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 < 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 < 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«
»
和 …
(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 brûlée
到 crme brle
与 creme 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, '&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
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