Fixes diactritics regressions

This commit is contained in:
Jellifi007
2022-04-15 20:14:13 +02:00
parent 5778541d2f
commit 56573f14b0
16 changed files with 81 additions and 14 deletions

View File

@@ -1,4 +1,8 @@
using System;
using System.Diagnostics;
using System.Globalization;
using System.Text;
using System.Text.RegularExpressions;
namespace Jellyfin.Extensions
{
@@ -7,6 +11,44 @@ namespace Jellyfin.Extensions
/// </summary>
public static class StringExtensions
{
// Matches non-conforming unicode chars
// https://mnaoumov.wordpress.com/2014/06/14/stripping-invalid-characters-from-utf-16-strings/
private static readonly Regex _nonConformingUnicode = new Regex("([\ud800-\udbff](?![\udc00-\udfff]))|((?<![\ud800-\udbff])[\udc00-\udfff])|(\ufffd)");
/// <summary>
/// Removes the diacritics character from the strings.
/// </summary>
/// <param name="text">The string to act on.</param>
/// <returns>The string without diacritics character.</returns>
public static string RemoveDiacritics(this string text)
{
string withDiactritics = _nonConformingUnicode
.Replace(text, string.Empty)
.Normalize(NormalizationForm.FormD);
var withoutDiactritics = new StringBuilder();
foreach (char c in withDiactritics)
{
UnicodeCategory uc = CharUnicodeInfo.GetUnicodeCategory(c);
if (uc != UnicodeCategory.NonSpacingMark)
{
withoutDiactritics.Append(c);
}
}
return withoutDiactritics.ToString().Normalize(NormalizationForm.FormC);
}
/// <summary>
/// Checks wether or not the specified string has diacritics in it.
/// </summary>
/// <param name="text">The string to check.</param>
/// <returns>True if the string has diacritics, false otherwise.</returns>
public static bool HasDiacritics(this string text)
{
return !string.Equals(text, text.RemoveDiacritics(), StringComparison.Ordinal);
}
/// <summary>
/// Counts the number of occurrences of [needle] in the string.
/// </summary>