Hi Dave, I'll get started with that and let you know if I have any difficulties. Thanks for the test case pointers.
-Corey > -----Original Message----- > From: David N Bertoni/Cambridge/IBM > [mailto:[EMAIL PROTECTED]] > Sent: Monday, November 18, 2002 8:44 AM > To: [EMAIL PROTECTED] > Subject: RE: EXSLT and Xalan-C++ > > > > > > > Hi Corey, > > Yes, the ability to create XML nodes from extension functions is being > developed by someone else. I've haven't heard back from that > person in a > while, so I don't know what the state of that is. You might > want to start > on the encode-uri and decode-uri functions, then move on to > replace. By > the time you're finished, the other functionality may be integrated. > > As far as the test cases are concerned -- I'm not sure how > the Java tests > are being run these days. We share many of the same test > directories, but > they've done a lot of work on their testing, so I'm not sure > how many new > tests they use that we don't. In addition, we do conformance > testing on > Windows with a new executable that hasn't yet been ported to > any of the > Unix platforms. On Unix, we still use the old java test > harness, which can > be adapted through a configuration file to run any pair of > directories. In > any case, we should be able to share many stylesheets, > although those that > do testing of Java extension functions will not work. > > Take a look at xml-xalan/test/tests/conf and > xml-xalan/test/tests/conf-gold > as an example of a pair of directories. > > Dave > > > > > > > Corey Tripp > > > <CoreyT@InformativeRe To: > "'[EMAIL PROTECTED]'" <[EMAIL PROTECTED]> > > search.com> cc: > (bcc: David N Bertoni/Cambridge/IBM) > > Subject: > RE: EXSLT and Xalan-C++ > > 11/18/2002 08:13 AM > > > Please respond to > > > xalan-dev > > > > > > > > > > > Hi Dave, > > I had a chance to look over the EXSLT website and here's > what I found. > Of the modules that have implementations (common, math, sets, > and strings) > the strings module is the only one that is missing some > functions. Then > there are three modules that do not have any implementations at all > (dates-and-times, dynamic, and regular-expressions). > > To me I think it would be nice to finish off the strings > module and then > move on to the others. The remaining strings functions are > encode-uri, > decode-uri, tokenize, split, and replace. I think that > replace will be the > hardest out of this group, the rest seem pretty straigh > forward. The split > and tokenize functions need to return xml nodes of the tokens > found, my > understanding is that this functionality is being developed, is that > correct? > > As far as the test cases, does Xalan-C++ use the same smoketest > stylesheets that Morris said were in the test/tests/extensions/library > directory or is there another directory? I would like to see > some existing > examples so that I can get the structure and layout correct. > > -Corey > > > > > > > Hi Corey, > > > > Why don't you take a look at the EXSLT web site and compare what's > > implemented in Xalan-C to look for gaps? I'm a little fuzzy on the > > details, but there are quite a few things that aren't > > implemented yet, so > > why don't you look for something that looks interesting, and > > we can go from > > there. If you can come back with a potential list, we can > figure out > > priorities, etc. > > > > We should also spend some time devising basic stylesheets > > that will test > > the EXSLT functionality. We'll need at least a set of smoketest > > stylesheets before we start claiming we're really > implementing EXSLT. > > > > Dave > > > > > >
