[sword-devel] Transliteration
Chris Little
sword-devel@crosswire.org
Tue, 07 Oct 2003 22:34:42 -0700
Martin Gruner wrote:
>>So, my thought is to separate the single transliterator option we
>>currently have into script-specific options, e.g. one for Greek, one for
>>Hebrew, one for Gothic, etc.
>>
>>Does that sound good or bad or does anyone have a better idea?
>
>
> I'm not sure I understand your proposal correctly. One for every source or
> target script?
One for every source script. For example, there will be options for
Greek texts that allow transliteration to Beta, BGreek, ALA-LC, IPA,
ISO, etc.
> One for every source script would enable us to only offer the transliterators
> which support a module's text, but would result in a multitude of
> transliteration filters.
There are around 20 different scripts that we support currently. (It
will actually get more complicated when I start writing
language-specific transliterators, but that won't be until some time
next year.) I think doing a transliterator per script will be
necessary, but when it is implemented, we can make the API check for
modules with a given script and only display those that are relevent to
the user. (This will be a little complicated too, since we only know
the language that appears in a module, not the script it uses, and we
only know the primary language, not languages that may be used in short
quotations.)
> One for every target scripts would avoid that, as there would only be few
> filters. Maybe the filters could have a function to ask them which source
> scripts they support?
One filter per target script would force all source scripts to
transliterate according to a single standard. I.e. Greek couldn't be
displayed as ALA-LC while Hebrew is displayed as SBL. It also breaks
down when transliteration to a particular standard does not exist for a
given script. E.g. the Hugoye only works for Syriac, so if it's
selected as the target standard, what should be done with Hebrew? This
is essentially how the transliterator filter works currently--and since
it applies to all scripts, we only need one filter.
--Chris