Character strings

TrustLang supports two string types:

• :StrChar — universal byte UTF-8 strings;
• :StrWide — system (wide-character) strings.

Byte strings :StrChar

A single element of a byte string is one byte (more precisely :Int8, an eight-bit signed integer). The size of the string is returned in bytes, and index access refers to the byte of the sequence. Since the data is interpreted as a sequence of UTF-8 characters, when modifying individual bytes care must be taken.

System strings :StrWide

A single element of a system string is the wide character wchar_t, and the size is returned in wide characters. The size of wchar_t depends on the operating system: in Windows one :StrWide character occupies 2 bytes, in Linux — 4. The main purpose of system strings is to simplify work in a text terminal: one character always corresponds to one character cell without converting UTF-8 code points.

Despite the different single element, the internal representation of both types is the same: the data of the strings is stored as a byte UTF-8 sequence.

String formatting

Any variable can be addressed as a function (by specifying parentheses after the name), thereby creating a copy/clone of the object; for strings, such a call is used as a data formatting template.

Formatting is possible in two ways:

• A printf-style format string (the type :FmtChar/:FmtWide) is not implemented in the current version — the type :FmtChar/:FmtWide does not exist (see Status). For native functions with a printf string, the attribute @[format("printf", pos, ...)] is used (parameter checking at compile time, example — hello).
• In all other cases (any string by default), the format corresponds to the {fmt}/std::format library: the ordinal number of the argument ({0}, {1}) or empty {}. Argument checking is performed when the string is cloned (at compile time and at runtime). Named arguments (name = …) are not implemented.

$template := "{0}: {1}";    # format as in {fmt}
$result := $template('value', 123);