Gary Gregory wrote: > On Jan 27, 2011, at 17:43, "Jörg Schaible" <joerg.schai...@gmx.de> wrote: > >> Gary Gregory wrote: >> >>>> -----Original Message----- >>>> From: Jörg Schaible [mailto:joerg.schai...@scalaris.com] >>>> Sent: Thursday, January 27, 2011 12:21 >>>> To: dev@commons.apache.org >>>> Subject: Re: [codec] Large test data set! >>>> >>>> Hi Julius, >>>> >>>> Julius Davies wrote: >>>> >>>>>>> However, it will probably do no harm to ask at legal@. >>>>>> >>>>>> Do we actually have to distribute it? Maybe we can add is as zip to >>>>>> the Maven repo and use the dependency plugin to download and extract >>>>>> it on the fly. >>>>>> >>>>>> - Jörg >>>>> >>>>> >>>>> Another option: ask the copyright holder to relicense to us under >>>>> the ASL 2.0. That often works! >>>> >>>> Why do you want to include it into our Subversion? If it's downloade >>>> automatically for the test like any other dependency, what's wrong with >>>> it? >>> >>> It's not in a Maven repository, it's in a Zip file on SourceForge. >> >> ... and it is a defined process to put something into central. > > Yes that's true and a good point. But how do you read a data file from > java that works from the build and any IDE. That code would need to be > able to resolve the right location in the local M2 repo. This must have > happened before. Maybe there is a way to add the right dir to the > classpath and get the file as a resource through a class loader.
?!? You would use the dependency plugin in the generate-resources phase to extract the zip file into your local target directory where you then access the file(s) in your test. Quite a normal action with Maven. And regarding an IDE there's no difference from any other task that generates resources or additional source files. - Jörg --------------------------------------------------------------------- To unsubscribe, e-mail: dev-unsubscr...@commons.apache.org For additional commands, e-mail: dev-h...@commons.apache.org
