我试图转换一些字符串,在法国加拿大,基本上,我想能够拿出法国重音标记在字母,同时保持字母。(例如,将é转换为e,那么crème brûlée就会变成creme brulee)

实现这一目标的最佳方法是什么?


当前回答

我没有使用过这种方法,但是Michael Kaplan在他的博客文章(有一个令人困惑的标题)中描述了一种方法,谈论剥离变音符:剥离是一项有趣的工作(又名剥离) 论无意义的意义,即一切 Mn字符是非空格的,但是 有些更非间距比 其他人)

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:剥离变音符....

该方法使用String。Normalize将输入字符串分割为组成符号(基本上是将“基本”字符与变音符符分开),然后扫描结果并仅保留基本字符。这只是有点复杂,但实际上你看到的是一个复杂的问题。

当然,如果你限制自己使用法语,你可能会使用@David Dibben推荐的如何在c++ std::string中删除重音和波浪号的简单基于表的方法。

其他回答

为所有找到Lucene的人。Net作为一个多余的删除变音符,我设法找到这个小库,利用ASCII音译为您。

https://github.com/anyascii/anyascii

没有足够的声誉,显然我不能评论亚历山大的优秀链接。Lucene似乎是唯一的解决方案在合理的通用情况下工作。

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

字符串试验台= " AAAACEIIOOØUUÞaaaaaaæceeeeiiiið人参公鸡øUUāăčĐęğıŁłńŌōřŞşšźžșțệủ”;

Console.WriteLine (Lucene.latinizeLucene(实验);

AAAACEIIOOOUUTHaaaaaaaeceeeeiiiidnoooouuaacDegiLlnOorSsszzsteu

//////////

public static class Lucene
{
    // source: https://raw.githubusercontent.com/apache/lucenenet/master/src/Lucene.Net.Analysis.Common/Analysis/Miscellaneous/ASCIIFoldingFilter.cs
    // idea: https://stackoverflow.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;
    }
}

如果有人感兴趣,我正在寻找类似的东西,最后写了如下:

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
}

与接受的答案相同,但更快,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);
}

此外,这是可扩展的额外字符替换,如抛光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要快得多,它为垃圾收集器减少了工作。1000是一个阈值,以避免在堆栈上分配大结构,这可能会导致StackOverflowException。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(100k在堆栈上分配最多200kB,而默认堆栈大小为1mb)。然而10万对我来说有点危险。

我没有使用过这种方法,但是Michael Kaplan在他的博客文章(有一个令人困惑的标题)中描述了一种方法,谈论剥离变音符:剥离是一项有趣的工作(又名剥离) 论无意义的意义,即一切 Mn字符是非空格的,但是 有些更非间距比 其他人)

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:剥离变音符....

该方法使用String。Normalize将输入字符串分割为组成符号(基本上是将“基本”字符与变音符符分开),然后扫描结果并仅保留基本字符。这只是有点复杂,但实际上你看到的是一个复杂的问题。

当然,如果你限制自己使用法语,你可能会使用@David Dibben推荐的如何在c++ std::string中删除重音和波浪号的简单基于表的方法。