jbominji

[John Leuner <jewel@subvert-the-dominant-paradigm.net>]

I've set up a simple bot called jbominji in the #lojban IRC channel.

Right now it logs its output to a room called #jbokumfa.

It listens in #lojban and tries to parse each line of text (with the PEG
lojban grammar) and then prints output something like this:

<< mi na'o na jimpe >> 
.i zo mi cu cmavo la'e zoi gy. me gy.
.i zo na cu cmavo la'e zoi gy. bridi negator gy.
.i zo jimpe cu gismu la'e zoi gy. understand gy.

The quotes indicate which part of the text parsed successfully. Any text
following the quotes did not parse.

It then provides a gloss for gismu and cmavo from gismu.txt and
cmavo.txt.

You can access the same parser via
http://subvert-the-dominant-paradigm.net:9072/generator to generate
graphviz graphs of the parse tree generated by parsing a lojban text.

Please feel free to send me any bug-fixes, feature requests etc

You can see the PEG grammar here:
http://subvert-the-dominant-paradigm.net/~jbominji/code/lojban_grammar.peg

TODO:

jbominji doesn't lookup most cmavo yet
have to check whether it will flood itself off the channel
fuhivla
explore other modes of output (Jabber, web page with AJAX etc)