Soenneker.Extensions.Char 4.0.102

Soenneker.Extensions.Char

Allocation-free character classification and casing helpers with explicit ASCII-only operations and Unicode-aware fast paths.

Installation

dotnet add package Soenneker.Extensions.Char

ASCII-only operations

using Soenneker.Extensions.Char;

bool ascii = 'A'.IsAscii();                    // true
bool letter = 'z'.IsAsciiLetter();             // true
bool digit = '٥'.IsAsciiDigit();                // false
bool identifierPart = '7'.IsAsciiLetterOrDigit(); // true

char upper = 'q'.ToAsciiUpper();               // 'Q'
char unchanged = 'é'.ToAsciiUpper();           // 'é'

The ASCII-only APIs never apply Unicode or culture rules:

• IsAscii, IsAsciiUpper, IsAsciiLower, IsAsciiLetter, IsAsciiDigit, IsAsciiLetterOrDigit
• IsAsciiWhiteSpace for space and U+0009 through U+000D
• ToAsciiUpper and ToAsciiLower, which leave non-ASCII characters unchanged
• EqualsAsciiIgnoreCase(c, lower), where lower must be a lowercase ASCII letter ('a' through 'z')

EqualsAsciiIgnoreCase is intentionally specialized; it does not validate or normalize its second argument and is not a general Unicode case-insensitive comparison.

Unicode-aware operations

bool unicodeDigit = '٥'.IsDigitFast();       // true
bool unicodeLetter = '日'.IsLetterFast();    // true
bool unicodeSpace = '\u2003'.IsWhiteSpaceFast(); // true

char invariantUpper = 'é'.ToUpperInvariant(); // 'É'

IsLetterOrDigitFast, IsDigitFast, IsLetterFast, IsUpperFast, IsLowerFast, and IsWhiteSpaceFast use an ASCII branch and fall back to the matching char BCL method for non-ASCII input. Their classification semantics match those char methods; “Fast” describes the implementation path, not a narrower character set.

ToUpperInvariant and ToLowerInvariant use an ASCII branch and fall back to char.ToUpperInvariant / char.ToLowerInvariant. Like all char casing APIs, they return one UTF-16 code unit and cannot represent mappings that require multiple characters.

Separators and newlines

IsTokenSeparator() recognizes exactly:

• ASCII whitespace: tab, line feed, vertical tab, form feed, carriage return, and space
• _, -, ., /, \, :, and ;

All non-ASCII characters return false from IsTokenSeparator().

IsAsciiNewLine() recognizes carriage return and line feed. IsNewLineFast() additionally recognizes next line (U+0085), line separator (U+2028), and paragraph separator (U+2029). Neither method treats a CRLF pair as one unit; call it for each character while parsing.

UTF-16 boundary

A .NET char is a UTF-16 code unit, not necessarily a complete Unicode scalar value. Characters outside the Basic Multilingual Plane are represented by surrogate pairs, and these extensions see each surrogate separately. Use System.Text.Rune when classification must operate on full Unicode scalar values.