Hi!
I'm aware that this topic has been dicussed before, but I couldn't
find exactly what I'm looking for in the archives.
There are several tools for maintaining a lexicon and generating
inflexions according to rules, and then there's software like
Langmaker, which besides being non-free and unmaintained, is quite
similar to what I want, but I don't think it's quite that. I'm judging
from what I've heard and some web apps that supposedly do something
similar. They just give you a random list of words, but not the full
list of possible words reponding to the regular expression or BNF
introduced.
So, what I need is a program/script which can:
1) Generate the FULL list of possible words according to a user-
defined specification, like a EBNF or regular expressions. For
instance, "Word := C V C | V C" with additional restrictions, like
"the first letter can't be "h" "
2) Let me define another list of "dummy names" for the words in the
lexicon, so that I can build example sentences like "#article-1 #noun-
dog #verb-barks" without using actual words of the conlang, which may
not even be assigned yet.
3) Let me build a third list of translations to other language or
languages.
4) Let me assign conlang words to their dummy names, and dummy names
to their translations
5) Tell me at every time which words are still available (unassigned
to dummy names) and vice versa, which dummy names have yet no
translation, and vice versa, which translations have no conlang word
and viceversa.
6)Output the results in a suitable table format (word- dummy name -
translation) to be inserted in text or HTML documents.
Most of this can be done with a database, like kexi (similar to MS
Access) or OOBase. There are even more flexible options, like
powerloom. The problem I have is how to interact with the database for
automated lexicon generation (point 1). So, maybe if a script can
generate a CSV text file of words from a BNF description, that could
be imported to the database.
All ideas are welcome :)
Regards,
|
