On Tue, Aug 7, 2012 at 2:14 AM, Jimmy O'Regan <[email protected]> wrote:
> On 6 August 2012 21:03, Jacob Nordfalk <[email protected]> wrote:
> > A workaround would be to split the code up in several methods.
> > However, I think we first should look if the rule could be simplified or
> > something. as 64 thousand bytecode instructuions probably means the rule
> > contains over 5000-10000 comparisons and invocations, which I'd is way
> too
> > much
>
> I've replaced ~10,000 lines of rules with one 20 line macro - that's
> only maybe 20% of what can be deleted from those two rules, but they
> seem to build without error now.
>
Thank you Jimmy!
The bytecode is generated without any error, and everything works fine when
run. But there is still a problem... I haven't been able to convert it to
Dalvik bytecode, dx simply hangs when performing the conversion and I have
to kill its process. I guess that Dalvik has a similar limitation for
method size, and that hasn't been solved yet...
Apart from that, I have added the modes for the Aragonese dialect at the
Occitan pairs, and removed ca_valencia-es as suggested by Xavi (btw, are
you sure that ca_valencia-es and ca-es do exactly the same thing? They seem
to be using different files...).
------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and
threat landscape has changed and how IT managers can respond. Discussions
will include endpoint security, mobile security and the latest in malware
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
Apertium-stuff mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/apertium-stuff
