","t","y","p","s","d","f","g","h","j","k","l","z","x","c","v","b","n","m",
"w","r","t","p","s","d","f","g","h","j","k","l","c","b","n","m",
"r","t","p","s","d","h","j","k","l","c","b","n","m",
"r","t","s","j","c","n","m",
"tr","dr","ch","wh","st",
"s","s"
};
private static readonly List<string> Vowels = new List<string>
{
"a","e","i","o","u",
"a","e","i","o","u",
"a","e","i",
"a","e",
"e",
"ar","as","ai","air","ay","al","all","aw",
"ee","ea","ear","em","er","el","ere",
"is","ir",
"ou","or","oo","ou","ow",
"ur"
};
}
而使用方法,则是简单的一个调用:
var identifier = new RandomIdentifier();
var pascal = _identifier.Generate(true);
var camel = _identifier.Generate(false);
传入 true
生成首字母大写的版本,传入 false
生成首字母小写的版本。