如何从 .NET 中的字符串中删除变音符号(重音符号)?

Posted

技术标签:

【中文标题】如何从 .NET 中的字符串中删除变音符号(重音符号)?【英文标题】:How do I remove diacritics (accents) from a string in .NET? 【发布时间】:2010-09-19 22:19:12 【问题描述】:

我正在尝试转换一些法语加拿大语的字符串,基本上,我希望能够在保留字母的同时去掉字母中的法语重音符号。 (例如,将é 转换为e,因此crème brûlée 将变为creme brulee

实现这一目标的最佳方法是什么?

【问题讨论】:

警告:这种方法在某些特定情况下可能有效,但通常不能只删除变音符号。在某些情况下和某些语言中,这可能会改变文本的含义。你没有说你为什么要这样做;如果是为了比较字符串或搜索,最好使用支持 unicode 的库。 由于实现这一点的大多数技术都依赖于 Unicode 规范化,因此描述该标准的文档可能对阅读有用:unicode.org/reports/tr15 我认为Azure团队修复了这个问题,我尝试上传一个名为“Mémo de la réunion.pdf”的文件,操作成功。 在我们的例子中,限制来自 Postgres 数据库中的 ltree 数据类型。其中 ltree 只允许[a-zA-Z0-9_]。对于我们的案例,确实有必要进行快速搜索。 【参考方案1】:

我没有使用过这种方法,但 Michael Kaplan 在他的博客文章(标题令人困惑)中描述了一种这样做的方法,其中谈到了剥离变音符号:Stripping is an interesting job (aka On the meaning of meaningless, aka All Mn characters are non-spacing, but some are more non-spacing than others)

static string RemoveDiacritics(string text) 

    var normalizedString = text.Normalize(NormalizationForm.FormD);
    var stringBuilder = new StringBuilder(capacity: normalizedString.Length);

    for (int i = 0; i < normalizedString.Length; i++)
    
        char c = normalizedString[i];
        var unicodeCategory = CharUnicodeInfo.GetUnicodeCategory(c);
        if (unicodeCategory != UnicodeCategory.NonSpacingMark)
        
            stringBuilder.Append(c);
        
    

    return stringBuilder
        .ToString()
        .Normalize(NormalizationForm.FormC);

请注意,这是他之前帖子的后续:Stripping diacritics....

该方法使用String.Normalize 将输入字符串拆分为组成字形(基本上将“基本”字符与变音符号分开),然后扫描结果并仅保留基本字符。这只是有点复杂,但实际上你正在研究一个复杂的问题。

当然,如果您限制自己使用法语,您可能会按照@David Dibben 的建议使用 How to remove accents and tilde in a C++ std::string 中基于表格的简单方法。

【讨论】:

这是错误的。德语字符 ä 和 ö 和 ü 拉丁化为 ae ue 和 oe,而不是 a, o u ... 另外,波兰字母 ł 被忽略。 Norse ø 也被忽略 @StefanSteiger 你知道,在捷克语中有 áčěů 之类的字母,我们通常将其“拉丁化”为 aceu,尽管它听起来不同,并且可能会在诸如“hrábě”/hra:bje/ 之类的词中引起混淆、“hrabě”/hrabje/ 和“hrabe”/hrabe/。对我来说,删除变音符号似乎是一个纯粹的图形问题,与字母的语音或历史无关。像ä ö ü 这样的字母是通过在基本字母中添加上标“e”来创建的,因此“ae”分解在历史上是有意义的。这取决于目标 - 删除图形标记,或将字母分解为 ASCII 字符。 此函数与语言无关。它不知道字符串是德语还是其他语言。如果我们考虑到在德语文本中用 oe 替换 ö 是可以的,但是用土耳其语这样做没有任何意义,那么我们会发现,如果不检测语言,这个问题实际上是无法解决的。 【参考方案2】:

如果有人感兴趣,我正在寻找类似的东西并结束了以下内容:

public static string NormalizeStringForUrl(string name)

    String normalizedString = name.Normalize(NormalizationForm.FormD);
    StringBuilder stringBuilder = new StringBuilder();

    foreach (char c in normalizedString)
    
        switch (CharUnicodeInfo.GetUnicodeCategory(c))
        
            case UnicodeCategory.LowercaseLetter:
            case UnicodeCategory.UppercaseLetter:
            case UnicodeCategory.DecimalDigitNumber:
                stringBuilder.Append(c);
                break;
            case UnicodeCategory.SpaceSeparator:
            case UnicodeCategory.ConnectorPunctuation:
            case UnicodeCategory.DashPunctuation:
                stringBuilder.Append('_');
                break;
        
    
    string result = stringBuilder.ToString();
    return String.Join("_", result.Split(new char[]  '_' 
        , StringSplitOptions.RemoveEmptyEntries)); // remove duplicate underscores

【讨论】:

您应该将 StringBuilder 缓冲区预分配给 name.Length 以最小化内存分配开销。最后一次拆分/加入调用以删除顺序重复的 _ 很有趣。也许我们应该避免在循环中添加它们。为前一个字符设置一个标志,如果为真,则不发出一个 _。 2 点非常好,如果我有时间回到这部分代码,我会重写它:) 不错。除了 IDisposables 注释之外,我们可能还应该检查 c &lt; 128,以确保我们不会拾取任何 UTF 字符 see here。 或者可能更有效c &lt; 123。 see ASCI【参考方案3】:

这对我有用...

string accentedStr;
byte[] tempBytes;
tempBytes = System.Text.Encoding.GetEncoding("ISO-8859-8").GetBytes(accentedStr);
string asciiStr = System.Text.Encoding.UTF8.GetString(tempBytes);

快&短!

【讨论】:

这是我见过的最好的方法。 我很喜欢这个解决方案,它适用于 Windows 应用商店应用程序。但是,它不适用于 Windows Phone 应用程序,因为编码 ISO-8859-8 似乎不可用。是否可以使用其他编码? 这将适用于最常见的字符,但许多特殊字符,如 « »(作为单个字符)将在此过程中发生变化,但情况并非如此接受的解决方案。 请注意,这不适用于 Linux 上的 .NET Core:System.ArgumentException: 'ISO-8859-8' is not a supported encoding name. 如果您在 .NET Core 上,请从 nuget 安装 System.Text.Encoding.CodePages,然后调用它来注册提供程序:Encoding.RegisterProvider(CodePagesEncodingProvider.Instance); - 完成此操作后,您可以使用 ISO- 8859-8【参考方案4】:

这是 VB 版本(适用于希腊语):

导入 System.Text

导入 System.Globalization

Public Function RemoveDiacritics(ByVal s As String)
    Dim normalizedString As String
    Dim stringBuilder As New StringBuilder
    normalizedString = s.Normalize(NormalizationForm.FormD)
    Dim i As Integer
    Dim c As Char
    For i = 0 To normalizedString.Length - 1
        c = normalizedString(i)
        If CharUnicodeInfo.GetUnicodeCategory(c) <> UnicodeCategory.NonSpacingMark Then
            stringBuilder.Append(c)
        End If
    Next
    Return stringBuilder.ToString()
End Function

【讨论】:

可能是一个旧答案,但为什么要使用单独的行进行变量声明和第一次赋值?【参考方案5】:

我经常使用基于我在这里找到的另一个版本的扩展方法 (见Replacing characters in C# (ascii)) 快速解释:

规范化以形成 D 将 è 等字符拆分为 e 和非间距 ` 从此,nospacing 字符被删除 结果被归一化回形式 C(我不确定这是否必要)

代码:

using System.Linq;
using System.Text;
using System.Globalization;

// namespace here
public static class Utility

    public static string RemoveDiacritics(this string str)
    
        if (null == str) return null;
        var chars =
            from c in str.Normalize(NormalizationForm.FormD).ToCharArray()
            let uc = CharUnicodeInfo.GetUnicodeCategory(c)
            where uc != UnicodeCategory.NonSpacingMark
            select c;

        var cleanStr = new string(chars.ToArray()).Normalize(NormalizationForm.FormC);

        return cleanStr;
    

    // or, alternatively
    public static string RemoveDiacritics2(this string str)
    
        if (null == str) return null;
        var chars = str
            .Normalize(NormalizationForm.FormD)
            .ToCharArray()
            .Where(c=> CharUnicodeInfo.GetUnicodeCategory(c) != UnicodeCategory.NonSpacingMark)
            .ToArray();

        return new string(chars).Normalize(NormalizationForm.FormC);
    

【讨论】:

【参考方案6】:

试试HelperSharp package。

有一个方法RemoveAccents:

 public static string RemoveAccents(this string source)
 
     //8 bit characters 
     byte[] b = Encoding.GetEncoding(1251).GetBytes(source);

     // 7 bit characters
     string t = Encoding.ASCII.GetString(b);
     Regex re = new Regex("[^a-zA-Z0-9]=-_/");
     string c = re.Replace(t, " ");
     return c;
 

【讨论】:

【参考方案7】:

这就是我在所有 .NET 程序中将变音符号替换为非变音符号的方法

C#:

//Transforms the culture of a letter to its equivalent representation in the 0-127 ascii table, such as the letter 'é' is substituted by an 'e'
public string RemoveDiacritics(string s)

    string normalizedString = null;
    StringBuilder stringBuilder = new StringBuilder();
    normalizedString = s.Normalize(NormalizationForm.FormD);
    int i = 0;
    char c = '\0';

    for (i = 0; i <= normalizedString.Length - 1; i++)
    
        c = normalizedString[i];
        if (CharUnicodeInfo.GetUnicodeCategory(c) != UnicodeCategory.NonSpacingMark)
        
            stringBuilder.Append(c);
        
    

    return stringBuilder.ToString().ToLower();

VB .NET:

'Transforms the culture of a letter to its equivalent representation in the 0-127 ascii table, such as the letter "é" is substituted by an "e"'
Public Function RemoveDiacritics(ByVal s As String) As String
    Dim normalizedString As String
    Dim stringBuilder As New StringBuilder
    normalizedString = s.Normalize(NormalizationForm.FormD)
    Dim i As Integer
    Dim c As Char

    For i = 0 To normalizedString.Length - 1
        c = normalizedString(i)
        If CharUnicodeInfo.GetUnicodeCategory(c) <> UnicodeCategory.NonSpacingMark Then
            stringBuilder.Append(c)
        End If
    Next
    Return stringBuilder.ToString().ToLower()
End Function

【讨论】:

【参考方案8】:

您可以使用 MMLib.Extensions nuget 包中的字符串扩展:

using MMLib.RapidPrototyping.Generators;
public void ExtensionsExample()

  string target = "aácčeéií";
  Assert.AreEqual("aacceeii", target.RemoveDiacritics());
 

Nuget 页面:https://www.nuget.org/packages/MMLib.Extensions/ Codeplex项目现场https://mmlib.codeplex.com/

【讨论】:

无源代码;掘金包指向一个无效的 url,托管在 codeplex【参考方案9】:
Imports System.Text
Imports System.Globalization

 Public Function DECODE(ByVal x As String) As String
        Dim sb As New StringBuilder
        For Each c As Char In x.Normalize(NormalizationForm.FormD).Where(Function(a) CharUnicodeInfo.GetUnicodeCategory(a) <> UnicodeCategory.NonSpacingMark)  
            sb.Append(c)
        Next
        Return sb.ToString()
    End Function

【讨论】:

使用 NFD 代替 NFC 会导致远远超出要求的更改。【参考方案10】:

What this person said:

Encoding.ASCII.GetString(Encoding.GetEncoding(1251).GetBytes(text));

它实际上将å 之类的字符拆分为@ 987654327@加上某种修饰符,然后ASCII转换去掉修饰符,只剩下a

【讨论】:

【参考方案11】:

我需要一些东西来转换所有主要的 unicode 字符并且投票的答案遗漏了一些,所以我创建了 CodeIgniter 的 convert_accented_characters($str) 的一个版本,它可以轻松定制为 C#:

using System;
using System.Text;
using System.Collections.Generic;

public static class Strings

    static Dictionary<string, string> foreign_characters = new Dictionary<string, string>
    
         "äæǽ", "ae" ,
         "öœ", "oe" ,
         "ü", "ue" ,
         "Ä", "Ae" ,
         "Ü", "Ue" ,
         "Ö", "Oe" ,
         "ÀÁÂÃÄÅǺĀĂĄǍΑΆẢẠẦẪẨẬẰẮẴẲẶА", "A" ,
         "àáâãåǻāăąǎªαάảạầấẫẩậằắẵẳặа", "a" ,
         "Б", "B" ,
         "б", "b" ,
         "ÇĆĈĊČ", "C" ,
         "çćĉċč", "c" ,
         "Д", "D" ,
         "д", "d" ,
         "ÐĎĐΔ", "Dj" ,
         "ðďđδ", "dj" ,
         "ÈÉÊËĒĔĖĘĚΕΈẼẺẸỀẾỄỂỆЕЭ", "E" ,
         "èéêëēĕėęěέεẽẻẹềếễểệеэ", "e" ,
         "Ф", "F" ,
         "ф", "f" ,
         "ĜĞĠĢΓГҐ", "G" ,
         "ĝğġģγгґ", "g" ,
         "ĤĦ", "H" ,
         "ĥħ", "h" ,
         "ÌÍÎÏĨĪĬǏĮİΗΉΊΙΪỈỊИЫ", "I" ,
         "ìíîïĩīĭǐįıηήίιϊỉịиыї", "i" ,
         "Ĵ", "J" ,
         "ĵ", "j" ,
         "ĶΚК", "K" ,
         "ķκк", "k" ,
         "ĹĻĽĿŁΛЛ", "L" ,
         "ĺļľŀłλл", "l" ,
         "М", "M" ,
         "м", "m" ,
         "ÑŃŅŇΝН", "N" ,
         "ñńņňʼnνн", "n" ,
         "ÒÓÔÕŌŎǑŐƠØǾΟΌΩΏỎỌỒỐỖỔỘỜỚỠỞỢО", "O" ,
         "òóôõōŏǒőơøǿºοόωώỏọồốỗổộờớỡởợо", "o" ,
         "П", "P" ,
         "п", "p" ,
         "ŔŖŘΡР", "R" ,
         "ŕŗřρр", "r" ,
         "ŚŜŞȘŠΣС", "S" ,
         "śŝşșšſσςс", "s" ,
         "ȚŢŤŦτТ", "T" ,
         "țţťŧт", "t" ,
         "ÙÚÛŨŪŬŮŰŲƯǓǕǗǙǛŨỦỤỪỨỮỬỰУ", "U" ,
         "ùúûũūŭůűųưǔǖǘǚǜυύϋủụừứữửựу", "u" ,
         "ÝŸŶΥΎΫỲỸỶỴЙ", "Y" ,
         "ýÿŷỳỹỷỵй", "y" ,
         "В", "V" ,
         "в", "v" ,
         "Ŵ", "W" ,
         "ŵ", "w" ,
         "ŹŻŽΖЗ", "Z" ,
         "źżžζз", "z" ,
         "ÆǼ", "AE" ,
         "ß", "ss" ,
         "IJ", "IJ" ,
         "ij", "ij" ,
         "Œ", "OE" ,
         "ƒ", "f" ,
         "ξ", "ks" ,
         "π", "p" ,
         "β", "v" ,
         "μ", "m" ,
         "ψ", "ps" ,
         "Ё", "Yo" ,
         "ё", "yo" ,
         "Є", "Ye" ,
         "є", "ye" ,
         "Ї", "Yi" ,
         "Ж", "Zh" ,
         "ж", "zh" ,
         "Х", "Kh" ,
         "х", "kh" ,
         "Ц", "Ts" ,
         "ц", "ts" ,
         "Ч", "Ch" ,
         "ч", "ch" ,
         "Ш", "Sh" ,
         "ш", "sh" ,
         "Щ", "Shch" ,
         "щ", "shch" ,
         "ЪъЬь", "" ,
         "Ю", "Yu" ,
         "ю", "yu" ,
         "Я", "Ya" ,
         "я", "ya" ,
    ;

    public static char RemoveDiacritics(this char c)
        foreach(KeyValuePair<string, string> entry in foreign_characters)
        
            if(entry.Key.IndexOf (c) != -1)
            
                return entry.Value[0];
            
        
        return c;
    

    public static string RemoveDiacritics(this string s) 
    
        //StringBuilder sb = new StringBuilder ();
        string text = "";


        foreach (char c in s)
        
            int len = text.Length;

            foreach(KeyValuePair<string, string> entry in foreign_characters)
            
                if(entry.Key.IndexOf (c) != -1)
                
                    text += entry.Value;
                    break;
                
            

            if (len == text.Length) 
                text += c;  
            
        
        return text;
    

用法

// for strings
"crème brûlée".RemoveDiacritics (); // creme brulee

// for chars
"Ã"[0].RemoveDiacritics (); // A

【讨论】:

您的实现可以完成这项工作,但在用于生产代码之前应该进行改进。 github.com/apache/lucenenet/blob/master/src/… 为什么不简单地将这个if (entry.Key.IndexOf(c) != -1)替换成if (entry.Key.Contains(c)) 我使用@Alexander 的链接在下面给出了答案:***.com/a/56797567/479701 我不明白为什么会有这么多的箍跳来使用 "äæǽ", "ae" 而不是 "ä", "ae" , "æ", "ae" , "ǽ", "ae" 而只是调用if (foreign_characters.TryGetValue(...)) ...。您已经完全破坏了字典已有索引的目的。【参考方案12】:

Greek (ISO)的CodePage可以做到

关于此代码页的信息在System.Text.Encoding.GetEncodings() 中。了解详情:https://msdn.microsoft.com/pt-br/library/system.text.encodinginfo.getencoding(v=vs.110).aspx

希腊语 (ISO) 有代码页 28597 和名称 iso-8859-7

转到代码... \o/

string text = "Você está numa situação lamentável";

string textEncode = System.Web.HttpUtility.UrlEncode(text, Encoding.GetEncoding("iso-8859-7"));
//result: "Voce+esta+numa+situacao+lamentavel"

string textDecode = System.Web.HttpUtility.UrlDecode(textEncode);
//result: "Voce esta numa situacao lamentavel"

所以,写这个函数……

public string RemoveAcentuation(string text)

    return
        System.Web.HttpUtility.UrlDecode(
            System.Web.HttpUtility.UrlEncode(
                text, Encoding.GetEncoding("iso-8859-7")));

注意...Encoding.GetEncoding("iso-8859-7") 等价于Encoding.GetEncoding(28597),因为第一个是名称,第二个是编码的代码页。

【讨论】:

太棒了!简短而高效! 好东西。我测试的几乎所有字符都通过了。 (äáčďěéíľľňôóřŕšťúůýž ÄÁČĎĚÉÍĽĽŇÔÓŘŔŠŤÚŮÝŽ ÖÜË łŁđĐ ţŢşŞçÇ øı)。只有ßə才发现问题,将其转换为?,但此类异常始终可以单独处理。在将其投入生产之前,最好对所有包含变音符号字母的 Unicode 区域进行测试。【参考方案13】:

有趣的是,这样的问题可以得到这么多答案,但没有一个符合我的要求 :) 周围有这么多语言,AFAIK 不可能提供完整的语言不可知论解决方案,因为其他人提到 FormC 或 FormD 是给出问题。

由于最初的问题与法语有关,因此最简单的工作答案确实是

    public static string ConvertWesternEuropeanToASCII(this string str)
    
        return Encoding.ASCII.GetString(Encoding.GetEncoding(1251).GetBytes(str));
    

1251应替换为输入语言的编码码。

然而,这仅用一个字符替换一个字符。由于我也使用德语作为输入,因此我进行了手动转换

    public static string LatinizeGermanCharacters(this string str)
    
        StringBuilder sb = new StringBuilder(str.Length);
        foreach (char c in str)
        
            switch (c)
            
                case 'ä':
                    sb.Append("ae");
                    break;
                case 'ö':
                    sb.Append("oe");
                    break;
                case 'ü':
                    sb.Append("ue");
                    break;
                case 'Ä':
                    sb.Append("Ae");
                    break;
                case 'Ö':
                    sb.Append("Oe");
                    break;
                case 'Ü':
                    sb.Append("Ue");
                    break;
                case 'ß':
                    sb.Append("ss");
                    break;
                default:
                    sb.Append(c);
                    break;
            
        
        return sb.ToString();
    

它可能无法提供最佳性能,但至少它非常易于阅读和扩展。 正则表达式是不行的,比任何字符/字符串的东西都要慢。

我还有一个很简单的去除空格的方法:

    public static string RemoveSpace(this string str)
    
        return str.Replace(" ", string.Empty);
    

最终,我使用了上述所有 3 个扩展的组合:

    public static string LatinizeAndConvertToASCII(this string str, bool keepSpace = false)
    
        str = str.LatinizeGermanCharacters().ConvertWesternEuropeanToASCII();            
        return keepSpace ? str : str.RemoveSpace();
    

以及成功通过的小单元测试(不详尽)。

    [TestMethod()]
    public void LatinizeAndConvertToASCIITest()
    
        string europeanStr = "Bonjour ça va? C'est l'été! Ich möchte ä Ä á à â ê é è ë Ë É ï Ï î í ì ó ò ô ö Ö Ü ü ù ú û Û ý Ý ç Ç ñ Ñ";
        string expected = "Bonjourcava?C'estl'ete!IchmoechteaeAeaaaeeeeEEiIiiiooooeOeUeueuuuUyYcCnN";
        string actual = europeanStr.LatinizeAndConvertToASCII();
        Assert.AreEqual(expected, actual);
    

【讨论】:

【参考方案14】:

我非常喜欢azrafe7 提供的简洁实用的代码。 所以,我稍微改动了一下,将其转换为扩展方法:

public static class StringExtensions

    public static string RemoveDiacritics(this string text)
    
        const string SINGLEBYTE_LATIN_ASCII_ENCODING = "ISO-8859-8";

        if (string.IsNullOrEmpty(text))
        
            return string.Empty;
        

        return Encoding.ASCII.GetString(
            Encoding.GetEncoding(SINGLEBYTE_LATIN_ASCII_ENCODING).GetBytes(text));
    

【讨论】:

这是唯一适用于所有波兰变音符号的方法。接受的答案不适用于 Ł 和 ł 字符。【参考方案15】:

如果您还没有考虑过,请在此处弹出此库。看起来有各种各样的单元测试。

https://github.com/thomasgalliker/Diacritics.NET

【讨论】:

【参考方案16】:

没有足够的声誉,显然我无法评论亚历山大的优秀链接。 - Lucene 似乎是在合理通用情况下工作的唯一解决方案。

对于那些想要一个简单的复制粘贴解决方案的人来说,这里就是利用 Lucene 中的代码:

string testbed = "ÁÂÄÅÇÉÍÎÓÖØÚÜÞàáââãåæçèéêëìíîïïðñóôööøúüāăčĐęğıŁłńŌōřŞşšźžžşțệủ";

Console.WriteLine(Lucene.latinizeLucene(testbed));

AAAACEIIOOOUUTHaaaaaaaaeeeeeiiiidnoooouuaacDegiLlnOorSsszzsteu

//////////

public static class Lucene

    // source: https://raw.githubusercontent.com/apache/lucenenet/master/src/Lucene.Net.Analysis.Common/Analysis/Miscellaneous/ASCIIFoldingFilter.cs
    // idea: https://***.com/questions/249087/how-do-i-remove-diacritics-accents-from-a-string-in-net (scroll down, search for lucene by Alexander)
    public static string latinizeLucene(string arg)
    
        char[] argChar = arg.ToCharArray();

        // latinizeLuceneImpl can expand one char up to four chars - e.g. Þ to TH, or æ to ae, or in fact ⑽ to (10)
        char[] resultChar = new String(' ', arg.Length * 4).ToCharArray();

        int outputPos = Lucene.latinizeLuceneImpl(argChar, 0, ref resultChar, 0, arg.Length);

        string ret = new string(resultChar);
        ret = ret.Substring(0, outputPos);

        return ret;
    

    /// <summary>
    /// Converts characters above ASCII to their ASCII equivalents.  For example,
    /// accents are removed from accented characters. 
    /// <para/>
    /// @lucene.internal
    /// </summary>
    /// <param name="input">     The characters to fold </param>
    /// <param name="inputPos">  Index of the first character to fold </param>
    /// <param name="output">    The result of the folding. Should be of size >= <c>length * 4</c>. </param>
    /// <param name="outputPos"> Index of output where to put the result of the folding </param>
    /// <param name="length">    The number of characters to fold </param>
    /// <returns> length of output </returns>
    private static int latinizeLuceneImpl(char[] input, int inputPos, ref char[] output, int outputPos, int length)
    
        int end = inputPos + length;
        for (int pos = inputPos; pos < end; ++pos)
        
            char c = input[pos];

            // Quick test: if it's not in range then just keep current character
            if (c < '\u0080')
            
                output[outputPos++] = c;
            
            else
            
                switch (c)
                
                    case '\u00C0': // À  [LATIN CAPITAL LETTER A WITH GRAVE]
                    case '\u00C1': // Á  [LATIN CAPITAL LETTER A WITH ACUTE]
                    case '\u00C2': // Â  [LATIN CAPITAL LETTER A WITH CIRCUMFLEX]
                    case '\u00C3': // Ã  [LATIN CAPITAL LETTER A WITH TILDE]
                    case '\u00C4': // Ä  [LATIN CAPITAL LETTER A WITH DIAERESIS]
                    case '\u00C5': // Å  [LATIN CAPITAL LETTER A WITH RING ABOVE]
                    case '\u0100': // Ā  [LATIN CAPITAL LETTER A WITH MACRON]
                    case '\u0102': // Ă  [LATIN CAPITAL LETTER A WITH BREVE]
                    case '\u0104': // Ą  [LATIN CAPITAL LETTER A WITH OGONEK]
                    case '\u018F': // Ə  http://en.wikipedia.org/wiki/Schwa  [LATIN CAPITAL LETTER SCHWA]
                    case '\u01CD': // Ǎ  [LATIN CAPITAL LETTER A WITH CARON]
                    case '\u01DE': // Ǟ  [LATIN CAPITAL LETTER A WITH DIAERESIS AND MACRON]
                    case '\u01E0': // Ǡ  [LATIN CAPITAL LETTER A WITH DOT ABOVE AND MACRON]
                    case '\u01FA': // Ǻ  [LATIN CAPITAL LETTER A WITH RING ABOVE AND ACUTE]
                    case '\u0200': // Ȁ  [LATIN CAPITAL LETTER A WITH DOUBLE GRAVE]
                    case '\u0202': // Ȃ  [LATIN CAPITAL LETTER A WITH INVERTED BREVE]
                    case '\u0226': // Ȧ  [LATIN CAPITAL LETTER A WITH DOT ABOVE]
                    case '\u023A': // Ⱥ  [LATIN CAPITAL LETTER A WITH STROKE]
                    case '\u1D00': // ᴀ  [LATIN LETTER SMALL CAPITAL A]
                    case '\u1E00': // Ḁ  [LATIN CAPITAL LETTER A WITH RING BELOW]
                    case '\u1EA0': // Ạ  [LATIN CAPITAL LETTER A WITH DOT BELOW]
                    case '\u1EA2': // Ả  [LATIN CAPITAL LETTER A WITH HOOK ABOVE]
                    case '\u1EA4': // Ấ  [LATIN CAPITAL LETTER A WITH CIRCUMFLEX AND ACUTE]
                    case '\u1EA6': // Ầ  [LATIN CAPITAL LETTER A WITH CIRCUMFLEX AND GRAVE]
                    case '\u1EA8': // Ẩ  [LATIN CAPITAL LETTER A WITH CIRCUMFLEX AND HOOK ABOVE]
                    case '\u1EAA': // Ẫ  [LATIN CAPITAL LETTER A WITH CIRCUMFLEX AND TILDE]
                    case '\u1EAC': // Ậ  [LATIN CAPITAL LETTER A WITH CIRCUMFLEX AND DOT BELOW]
                    case '\u1EAE': // Ắ  [LATIN CAPITAL LETTER A WITH BREVE AND ACUTE]
                    case '\u1EB0': // Ằ  [LATIN CAPITAL LETTER A WITH BREVE AND GRAVE]
                    case '\u1EB2': // Ẳ  [LATIN CAPITAL LETTER A WITH BREVE AND HOOK ABOVE]
                    case '\u1EB4': // Ẵ  [LATIN CAPITAL LETTER A WITH BREVE AND TILDE]
                    case '\u1EB6': // Ặ  [LATIN CAPITAL LETTER A WITH BREVE AND DOT BELOW]
                    case '\u24B6': // Ⓐ  [CIRCLED LATIN CAPITAL LETTER A]
                    case '\uFF21': // A  [FULLWIDTH LATIN CAPITAL LETTER A]
                        output[outputPos++] = 'A';
                        break;
                    case '\u00E0': // à  [LATIN SMALL LETTER A WITH GRAVE]
                    case '\u00E1': // á  [LATIN SMALL LETTER A WITH ACUTE]
                    case '\u00E2': // â  [LATIN SMALL LETTER A WITH CIRCUMFLEX]
                    case '\u00E3': // ã  [LATIN SMALL LETTER A WITH TILDE]
                    case '\u00E4': // ä  [LATIN SMALL LETTER A WITH DIAERESIS]
                    case '\u00E5': // å  [LATIN SMALL LETTER A WITH RING ABOVE]
                    case '\u0101': // ā  [LATIN SMALL LETTER A WITH MACRON]
                    case '\u0103': // ă  [LATIN SMALL LETTER A WITH BREVE]
                    case '\u0105': // ą  [LATIN SMALL LETTER A WITH OGONEK]
                    case '\u01CE': // ǎ  [LATIN SMALL LETTER A WITH CARON]
                    case '\u01DF': // ǟ  [LATIN SMALL LETTER A WITH DIAERESIS AND MACRON]
                    case '\u01E1': // ǡ  [LATIN SMALL LETTER A WITH DOT ABOVE AND MACRON]
                    case '\u01FB': // ǻ  [LATIN SMALL LETTER A WITH RING ABOVE AND ACUTE]
                    case '\u0201': // ȁ  [LATIN SMALL LETTER A WITH DOUBLE GRAVE]
                    case '\u0203': // ȃ  [LATIN SMALL LETTER A WITH INVERTED BREVE]
                    case '\u0227': // ȧ  [LATIN SMALL LETTER A WITH DOT ABOVE]
                    case '\u0250': // ɐ  [LATIN SMALL LETTER TURNED A]
                    case '\u0259': // ə  [LATIN SMALL LETTER SCHWA]
                    case '\u025A': // ɚ  [LATIN SMALL LETTER SCHWA WITH HOOK]
                    case '\u1D8F': // ᶏ  [LATIN SMALL LETTER A WITH RETROFLEX HOOK]
                    case '\u1D95': // ᶕ  [LATIN SMALL LETTER SCHWA WITH RETROFLEX HOOK]
                    case '\u1E01': // ạ  [LATIN SMALL LETTER A WITH RING BELOW]
                    case '\u1E9A': // ả  [LATIN SMALL LETTER A WITH RIGHT HALF RING]
                    case '\u1EA1': // ạ  [LATIN SMALL LETTER A WITH DOT BELOW]
                    case '\u1EA3': // ả  [LATIN SMALL LETTER A WITH HOOK ABOVE]
                    case '\u1EA5': // ấ  [LATIN SMALL LETTER A WITH CIRCUMFLEX AND ACUTE]
                    case '\u1EA7': // ầ  [LATIN SMALL LETTER A WITH CIRCUMFLEX AND GRAVE]
                    case '\u1EA9': // ẩ  [LATIN SMALL LETTER A WITH CIRCUMFLEX AND HOOK ABOVE]
                    case '\u1EAB': // ẫ  [LATIN SMALL LETTER A WITH CIRCUMFLEX AND TILDE]
                    case '\u1EAD': // ậ  [LATIN SMALL LETTER A WITH CIRCUMFLEX AND DOT BELOW]
                    case '\u1EAF': // ắ  [LATIN SMALL LETTER A WITH BREVE AND ACUTE]
                    case '\u1EB1': // ằ  [LATIN SMALL LETTER A WITH BREVE AND GRAVE]
                    case '\u1EB3': // ẳ  [LATIN SMALL LETTER A WITH BREVE AND HOOK ABOVE]
                    case '\u1EB5': // ẵ  [LATIN SMALL LETTER A WITH BREVE AND TILDE]
                    case '\u1EB7': // ặ  [LATIN SMALL LETTER A WITH BREVE AND DOT BELOW]
                    case '\u2090': // ₐ  [LATIN SUBSCRIPT SMALL LETTER A]
                    case '\u2094': // ₔ  [LATIN SUBSCRIPT SMALL LETTER SCHWA]
                    case '\u24D0': // ⓐ  [CIRCLED LATIN SMALL LETTER A]
                    case '\u2C65': // ⱥ  [LATIN SMALL LETTER A WITH STROKE]
                    case '\u2C6F': // Ɐ  [LATIN CAPITAL LETTER TURNED A]
                    case '\uFF41': // a  [FULLWIDTH LATIN SMALL LETTER A]
                        output[outputPos++] = 'a';
                        break;
                    case '\uA732': // Ꜳ  [LATIN CAPITAL LETTER AA]
                        output[outputPos++] = 'A';
                        output[outputPos++] = 'A';
                        break;
                    case '\u00C6': // Æ  [LATIN CAPITAL LETTER AE]
                    case '\u01E2': // Ǣ  [LATIN CAPITAL LETTER AE WITH MACRON]
                    case '\u01FC': // Ǽ  [LATIN CAPITAL LETTER AE WITH ACUTE]
                    case '\u1D01': // ᴁ  [LATIN LETTER SMALL CAPITAL AE]
                        output[outputPos++] = 'A';
                        output[outputPos++] = 'E';
                        break;
                    case '\uA734': // Ꜵ  [LATIN CAPITAL LETTER AO]
                        output[outputPos++] = 'A';
                        output[outputPos++] = 'O';
                        break;
                    case '\uA736': // Ꜷ  [LATIN CAPITAL LETTER AU]
                        output[outputPos++] = 'A';
                        output[outputPos++] = 'U';
                        break;

        // etc. etc. etc.
        // see link above for complete source code
        // 
        // unfortunately, postings are limited, as in
        // "Body is limited to 30000 characters; you entered 136098."

                    [...]

                    case '\u2053': // ⁓  [SWUNG DASH]
                    case '\uFF5E': // ~  [FULLWIDTH TILDE]
                        output[outputPos++] = '~';
                        break;
                    default:
                        output[outputPos++] = c;
                        break;
                
            
        
        return outputPos;
    

【讨论】:

【参考方案17】:

TL;DR - C# string extension method

我认为保留字符串含义的最佳解决方案是转换字符而不是剥离它们,这在 crème brûléecrme brlecreme brulee 的示例中得到了很好的说明。

我查看了Alexander's comment above 并看到 Lucene.Net 代码是 Apache 2.0 许可的,所以我将这个类修改为一个简单的字符串扩展方法。你可以这样使用它:

var originalString = "crème brûlée";
var maxLength = originalString.Length; // limit output length as necessary
var foldedString = originalString.FoldToASCII(maxLength); 
// "creme brulee"

函数太长,无法在 *** 中发布答案(允许 30k 的约 139k 个字符,哈哈)所以I made a gist and attributed the authors:

/*
 * Licensed to the Apache Software Foundation (ASF) under one or more
 * contributor license agreements.  See the NOTICE file distributed with
 * this work for additional information regarding copyright ownership.
 * The ASF licenses this file to You under the Apache License, Version 2.0
 * (the "License"); you may not use this file except in compliance with
 * the License.  You may obtain a copy of the License at
 *
 *     http://www.apache.org/licenses/LICENSE-2.0
 *
 * Unless required by applicable law or agreed to in writing, software
 * distributed under the License is distributed on an "AS IS" BASIS,
 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 * See the License for the specific language governing permissions and
 * limitations under the License.
 */

/// <summary>
/// This class converts alphabetic, numeric, and symbolic Unicode characters
/// which are not in the first 127 ASCII characters (the "Basic Latin" Unicode
/// block) into their ASCII equivalents, if one exists.
/// <para/>
/// Characters from the following Unicode blocks are converted; however, only
/// those characters with reasonable ASCII alternatives are converted:
/// 
/// <ul>
///   <item><description>C1 Controls and Latin-1 Supplement: <a href="http://www.unicode.org/charts/PDF/U0080.pdf">http://www.unicode.org/charts/PDF/U0080.pdf</a></description></item>
///   <item><description>Latin Extended-A: <a href="http://www.unicode.org/charts/PDF/U0100.pdf">http://www.unicode.org/charts/PDF/U0100.pdf</a></description></item>
///   <item><description>Latin Extended-B: <a href="http://www.unicode.org/charts/PDF/U0180.pdf">http://www.unicode.org/charts/PDF/U0180.pdf</a></description></item>
///   <item><description>Latin Extended Additional: <a href="http://www.unicode.org/charts/PDF/U1E00.pdf">http://www.unicode.org/charts/PDF/U1E00.pdf</a></description></item>
///   <item><description>Latin Extended-C: <a href="http://www.unicode.org/charts/PDF/U2C60.pdf">http://www.unicode.org/charts/PDF/U2C60.pdf</a></description></item>
///   <item><description>Latin Extended-D: <a href="http://www.unicode.org/charts/PDF/UA720.pdf">http://www.unicode.org/charts/PDF/UA720.pdf</a></description></item>
///   <item><description>IPA Extensions: <a href="http://www.unicode.org/charts/PDF/U0250.pdf">http://www.unicode.org/charts/PDF/U0250.pdf</a></description></item>
///   <item><description>Phonetic Extensions: <a href="http://www.unicode.org/charts/PDF/U1D00.pdf">http://www.unicode.org/charts/PDF/U1D00.pdf</a></description></item>
///   <item><description>Phonetic Extensions Supplement: <a href="http://www.unicode.org/charts/PDF/U1D80.pdf">http://www.unicode.org/charts/PDF/U1D80.pdf</a></description></item>
///   <item><description>General Punctuation: <a href="http://www.unicode.org/charts/PDF/U2000.pdf">http://www.unicode.org/charts/PDF/U2000.pdf</a></description></item>
///   <item><description>Superscripts and Subscripts: <a href="http://www.unicode.org/charts/PDF/U2070.pdf">http://www.unicode.org/charts/PDF/U2070.pdf</a></description></item>
///   <item><description>Enclosed Alphanumerics: <a href="http://www.unicode.org/charts/PDF/U2460.pdf">http://www.unicode.org/charts/PDF/U2460.pdf</a></description></item>
///   <item><description>Dingbats: <a href="http://www.unicode.org/charts/PDF/U2700.pdf">http://www.unicode.org/charts/PDF/U2700.pdf</a></description></item>
///   <item><description>Supplemental Punctuation: <a href="http://www.unicode.org/charts/PDF/U2E00.pdf">http://www.unicode.org/charts/PDF/U2E00.pdf</a></description></item>
///   <item><description>Alphabetic Presentation Forms: <a href="http://www.unicode.org/charts/PDF/UFB00.pdf">http://www.unicode.org/charts/PDF/UFB00.pdf</a></description></item>
///   <item><description>Halfwidth and Fullwidth Forms: <a href="http://www.unicode.org/charts/PDF/UFF00.pdf">http://www.unicode.org/charts/PDF/UFF00.pdf</a></description></item>
/// </ul>
/// <para/>
/// See: <a href="http://en.wikipedia.org/wiki/Latin_characters_in_Unicode">http://en.wikipedia.org/wiki/Latin_characters_in_Unicode</a>
/// <para/>
/// For example, '&amp;agrave;' will be replaced by 'a'.
/// </summary>
public static partial class StringExtensions

    /// <summary>
    /// Converts characters above ASCII to their ASCII equivalents.  For example,
    /// accents are removed from accented characters. 
    /// </summary>
    /// <param name="input">     The string of characters to fold </param>
    /// <param name="length">    The length of the folded return string </param>
    /// <returns> length of output </returns>
    public static string FoldToASCII(this string input, int? length = null)
    
        // See https://gist.github.com/andyraddatz/e6a396fb91856174d4e3f1bf2e10951c
    

希望对其他人有所帮助,这是我找到的最强大的解决方案!

【讨论】:

注意事项:1) 概念取决于区域设置。例如,“ä”可以是“a”或“aa”。 2) 错误命名/错误描述:结果不一定仅来自 C0 Controls 和 Basic Latin 块。它仅将拉丁字母和一些符号变体转换为“等价物”。 (当然,之后可以再通过一次来替换或删除非 C0 控件和基本拉丁语块字符。)但这会做得很好。 感谢您发布此内容。我相信您在文件末尾有一个尾随 括号。【参考方案18】:

这段代码对我有用:

var updatedText = text.Normalize(NormalizationForm.FormD)
     .Where(c => CharUnicodeInfo.GetUnicodeCategory(c) != UnicodeCategory.NonSpacingMark)
     .ToArray();

但是,请不要对名字这样做。这不仅是对名字中带有元音变音/重音的人的侮辱,在某些情况下也可能是危险的错误(见下文)。除了去掉重音之外,还有其他的写作方式。

此外,这完全是错误和危险的,例如如果用户必须准确地提供他的姓名在护照上的显示方式。

例如,我的名字写成Zuberbühler,在我护照的机器可读部分,您会找到Zuberbuehler。通过删除元音变音,名称将不匹配任何部分。这可能会给用户带来问题。

您应该在姓名输入表单中禁止使用元音变音/重音,这样用户就可以正确地写出他的名字,而不会使用元音变音或重音。

实际示例,如果申请 ESTA 的 Web 服务 (https://www.application-esta.co.uk/special-characters-and) 使用上述代码而不是正确转换元音变音,则 ESTA 申请将被拒绝,或者旅行者在进入美国边境管制局时会遇到问题国家。

另一个例子是机票。假设您有一个机票预订 Web 应用程序,用户为他的名字提供了一个重音符号,而您的实现只是删除重音符号,然后使用航空公司的 Web 服务来预订机票!您的客户可能不被允许登机,因为姓名与他/她护照的任何部分都不匹配。

【讨论】:

这对韩语不起作用,需要FormC。【参考方案19】:

与接受的答案相同,但速度更快,使用 Span 而不是 StringBuilder。 需要 .NET Core 3.1 或更新的 .NET。

static string RemoveDiacritics(string text) 

    ReadOnlySpan<char> normalizedString = text.Normalize(NormalizationForm.FormD);
    int i = 0;
    Span<char> span = text.Length < 1000
        ? stackalloc char[text.Length]
        : new char[text.Length];

    foreach (char c in normalizedString)
    
        if (CharUnicodeInfo.GetUnicodeCategory(c) != UnicodeCategory.NonSpacingMark)
            span[i++] = c;
    

    return new string(span).Normalize(NormalizationForm.FormC);

这也可扩展用于其他字符替换,例如波兰语 Ł.

span[i++] = c switch

    'Ł' = 'L',
    'ł' => 'l',
    _ => c
;

一个小提示:堆栈分配stackalloc 比堆分配new 更快,并且它减少了垃圾收集器的工作。 1000 是避免在堆栈上分配大型结构的阈值,这可能导致***Exception。虽然 1000 是一个相当安全的值,但在大多数情况下 10000 甚至 100000 也可以工作(100k 在堆栈上分配高达 200kB 而默认堆栈大小为 1 MB),但是 100k 对我来说有点危险。

【讨论】:

【参考方案20】:

接受的答案是完全正确的,但现在应该更新它以使用Rune 类而不是CharUnicodeInfo,因为 C# 和 .NET 更新了最新版本中分析字符串的方式(Rune 类已添加到.NET Core 3.0)。

现在推荐使用以下 .NET 5+ 的代码,因为它更适合非拉丁字符:

static string RemoveDiacritics(string text) 

    var normalizedString = text.Normalize(NormalizationForm.FormD);
    var stringBuilder = new StringBuilder();

    foreach (var c in normalizedString.EnumerateRunes())
    
        var unicodeCategory = Rune.GetUnicodeCategory(c);
        if (unicodeCategory != UnicodeCategory.NonSpacingMark)
        
            stringBuilder.Append(c);
        
    

    return stringBuilder.ToString().Normalize(NormalizationForm.FormC);

【讨论】:

以上是关于如何从 .NET 中的字符串中删除变音符号(重音符号)?的主要内容,如果未能解决你的问题,请参考以下文章

在 JavaScript 中删除字符串中的重音符号/变音符号

从雪花中的字符串中删除变音符号[重复]

用英语等价物替换德语字符(变音符号、重音符号)

使用正则表达式剥离字符失败,使用带有变音符号,撇号,重音符号等的文字字符

将非 ASCII 字符(变音符号、重音符号...)转换为最接近的 ASCII 等效字符(创建 slug)

如何在 Sqlite 中实现重音/变音符号不敏感搜索? [复制]