🚀 go-pugleaf

andrew@cursci.co.uk (Andrew Trotman) writes:


>Anyone out there got a description of any character sets and collating
>sequences that include all the roman characters and accented roman
>characters?

I am not sure if anybody really knows how many roman letters and
accented letters there exist, but a fair part have been tabled and
sorted in some POSIX compliant locales available at dkuug.dk in
the directory i18n. The locales also include specification of
greek and Cyrillic and Kana sorting.

As noted in another article the sorting sequences are different for
at least some languages. The locales available at Dkuug.dk covers 

                        da_DK                 Danish
                        en_DK                 English

The English one is actually intended to be quite general, also
covering languages as German, French, Italian, Portuguese, Dutch,
Vietnamese, Greek, Russian and Japanese

>Someone was telling me about ISO stuff and how there were a whole load of
>8-bit sets that define most of the european characters.  That would be kind
>of useful, but what I really want is a full description of either a 16-bit or
>32-bit character sequences that will include all the roman chars and their
>accented chars in a sequence that I can sort on.

There are also provided POSIX charmaps for ISO 10646 - only a part tho,
something like 2000 characters. And about 100 other character sets
like the ISO 8859 series, PC codepages, EBCDICs are also described
via POSIX charmaps.

Keld Simonsen