lucene-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Chris Hostetter <hossman_luc...@fucit.org>
Subject Re: hudson failure
Date Wed, 10 Dec 2008 23:30:41 GMT

: Thanks. I gave excluding the class a shot. Clover is a code coverage analysis
: tool. http://hudson.zones.apache.org/hudson/job/Lucene-trunk/clover/

A random idea that occured to me as a result of this thread...

Given the density of the source code, and the size of the resulting byte 
code, it's not too suprising that instrumentation makes pushed hte class 
file over the limit.

it might be prudent to break the source up into multiple classes (one 
per Unicode block perhaps?) as a way to improve readablity (and to allow 
for instrumentation)

It could even allow for specialized TokenFilters: if you know for certain 
your source text only contains chars from the "Latin Extended-*" blocks you could 
write a short 4 line TokenFilter that leveraged only the code dealing with 
those blocks (instead of the full ASCIIFoldingFilter).


...just tossing it out there as a suggestion if someone with a similar 
usecase wants to scratch the itch.



-Hoss


---------------------------------------------------------------------
To unsubscribe, e-mail: java-dev-unsubscribe@lucene.apache.org
For additional commands, e-mail: java-dev-help@lucene.apache.org


Mime
View raw message