4.6.7.5 Text Stream Encodings

SICStus Prolog supports character codes up to 31 bits wide where the codes are interpreted as for Unicode for the common subset.

When a character code (a “code point” in Unicode terminology) is read or written to a stream, it must be encoded into a byte sequence. The method by which each character code is encoded to or decoded from a byte sequence is called “character encoding”.

The following character encodings are currently supported by SICStus Prolog.

ANSI_X3.4-1968

The 7-bit subset of Unicode, commonly referred to as ASCII.

ISO-8859-1

The 8-bit subset of Unicode, commonly referred to as Latin 1.

ISO-8859-2

A variant of ISO-8859-1, commonly referred to as Latin 2.

ISO-8859-15

A variant of ISO-8859-1, commonly referred to as Latin 9.

windows 1252

The Microsoft Windows code page 1252.

UTF-8

UTF-16

UTF-16LE

UTF-16BE

UTF-32

UTF-32LE

UTF-32BE

The suffixes LE and BE denote respectively little endian and big endian.

These encodings can be auto-detected if a Unicode signature is present in a file opened for read. A Unicode signature is also known as a Byte order mark (BOM).

In addition, it is possible to use all alternative names defined by the IANA registry http://www.iana.org/assignments/character-sets.

All encodings in the table above, except the UTF-XXX encodings, supports the reposition(true) option to open/4 (see mpg-ref-open).

The encoding to use can be specified when using open/4 and similar predicates using the option encoding/1. When opening a file for input, the encoding can often be determined automatically. The default is ISO-8859-1 if no encoding is specified and no encoding can be detected from the file contents.

The encoding used by a text stream can be queried using stream_property/2.

See mpg-ref-open for details on how character encoding is auto-detected when opening text files.