{ju'anai} is not a lujvo.

[lagleki]

http://vlasisku.lojban.org/vlasisku/ju'anai says Components in this compound:. Clearly, the components should not be {jufra} and {nabmi}.

[mantycore]

I guess the part of issue is what vlasisku do not check for double consonats cluster in the first five letters of lujvo, maybe? http://vlasisku.lojban.org/vlasisku/be%27ebe%27e

[durka]

Correct, the lujvo decomposer function that vlasisku uses creates many
false positives because it doesn't check for required hyphens. The version
in my fork already uses the parser to break apart lujvo so this issue is
solved (though my version doesn't break up cmavo compounds, it just says
not found, maybe I should add that).
On Feb 15, 2015 12:28 AM, "mantycore" notifications@github.com wrote:

I guess the part of issue is what vlasisku do not check for double
consonats cluster in the first five letters of lujvo, maybe?
http://vlasisku.lojban.org/vlasisku/be%27ebe%27e

—
Reply to this email directly or view it on GitHub
#2 (comment).

[lagleki]

should be just using ilmentufa to split words.

2015-02-15 8:56 GMT+03:00 Alex Burka notifications@github.com:

Correct, the lujvo decomposer function that vlasisku uses creates many
false positives because it doesn't check for required hyphens. The version
in my fork already uses the parser to break apart lujvo so this issue is
solved (though my version doesn't break up cmavo compounds, it just says
not found, maybe I should add that).

On Feb 15, 2015 12:28 AM, "mantycore" notifications@github.com wrote:

I guess the part of issue is what vlasisku do not check for double
consonats cluster in the first five letters of lujvo, maybe?
http://vlasisku.lojban.org/vlasisku/be%27ebe%27e

—
Reply to this email directly or view it on GitHub
#2 (comment).

—
Reply to this email directly or view it on GitHub
#2 (comment).

[lagleki]

or la cmaxes which is a purely morphological parser http://mw.lojban.org/extensions/ilmentufa/cmaxes.html

[Pendrokar]

Seems fixed.