| 123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295296297298299300301302303304305306307308309310311312313314315316317318319320321322323324325326327328329330331332333334335336337338339340341342343344345346347348349350351352353354355356357358359360361362363364365366367368 |
- using System.Diagnostics.CodeAnalysis;
- using System.Globalization;
- using System.Runtime.CompilerServices;
- using System.Text;
- using System.Text.RegularExpressions;
- namespace Lua.Internal;
- internal static class StringHelper
- {
- [MethodImpl(MethodImplOptions.AggressiveInlining)]
- public static ReadOnlySpan<char> Slice(string s, int i, int j)
- {
- if (i < 0) i = s.Length + i + 1;
- if (j < 0) j = s.Length + j + 1;
- if (i < 1) i = 1;
- if (j > s.Length) j = s.Length;
- return i > j ? "" : s.AsSpan()[(i - 1)..j];
- }
- public static bool TryFromStringLiteral(ReadOnlySpan<char> literal, [NotNullWhen(true)] out string? result)
- {
- var builder = new ValueStringBuilder(literal.Length);
- for (int i = 0; i < literal.Length; i++)
- {
- var c = literal[i];
- if (c is '\\' && i < literal.Length - 1)
- {
- i++;
- c = literal[i];
- switch (c)
- {
- case '\n':
- builder.Append('\n');
- break;
- case '\r':
- builder.Append('\n');
- // check CRLF
- if (i + 1 < literal.Length && literal[i + 1] is '\n')
- {
- i++;
- }
- break;
- case 'a':
- builder.Append('\a');
- break;
- case 'b':
- builder.Append('\b');
- break;
- case 'f':
- builder.Append('\f');
- break;
- case 'n':
- builder.Append('\n');
- break;
- case 'r':
- builder.Append('\r');
- break;
- case 't':
- builder.Append('\t');
- break;
- case 'v':
- builder.Append('\v');
- break;
- case '\\':
- builder.Append('\\');
- break;
- case '\"':
- builder.Append('\"');
- break;
- case '\'':
- builder.Append('\'');
- break;
- case '[':
- builder.Append('[');
- break;
- case ']':
- builder.Append(']');
- break;
- case 'x':
- i++;
- if (i >= literal.Length)
- {
- result = null;
- return false;
- }
- c = literal[i];
- if (IsDigit(c))
- {
- var start = i;
- for (int j = 0; j < 2; j++)
- {
- i++;
- if (i >= literal.Length) break;
- c = literal[i];
- if (!IsDigit(c)) break;
- }
- builder.Append((char)int.Parse(literal[start..i], NumberStyles.HexNumber));
- i--;
- }
- else
- {
- result = null;
- return false;
- }
- break;
- default:
- if (IsNumber(c))
- {
- var start = i;
- for (int j = 0; j < 3; j++)
- {
- i++;
- if (i >= literal.Length) break;
- c = literal[i];
- if (!IsNumber(c)) break;
- }
- builder.Append((char)int.Parse(literal[start..i]));
- i--;
- }
- else
- {
- result = null;
- return false;
- }
- break;
- }
- }
- else
- {
- builder.Append(c);
- }
- }
- result = builder.ToString();
- return true;
- }
- public static string Escape(ReadOnlySpan<char> str)
- {
- var builder = new ValueStringBuilder(str.Length);
- for (int i = 0; i < str.Length; i++)
- {
- var c = str[i];
- switch (c)
- {
- case '\a':
- builder.Append("\\\a");
- break;
- case '\b':
- builder.Append("\\\b");
- break;
- case '\f':
- builder.Append("\\\f");
- break;
- case '\n':
- builder.Append("\\\n");
- break;
- case '\r':
- builder.Append("\\\r");
- break;
- case '\t':
- builder.Append("\\\t");
- break;
- case '\v':
- builder.Append("\\\v");
- break;
- case '\\':
- builder.Append("\\\\");
- break;
- case '\"':
- builder.Append("\\\"");
- break;
- case '\'':
- builder.Append("\\\'");
- break;
- default:
- builder.Append(c);
- break;
- }
- }
- return builder.ToString();
- }
- public static Regex ToRegex(ReadOnlySpan<char> pattern)
- {
- var builder = new ValueStringBuilder();
- var isEscapeSequence = false;
- var isInSet = false;
- for (var i = 0; i < pattern.Length; i++)
- {
- var c = pattern[i];
- if (isEscapeSequence)
- {
- if (c == '%' || c == '_')
- {
- builder.Append(c);
- isEscapeSequence = false;
- }
- else
- {
- switch (c)
- {
- case 'a': // all letters
- builder.Append("\\p{L}");
- break;
- case 'A': // all Non letters
- builder.Append("\\P{L}");
- break;
- case 's': // all space characters
- builder.Append("\\s");
- break;
- case 'S': // all NON space characters
- builder.Append("\\S");
- break;
- case 'd': // all digits
- builder.Append("\\d");
- break;
- case 'D': // all NON digits
- builder.Append("\\D");
- break;
- case 'w': // all alphanumeric characters
- builder.Append("\\w");
- break;
- case 'W': // all NON alphanumeric characters
- builder.Append("\\W");
- break;
- case 'c': // all control characters
- builder.Append("\\p{C}");
- break;
- case 'C': // all NON control characters
- builder.Append("[\\P{C}]");
- break;
- case 'g': // all printable characters except space
- builder.Append("[^\\p{C}\\s]");
- break;
- case 'G': // all NON printable characters including space
- builder.Append("[\\p{C}\\s]");
- break;
- case 'p': // all punctuation characters
- builder.Append("\\p{P}");
- break;
- case 'P': // all NON punctuation characters
- builder.Append("\\P{P}");
- break;
- case 'l': // all lowercase letters
- builder.Append("\\p{Ll}");
- break;
- case 'L': // all NON lowercase letters
- builder.Append("\\P{Ll}");
- break;
- case 'u': // all uppercase letters
- builder.Append("\\p{Lu}");
- break;
- case 'U': // all NON uppercase letters
- builder.Append("\\P{Lu}");
- break;
- case 'x': // all hexadecimal digits
- builder.Append("[0-9A-Fa-f]");
- break;
- case 'X': // all NON hexadecimal digits
- builder.Append("[^0-9A-Fa-f]");
- break;
- case 'b':
- if (i < pattern.Length - 2)
- {
- var c1 = pattern[i + 1];
- var c2 = pattern[i + 2];
- var c1Escape = Regex.Escape(c1.ToString());
- var c2Escape = Regex.Escape(c2.ToString());
- builder.Append("(");
- builder.Append(c1Escape);
- builder.Append("(?>(?<n>");
- builder.Append(c1Escape);
- builder.Append(")|(?<-n>");
- builder.Append(c2Escape);
- builder.Append(")|(?:[^");
- builder.Append(c1Escape);
- builder.Append(c2Escape);
- builder.Append("]*))*");
- builder.Append(c2Escape);
- builder.Append("(?(n)(?!)))");
- i += 2;
- }
- else
- {
- throw new Exception(); // TODO: add message
- }
- break;
- default:
- builder.Append('\\');
- builder.Append(c);
- break;
- }
- isEscapeSequence = false;
- }
- }
- else if (c == '%')
- {
- isEscapeSequence = true;
- }
- else if (c == '\\')
- {
- builder.Append("\\\\");
- }
- else if (isInSet)
- {
- if (c == ']') isInSet = false;
- builder.Append(c);
- }
- else if (c == '-')
- {
- builder.Append("*?");
- }
- else if (c == '[')
- {
- builder.Append('[');
- isInSet = true;
- }
- else if (c == '^' && !isInSet)
- {
- builder.Append("\\G");
- }
- else if (c == '(')
- {
- builder.Append('(');
- }
- else
- {
- builder.Append(c);
- }
- }
- return new Regex(builder.ToString());
- }
- [MethodImpl(MethodImplOptions.AggressiveInlining)]
- public static bool IsNumber(char c)
- {
- return '0' <= c && c <= '9';
- }
- [MethodImpl(MethodImplOptions.AggressiveInlining)]
- public static bool IsDigit(char c)
- {
- return IsNumber(c) ||
- ('a' <= c && c <= 'f') ||
- ('A' <= c && c <= 'F');
- }
- }
|