🚀 go-pugleaf

Here are the consolidated answers to the original question from Luis
Miguel (luism@sybase.com):

     What are the most commonly sort orders (collating sequences) used
     for the (economically) major non-English European languages?

     Spanish

     German

     French


Jeff
INSOFT-L List Manager

*========================================================================*
 Jeff Bowyer                         EMail:
 Computing Center		     jbowyer@cis.vutbr.cz
 Technical University of Brno
 Udolni 19, 602 00 BRNO
 Czech Republic
*========================================================================*


Sender: Patrick O'Callaghan <poc@usb.ve>

Spanish is the same as English except:

	'ch' is regarded as a single letter and sorts after 'c', for
example casa, circo, clase, cumbre, chatarra, chola.

	'll' is a distinct letter and sorts after 'l', thus lastima,
logro, llave, lluvia.

	'n~' ("enye") is a distinct letter. It's not an 'n' with an
accent. It sorts after 'n'.


Sender: James Kanze <kanze@us-es.sel.de>

|> 	Spanish

Not much experience, but I believe that the accents are ignored, ll
sorts as one letter, following l, and n~ sorts as a separate letter,
following n.  I think that ch also sorts as a single letter, following
c, but am not sure of this.

|> 	German

There are two different collating sequences in wide spread use.

Classically, you simply ignored the umlaut, sorting a" as a, etc.  The
s-tset (looks like a Greek beta) sorts as 'ss'.

DIN has adapted a different standard: a letter with umlaut is sorted
as if it were the letter without the umlaut, followed by an 'e', thus:
a" sorts as 'ae', etc.

|> 	French

Just ignore the accents (and the cedilla), and sort as in English.


Sender: Arne Larsson (LARSSON@ntcclu.ntc.nokia.com)

	This is an apparently innocent looking problem with a multitude
	of different answers.

	Problems due to ligatures like Danish ae or French oe,
	digraphs such as Spanish ch or ij in Dutch, variant letters
	like German sharp s, or thorn or edh in Icelandic, and
	diacritical marks (e grave, etc, which can not be rendered
	on the Internet), logograms like %,&, and punctuation
	marks are evidently no surprise. These are alphabetized
	differently in different languages: German a,<adiaeresis>
	o,<odiaeresis>,u,<udiaeresis>, for example:

	...
	Augiastall
	<adiaeresis>ugig	(read: [oigig])
	<Adiaerisis>uglein
	Augment
	...

	In Swedish the order is ...x y z <aring> <adiaeresis>
	<odiaeresis>, whereas Danish exhibits ... x y z <aeligature>
	<ooblique> <aring>.

	In addition to those character set oriented problems, we
	have variation due to the fact that the set of RANKING
	WORDS are selected differently according to the field of
	appication. In bibliographic filing some prepositions,
	conjunctions, definite and indefinite articles are excluded
	from the set of ranking words (e.g. and, und, et, och, og
	will not be used in alphabetizing). In phone books and similar
	directories certain prefixes (Sir, af, van, van der, von, etc),
	and attributes (General, International, Dept. of, and national
	abbreviations meaning 'limited liability company' such as Ltd,
	AB, AS, Oy, etc) are not used as ranking words.

	One solution proposed by Dr Rolf Gavare of the University
	of Gothenburg (Sweden) is to establish a literal character string
	for each ranking word. The alphabetical comparison should
	the be based on this character string (called ORDINAL FORM).
	Subsequently, SORTING KEYS will be generated to be used by
	standard sorting software for the actual sorting of entries.

	For details, please refer to Gavare, R, Alphabetic Ordering
	in a Lexicological Perspective. A computational model
	for exacting alphabetization of dictionary and encyclopedic
	material. In: Gellerstam, M (ed), Studies in Computer-Aided
	Lexicology, Almqvist & Wiksell, Stockholm 1988, p.63-102.
	Data Linguistica, Vol. 18. ISBN 91-22-01258-3.


Sender: Ciaran O'Duibhin (CSG0070@v2.qub.ac.uk)

A single-level collating sequence is not enough. To compare two strings,
you need in general to form three keys:
  primary key: letters with accents removed and converted to one case
  secondary key: formed from accents only
  tertiary key: formed from case indications only
This is because, e.g. an accent on the first letter is less significant
than an alphabetic difference in the last letter.

Comprehensive tables for many European languages are available from
Otto Stolz (RZOTTO@EARN.DKNKURZ1).


Sender: Goran Eriksson (get@get.pp.se)

If you have access to a MS-DOS machine (with MS-DOS 3.30 or later) and make a
little program (I don't know of any generally available utility for this) you
may obtain Microsoft's view on the proper collating sequence for many a
European language. You just retrieve whatever information MS-DOS function 65H
subfunction 06 returns.

As for the validity of the information ... well, I can at least tell that the
Swedish version isn't too bad.


Sender: Edgar Sanchez (esanchez@mmrree.gov.ec)

I see that many people targets Europe as their main localization market.
Here in Southamerica we also speak Spanish, I understand that is a smaller
market (specially for expensive products like Sybase) but I also think it's
got tremendous potential. By the way, Latinamerican Spanish is not exactly
the same as Spain's Spanish, of course we understand everything but some
european vocabulary sound very weird to us, even IBM makes different
keyboards for them and us (although I don't understan why), does this
qualify us as a different localization language?

Going back to the original question: I vote Spanish of course.

P.S. No pun intended on Sybase: we just bought it and plan to use it
massively here, we think is one of best RDBMS in the market (if not the
very best), still, we find it expensive.


Sender: Jiri Kuchta (kuchta@dcse.fee.vutbr.cs)