12.1.1 Encodings and Locales

When a port is provided to a character-based operation, such as read-char or read, the port’s bytes are read and interpreted as a UTF-8 encoding of characters. Thus, reading a single character may require reading multiple bytes, and a procedure like char-ready? may need to peek several bytes into the stream to determine whether a character is available. In the case of a byte stream that does not correspond to a valid UTF-8 encoding, functions such as read-char may need to peek one byte ahead in the stream to discover that the stream is not a valid encoding.

When an input port produces a sequence of bytes that is not a valid UTF-8 encoding in a character-reading context, then bytes that constitute an invalid sequence are converted to the character #\uFFFD. Specifically, bytes 255 and 254 are always converted to #\uFFFD, bytes in the range 192 to 253 produce #\uFFFD when they are not followed by bytes that form a valid UTF-8 encoding, and bytes in the range 128 to 191 are converted to #\uFFFD when they are not part of a valid encoding that was started by a preceding byte in the range 192 to 253. To put it another way, when reading a sequence of bytes as characters, a minimal set of bytes are changed to the encoding of #\uFFFD so that the entire sequence of bytes is a valid UTF-8 encoding.

See Byte Strings for procedures that facilitate conversions using UTF-8 or other encodings. See also reencode-input-port and reencode-output-port for obtaining a UTF-8-based port from one that uses a different encoding of characters.

A locale captures information about a user’s language-specific interpretation of character sequences. In particular, a locale determines how strings are “alphabetized,” how a lowercase character is converted to an uppercase character, and how strings are compared without regard to case. String operations such as string-ci=? are not sensitive to the current locale, but operations such as string-locale-ci=? (see Strings) produce results consistent with the current locale.

A locale also designates a particular encoding of code-point sequences into byte sequences. Racket generally ignores this aspect of the locale, with a few notable exceptions: command-line arguments passed to Racket as byte strings are converted to character strings using the locale’s encoding; command-line strings passed as byte strings to other processes (through subprocess) are converted to byte strings using the locale’s encoding; environment variables are converted to and from strings using the locale’s encoding; filesystem paths are converted to and from strings (for display purposes) using the locale’s encoding; and, finally, Racket provides functions such as string->bytes/locale to specifically invoke a locale-specific encoding.

A Unix user selects a locale by setting environment variables, such as LC_ALL. On Windows and Mac OS X, the operating system provides other mechanisms for setting the locale. Within Racket, the current locale can be changed by setting the current-locale parameter. The locale name within Racket is a string, and the available locale names depend on the platform and its configuration, but the "" locale means the current user’s default locale; on Windows and Mac OS X, the encoding for "" is always UTF-8, and locale-sensitive operations use the operating system’s native interface. (In particular, setting the LC_ALL and LC_CTYPE environment variables does not affect the locale "" on Mac OS X. Use getenv and current-locale to explicitly install the environment-specified locale, if desired.) Setting the current locale to #f makes locale-sensitive operations locale-insensitive, which means using the Unicode mapping for case operations and using UTF-8 for encoding.

When locale sensitivity is disabled by setting the parameter to #f, strings are compared, etc., in a fully portable manner, which is the same as the standard procedures. Otherwise, strings are interpreted according to a locale setting (in the sense of the C library’s setlocale). The "" locale is always an alias for the current machine’s default locale, and it is the default. The "C" locale is also always available; setting the locale to "C" is the same as disabling locale sensitivity with #f only when string operations are restricted to the first 128 characters. Other locale names are platform-specific.

String or character printing with write is not affected by the parameter, and neither are symbol case or regular expressions (see Regular Expressions).