textcode

Text encoding and decoding between legacy charsets and UTF-8, plus DVB text (ETSI EN 300 468 Annex A) with charset selectors.

Conversion is non-strict: bytes without a mapping decode to U+FFFD, characters without a mapping encode to ?.

Charsets

• Utf8
• Utf16 - decodes BE and LE with a BOM, encodes BE without a BOM
• Iso6937 - Latin superset of ISO/IEC 6937 with the Euro sign
• Iso8859_1 - Western European
• Iso8859_2 - Central European
• Iso8859_3 - South European
• Iso8859_4 - North European
• Iso8859_5 - Cyrillic
• Iso8859_6 - Arabic
• Iso8859_7 - Greek
• Iso8859_8 - Hebrew
• Iso8859_9 - Turkish
• Iso8859_10 - Nordic
• Iso8859_11 - Thai
• Iso8859_13 - Baltic Rim
• Iso8859_14 - Celtic
• Iso8859_15 - Western European
• Iso8859_16 - South-Eastern European
• Gb2312 - Simplified Chinese
• Geo - DVB single-byte Georgian encoding (not part of EN 300 468)

Usage

Every charset is a type implementing the Textcode trait; the generic functions pick the codec by type parameter:

use textcode::{Iso8859_5, decode, encode};

let text = decode::<Iso8859_5>(b"\xbf\xe0\xd8\xd2\xd5\xe2!");
assert_eq!(text, "Привет!");

let bytes = encode::<Iso8859_5>("Привет!");
assert_eq!(bytes, b"\xbf\xe0\xd8\xd2\xd5\xe2!");

decode_to_slice and encode_to_slice write into a caller-provided buffer and return the number of bytes written:

use textcode::{Utf8, decode_to_slice};

let mut buf = [0u8; 16];
let len = decode_to_slice::<Utf8>(b"abc", &mut buf);
assert_eq!(&buf[.. len], b"abc");

DVB text

The dvb module handles EN 300 468 Annex A text, where the first bytes select the charset:

Selector	Charset
0x20-0xFF (none) or 0x00	Iso6937
0x01-0x0B	Iso8859_5-Iso8859_15
0x10 0x00 NN	Iso8859_NN (NN = 1-16)
0x11	Utf16 (big endian)
0x13	Gb2312
0x15	Utf8
0x1E	Geo

use textcode::dvb::{self, Charset};

// decoding follows the selector
assert_eq!(dvb::decode(b"\x01\xbf\xe0\xd8\xd2\xd5\xe2!"), "Привет!");

// encoding emits the selector for the chosen charset
assert_eq!(dvb::encode("Привет!", Charset::Iso8859_5), b"\x01\xbf\xe0\xd8\xd2\xd5\xe2!");

DvbText owns raw DVB bytes as they come off the wire and is never validated on construction; DvbTextRef is a borrowed view with a verified selector. Both decode through Display:

use textcode::dvb::{Charset, DvbText, DvbTextRef};

let text = DvbText::from(&b"\x15Menu"[..]);
assert_eq!(text.to_string(), "Menu");

let text_ref = DvbTextRef::try_from(text.as_bytes()).unwrap();
assert_eq!(text_ref.charset(), Charset::Utf8);
assert_eq!(text_ref.text(), b"Menu");

// an explicit charset overrides the selector
assert_eq!(text_ref.decode_with(Charset::Iso8859_1), "Menu");

// an undecodable selector displays as empty, the bytes stay reachable
let bad = DvbText::from(&b"\x1f?"[..]);
assert_eq!(bad.to_string(), "");
assert_eq!(bad.as_bytes(), b"\x1f?");

License

MIT