Re: Access/Excel, etc. (was: Has anyone made a real conlang? )
From: | Elyse Grasso <emgrasso@...> |
Date: | Friday, April 25, 2003, 1:49 |
On Thursday 24 April 2003 01:39 pm, BP Jonsson wrote:
> At 13:52 23.4.2003 -0600, Elyse Grasso wrote:
>
> >it still might need some perl for the conlang sorting order.
>
> How do you do the sorting by conlang sorting order?
> Run everything thru tr///, then thru sort() and then again thru tr///,
> or is there a more elegant way?
>
> / B.Philip Jonsson B^)
This is going to take some explaining....
when I first enter a term into the lexicon, the record looks something
like this:
<entry>
<term>adth</term><class>root: adjectival</class><xsampa></xsampa>
<gloss>to be sick, ill</gloss>
<keyword>sick</keyword><keyword>ill</keyword>
</entry>
after the first time I run the html generation script against it, it
looks like this:
<entry><term>adth</term>
<class>root: adjectival</class>
<xsampa>/ad_T/</xsampa>
<gloss>to be sick, ill</gloss>
<keyword>sick</keyword>
<keyword>ill</keyword>
<glyphs>48 2A</glyphs>
</entry>
The translation tables between the Latin orthography used in the 'term'
field and the values used in the 'glyphs' and 'xsampa' fields are
recorded on the orthography section of the web page.
The glyphs translation is a precursor to the xsampa generation, and is
done with a carefully tuned sequence of s/// commands. ('dth' needs to
be glyphed before 'th', and 'th' before 't' or 'h', etc.)
The native sort is basically a Jouevyaix ascii sort, performed on the
glyphs field. The glyph data will let me generate Jouevyaix text fields
in the lexicon displays if/when I get my script and font fully defined.
I currently do manual tweaks for the glyphs for medial vowels in closed
syllables, which take a different form from open ones. I'd need to add
syllabification info to the term field to get the native spelling
really right with automation, or else make the script interactive.
Haven't decided which way to play it yet. It doesn't matter a lot for
sorting purposes yet: medial vowels are adjacent to the non-medial
forms, and can never happen initially. (Imperial glyph translation is
completely automatable, since initial vowel are easy to identify.)
There is an optional <usage> field in my lexicon.dtd, in case I get
ambitious enough to add usage examples for some terms.
--
Elyse Grasso
The World of Cherani Station
www.data-raptors.com/cherani/index.html
Cherani Tradespeech
www.data-raptors.com/cherani/tradespeech.html