ALA character set

[Peter C. Gorman]

Urban Andersson writes:

>How is the ALA character set related/similair to ISO 8859-1/Latin 1,
>that is, does it cover all Western European/Scandinavian characters,
>such as =C5 (A with ring over), =C4 (A with diaeresis/umlaut) and other
>characters with different diacritics used in those languages?
>I understand it uses 8-bit characters.

Like ISO8859-1, the ALA character set (formally, ANSI/NISO Z39.47-1993
Extended Latin Alphabet Coded Character Set for Bibliographic Use (ANSEL))
uses 8-bit characters. The ANSEL character set can represent more
characters within that space because it encodes diacritics separately from
the characters they modify. The cost of this is that a character/diacritic
combination takes 2 bytes (diacritic first, then the modified character).
The benefit is that it allows arbitrary diacritic/character pairings,
including multiple diacritics on a single character. ISO Latin 1, by
contrast, has a separate (1 byte) encoding for each character-diacritic
combination.

ANSEL also contains single-byte non-English characters (Icelandic thorn,
inverted question mark, etc.), some of which are also found in ISO Latin 1,
but not necessarily with the same numeric value. I don't know offhand
whether ANSEL represents a strict superset of the characters in Latin 1.

>Is there a specification/list of the ALA character set anywhere on the
>web?

I couldn't find a table of the characters anywhere on the web, but you can
find bibliographic information on the published standard at
http://www.faxon.com/Standards/NISOcat/NISO_Automation.html

It is also contained in _USMARC Specifications for Record Structure,
Character Sets, and Exchange Media_, available from the Library of Congress
Cataloging Distribution Service.

Hope this helps,


PG
_______________________________
Peter C. Gorman
University of Wisconsin
General Library System
Automation Services
pcgorman at facstaff.wisc.edu
(608) 265-5291