Thanks Vijay. I'm going to do the rebuild, but one thing I noticed is that YTEX is using the 2008 version of LVG and the old PAD term spotter looks like it is using the one from 2004!
I guess they can both use different versions (each one living in its own directory), but given how dated that LVG is (and the number of improvements) maybe we should upgrade and/or share LVG? What do others think? I'm likely going to do the rebuild of the ytex branch with an newer version of LVG and see how it goes. I'm not sure how the PAD term spotter would like that though - it looks like it is using MySQL instead of HSQL. -- John David Osborne Research Associate University of Alabama at Birmingham Biomedical Informatics Center for Clinical and Translational Science 1720 7th Avenue South Sparks Building, Suite 175 Birmingham, AL, 35294 On 2/11/14 8:19 PM, "vijay garla" <[email protected]> wrote: >Hi John, > >Thanks for this. I've updated the YTEXPipeline, fixed the lvg paths in >SetupAUIFirstWord. > >If you want to re-run SetupAUIFirstWord (not necessary unless you are >using >the stemmed words for dictionary lookup), just svn update, rebuild >ctakes-ytex-uima, and copy the jar to the lib dir of your ctakes install. > >Best, > >VJ > > >On Mon, Feb 10, 2014 at 6:25 PM, John David Osborne (Campus) ><[email protected] >> wrote: > >> These were the changes I made to get the YTEX pipeline working with LVG >> (2008). It looks like there were just a couple of spots with some old >> hard-coded paths in SetupAUIFirstWord.java that were appropriate to the >>old >> ytex directory structure. >> >> For now I have just swapped them out to fit with the new directory >> structure, but I suppose the correct fix may be to extract them out >> somewhere... In any case I don't have write privileges, some someone >>else >> may want to fix this (Vijay?) >> >> I also included the YTEXPipeline.xml descriptor file I fixed as well in >> case anybody needs it. >> >> -John >> >> >>
