[tex-hyphen] New language classiclatin – question about the language code

[Mojca Miklavec]

On Wed, Jun 4, 2014 at 12:11 AM, Claudio Beccari wrote:
>
> Ih had to do so because the hyphenation rules used by modern scholars are
> different; the spelling of the three variants is different, but one pattern
> set can accomodate both modern and medieval Latin, while such procedure was
> impossible with classic Latin; I had to work quite a lot to produce the
> classic Latin patterns, because the rules aro so incompatible that a new
> different pattern set is required. Just to make an example the word
> transubstantialis is hyphenated as tran-sub-stan-tia-lis with
> modern/medieval latin patterns and as tran-subs-tan-ti-a-lis with classic
> latin patterns.

Thank you. This was exactly the piece of information that I was
interested in. If that's the case I'm at least sure that we cannot
have one set of patterns no matter what.

(I really wonder how people know how Cicero hyphenated words ;)

> I considered that the tags la and  lac were sufficient to distinguish
> la(modern&medieval) from la(classic). I was not aware of the existance of
> special tags and subtags for the various languages as Mojca showed me
> through the link:
>
> http://www.iana.org/assignments/language-subtag-registry/language-subtag-registry
>
> The contents of that language-subtag-registry file is a little misterious
> for me; neverthelss  I think I made a reasonable suggestion by proposing the
> addition of
>
>
> %%
> Type: variant
> Subtag: classic
> Description: Classic Latin
> Added: 2014-06-03
> Prefix: la

Yes, that's reasonable, but it needs to be officially registered or we
may use a private tag instead.

> but the overall correspondece on this topic shows that a subtag classic may
> not be acceptable with the prefix la.

> OK. And what about tags ans subtags for Greek? they contain some identical
> tags with different subtags (and I did not list all of them). Why la-classic
> is not acceptable, while, say, el and el-polyton are acceptable?

No, I didn't say it wasn't acceptable. I just mentioned that once
la-classic is registered, el-classic may not be registered later for
example. Or, if one registers "la-new", "<anyotherlanguage>-new" is
forbidden from that point on. (You may not register "la-polyton"
however.)

And again: official registration of a new tag is not a prerequisite to
use these patterns.

I added the patterns, using a private tag "x-classic" for now.

> Mojca shows that Wikipedia lists more than half a dozen variants of Latin; I
> am not going to add so many modifiers/attributes for all these Latin
> languages; I would like that real linguists took care of their necessities
> with their professional competence, that I miss. But up to now I wrote some
> 15 pattern files for different languages (not all of them are or have been
> on CTAN); I got some experience in creating such pattern files without using
> patgen (because it assumes the existance of a very large list of correctly
> hyphenated words) but just using grammar rules. Now some difficulties arise
> because the language-subtag-registry file is incomplete.

It's not really a problem. I just wanted to try to name the patterns
"by the book" whenever possible to avoid situations like those in some
macro package when users would ask for language "uk" and then wonder
why they got Ukrainian settings as opposed to the British English. Or
vice versa.

> Who is in charge of maintaining that file should provide the missing
> entries;

See http://www.langtag.net/register-new-subtag.html

> I am not going to submit any request for updating that list.

OK.

> Is it
> really necessary for the good working of the TeX system

No, it is not.

> or is it just one of
> those constraints that are becomming so common in modern times.

It's not a constraint. Just a discussion of whether we want to
register a new tag or use a private one (and if so, which one).

> I added the
> classiclatin language to my personal language files; recreated the formats
> and everything seems to work properly so that I succeeded testing and
> correcting as much as necessary.

Formats build fine for me as well.

> If and when a decision on tags and subtags is taken, I accordingly change
> the contents of the babel-latin and gloss-latin language description files,
> and of course the hyph-*.tex files.
>
> Please let me know the decisions taken on this matter.

I now used "la-x-classic", but we can change that if needed. The
letter "x" means that we are using an unregistered/private tag.

We didn't make any decision yet about whether or not a new subtag
should be officially requested/registered or not. But nobody seemed
too eager to try to do that.

However, babel and polyglossia probably shouldn't depend on that tag.
What matters to babel/polyglossia is just the "tex name" which I have
set to "classiclatin" as you suggested.

Mojca