+1 to Jeremy's suggestion if it works. This is the strategy we've employed in WTP and have found it works well in many cases. I'd like to keep the URI resolver interface as simple as possible (meaning I'd rather they not be required to create and possibly open streams).
Lawrence "Jeremy Hughes" <[EMAIL PROTECTED]> Sent by: [EMAIL PROTECTED] 11/17/2006 11:47 AM Please respond to [email protected] To [email protected] cc Subject Re: URIResolver requirement to return InputSource On 11/17/06, Graham Turrell <[EMAIL PROTECTED]> wrote: > Could we reschedule a call for this afternoon? > > I'd like to understand the axis2 requirement better and also get a feel for > urgency since the forthcoming OASIS > resolver [1] can do all that's requested (if I read the requirements > right) witht he existing resolver interface. Having said that, the Simple > URI resolver, whilst designed primarily to > resolve to filepath URIs *should* resolve to any URI including jar, > provided the jar/aar is on the classpath. I'm happy to create & run some The jar/aar doesn't have to be on the classpath. This mechanism doesn't use classloaders. You should be able to put jar: in front of a URL that happens to be to a jar file and !/mypath/myfile at the end of the URL to address the file you want from inside the jar. > tests to verify this (or ortherwise). > > > [1] http://issues.apache.org/jira/browse/WODEN-72 > > Kind Regards, > > Graham. Cheers, Jeremy --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
